sunday parenteral posters s1 to s33 final€¦ · parenteral nutrition s1–s33 1 critical care...

Clinical Nutrition Week 2016 1

CLINICAL NUTRITION WEEK 2016 ABSTRACTS

January 16–19, 2016, Austin, Texas

Austin Convention Center

POSTER PRESENTATIONS

Topic Poster Numbers Page

Parenteral Nutrition S1–S33 1

Critical Care S34–S82 45

Pediatric and Neonatal S83–S118 115

Parenteral Nutrition Posters S1–S33

S1—Nutritional Status of Patients Before, During, and After Receiving Parenteral Nutrition in Hospital Amnah M. Alhanaee, MD1; Johane Allard, MD, FRCPC1; Bianca Arendt, PhD, CCRP1; Anastasia Teterina, PhD1; Barbara Bielawska1; Sultan Alenezi1 1Gasteroentrology, Toronto General Hospital, Toronto, Canada Purpose: Malnutrition is a common problem in hospitalized patients. Parenteral nutrition (PN) is indicated when patients have temporary intestinal failure. While on PN, patients’ nutritional status is monitored by the PN team. Once PN is discontinued, nutritional status is monitored by the ward dietitian. The aim of this study was to evaluate the nutritional status of PN patients throughout the continuum of care in hospital.

Methods: Prospective cohort study of patients on medical, surgical and ICU wards who received PN as part of their standard clinical care. Patients had their nutritional status assessed at the start of PN, every 10 days while receiving PN, at the time of PN discontinuation, and then every 10 days until discharge. Nutritional assessment was performed using Subjective Global Assessment (SGA). Statistical analysis was conducted using SPSS Version 22.0. Descriptive statistics are presented.

Results: So far 19 patients completed the study with a mean age of 51.4 y (range 22–80). Thirty five percent of patients were on surgical wards, 20% in ICU and 45% were on medical wards. At PN start, 10% of patients were SGA A (well nourished), 70% were SGA B (moderately malnourished) and 20% were SGA C (severely malnourished). Average BMI at start of PN was 22.5 kg/m2 (range 17.1–30.1). Median duration of PN was 17.5 days (range 7–27). One patient continued PN throughout hospitalization and was discharged on home PN. At the end of PN, 15.8% of patients were SGA A, 57.9% were SGA B and 26.3% were SGA C. One ICU patient had worsening of SGA while on PN (B to C) and died 12 days after stopping PN. The remaining 18 patients were followed until discharge. Median length of stay was 30.5 days (range 15–71) and median time to discharge from end of PN was 5 days (range 0– 29). At discharge, 22.2% of patients were SGA A, 61.1% were SGA B and 16.7% were SGA C. Of these 18 patients, 15 had no change in SGA status from start of PN to discharge, 1 patient improved from SGA B to A while on TPN, and 2 patients improved SGA status by one class between end of PN and discharge. There were no significant correlations between SGA status and length of stay or duration of PN use.

Conclusions: PN is used for in-hospital nutrition support in a variety of clinical settings and in patients across the spectrum of malnutrition. In this small cohort, we have found that the majority of patients requiring PN in hospital are SGA class B and that PN use can lead to stabilization of nutritional status in the majority of patients, and improvement in nutritional status either during or after stopping PN in a minority. Future analysis plans to determine predictors of improvement and decline in nutritional status during and after discontinuation of PN. Financial support: Baxter.


S2— Impact of the Presence of a Baseline Clinical Condition Warranting the Cautious Use of Parenteral Nutrition on Subsequent Metabolic Complications Angela L. Bingham, PharmD, BCPS, BCNSP1,2; Brandon J. Patterson, PharmD, PhD1; Laura A. Siemianowski, PharmD, BCPS2; Colleen E. Smith, PharmD2; Richard J. Song, PharmD2; James M. Hollands, PharmD, BCPS1,2 1Philadelphia College of Pharmacy, Philadelphia, PA, USA; 2Cooper University Hospital, Camden, NJ, USA Purpose: The American Society for Parenteral and Enteral Nutrition Adult Nutrition Support Core Curriculum describes clinical conditions that warrant cautious use of parenteral nutrition (PN): hyperglycemia (glucose >300 mg/dL), azotemia (BUN >100 mg/dL), hyperosmolarity (serum osmolarity >350 mOsm/kg), hypernatremia (sodium >150 mEq/L), hypokalemia (potassium <3 mEq/L), hyperchloremic metabolic acidosis (chloride >115 mEq/L), hypophosphatemia (phosphorus <2 mg/dL), hypochloremic metabolic alkalosis (chloride <85 mEq/L). However, the authors acknowledge there is no evidence in the literature for the specific criteria suggested for the clinical conditions. Consequently, the purpose of this study is to determine the impact of the presence of a baseline clinical condition warranting cautious use of PN on the development of subsequent metabolic complications [acid-base disturbance, hypernatremia (sodium >150 mEq/L), hypokalemia (potassium <3 mEq/L), hyperchloremia (chloride >115 mEq/L), hypochloremia (chloride <85 mEq/L), hypophosphatemia (phosphorus <2 mg/dL), hypoglycemia (glucose ≤60 mg/dL), hyperglycemia (glucose >200 mg/dL), hypercapnia, hypertriglyceridemia (triglycerides ≥400 mg/dL), hepatic complications, metabolic bone disease, refeeding syndrome] in patients receiving PN.

Methods: Adult patients initiated on PN during hospitalization from May 1, 2014 to July 31, 2015 at Cooper University Hospital were included in this retrospective study. The primary objective was to determine the impact of the presence of a baseline clinical condition warranting the cautious use of PN on development of subsequent metabolic complications. Subgroup analyses were completed to determine the impact of each individual baseline clinical condition on development of subsequent metabolic complications. Pearson’s chi-square tests were performed for categorical variables. The a priori significance level was 0.05.

Results: There were 342 patients included in the analysis. The mean age was 61.7 years and 51% of patients were female. The mean duration of PN was 8.5 days and 97% received central PN. Overall, metabolic complications developed in 57% when receiving PN. Metabolic complications occurred more frequently in patients with a baseline clinical condition warranting the cautious use of PN than patients without one of these conditions (77% versus 53%, p=0.001). Subgroup analyses for the development of metabolic complications in patients with or without each individual baseline condition warranting cautious use of PN yielded these results: hyperglycemia (67% versus 58%, p=0.762), azotemia (50% versus 58%, p=0.816), hyperosmolarity (no occurrences), hypernatremia (93% versus 57%, p=0.007), hypokalemia (77% versus 57%, p=0.164), hyperchloremic metabolic acidosis (86% versus 57%, p=0.033), hypophosphatemia (77% versus 57%, p=0.059), hypochloremic metabolic alkalosis (100% versus 58%, p=0.230).

Conclusions: Hospitalized adult patients with a baseline clinical condition warranting cautious use of PN are more likely to subsequently develop a metabolic complication during their PN course. Baseline hypernatremia and hyperchloremic metabolic acidosis are associated with the development of metabolic complications. The low occurrence of baseline conditions warranting cautious use of PN may have limited our ability to demonstrate a statistically significant impact of other individual criteria. Baseline conditions warranting cautious use of PN should be recognized when initiating PN and prompt close monitoring to minimize metabolic complications.

Financial support: None.


S3—Impact of RD Order Writing Privileges on Parenteral Nutrition Administration: A 15 Year Review Anne Coltman, MS, RD, CNSC1; Sarah Peterson, PhD, RD, CNSC1; Diane Sowa, MBA, RD1; Omar Lateef, DO2 1Clinical Nutrition, Rush University Medical Center, New Lenox, IL, USA; 2Department of Pulmonary Critical Care Medicine, Rush University Medical Center, Chicago, IL, USA Purpose: The use of a nutrition support team has been shown to have a positive impact on administration of parenteral nutrition in hospitalized patients. One previous study examined the use specifically of registered dietitian (RD) order-writing privileges on use of PN, finding a significant decrease in administration of PN once privileges were established. The purpose of the quality improvement (QI) initiative is to further describe changes in PN use following initiation of RD order writing privileges.

Methods: The institution’s PN QI database was utilized to identify patients receiving PN prior to (2001–2004) and after (2006–2015) RD order writing privileges were established. The PN QI database contains nutrition-specific information on all patients receiving PN at the institution, including date of initiation, days receiving PN and appropriateness of PN (inappropriate reasons included poor oral intakes, chyle leak, pancreatitis, hyperemesis, malnutrition, mucositis, GI bleed or ileus for less than seven days). The sample was divided into three time periods for analysis: 2001–2005, 2006–2010, 2011–Present. Counts and means were used to describe the sample; one-way ANOVA was used to compare groups.

Results: New PN starts decreased from 488 in 2001 to 280 in 2014, representing a 43% decrease (Figure 1). Mean PN starts, average days of PN, and percent inappropriate starts all significantly decreased over the three time periods (Table 1).

Conclusions: RDs with order-writing privileges can decrease PN initiation, days of PN, and inappropriate use of PN in a hospital setting.


Table 1. Differences in PN Use by Group

Group 1 (2001–2005) Group 2 (2006–2010 Group 3 (2011–2015) p-value

New PN Starts (mean ± SD) 523.6 ± 32.2 458.8 ± 72.3 308.5 ± 48.4

Days on PN (mean ± SD) 10.1 ± 1.3 9.1 ± 0.5 8.3 ± 0.5 0.031

PN Starts Deemed Appropriate (%) 53% 33% 29%

S4—Accumulation of Rare Earth Elements in the Bones of Patients Following Long-Term Parenteral Nutrition Aubrey L. Galusha, PhD1,2; Pamela C. Kruger, PhD1; Patrick J. Parsons, PhD, FRSC1,2; Lyn Howard, BM, BCh, FRCP3 1Wadsworth Center, New York State Department of Health, Albany, NY, USA; 2Environmental Health Sciences, University at Albany, Albany, NY, USA; 3Division of Gastroenterology and Nutrition, Albany Medical College, Albany, NY, USA Purpose: Patients on long-term Parenteral Nutrition (PN) can be exposed to trace contaminants, e.g., aluminium (Al), in PN solutions. We previously reported excessive accumulation of Al in the bones of PN patients. Here we report additional analyses of the bones, obtained post-mortem, for Rare Earth Elements (REEs).

Methods: Trabecular bones were collected at autopsy from 7 PN patients who received therapy for 2–21 (ave. 15) years; control bones consisted of 18 samples from hip/knee replacement surgeries. Digested bones and PN solutions were analyzed for REEs by inductively coupled plasma tandem mass spectrometry. Non-parametric statistics were used to explore differences between patient and control bones. Bone REE content was normalized to known values for the earth’s upper crust.

Results: Up to 16 REEs were detected in patient bone samples. REE content varied from 3–16000 ng/g. In the control group the lighter REEs La, Ce, and Nd were detected in 16 of the samples and Pr in 15; the remaining REEs were detected in <5 of the samples. PN bones were significantly enriched (p <0.001–0.0004) with REEs compared


to controls. The REE Gadolinium (Gd) was highly enriched in two of the patient bones. Analysis of PN solutions revealed trace contamination with many REEs (10–3000 ng/L), and the REE pattern was similar to that found in PN bones, normalized to crustal content and illustrated in Figure 1.

Conclusions: These data indicate that patients on long-term PN are exposed to REEs at trace levels. The likely source of most REEs is trace contamination of PN solutions, although the specific source(s) remains to be elucidated. Excessive accumulation of Gd in two of the patients is more likely due to prior MRI studies. While the levels of REEs detected in patient bones are considered trace, or even ultra trace, their effect on human health, and bone health in particular is unknown.

Financial support: Funded in part by The Oley Foundation.

Figure 1. Mean REE content of PN bones (left axis) and PN solutions (right axis) normalized to REE values found in the earth's upper crust. Fifteen of the REEs detected are shown on the x axis in order of increasing ionic radius.

S5—Reducing PN Usage While Enhancing PN Effectiveness Through a Consolidated, Collaborative, Standardized Approach Larry Feinman, DO, FACOS2; Joyce HagenFlint, MSA1; Jeff Bush, PharmD, CPh3; Betty L. Davis, MS, RDN, LDN, CNSC1; Eric Wood, BS, RDN, LDN1; Susan Alverio, BS, RDN, LDN1; Grace Lee, MS, RDN, LDN1; Lisa Schaedler, BS, RDN, LDN1 1Food and Nutrition Services, Parallon, HCA West Florida, Largo, FL, USA; 2Administration, HCA West Florida Division, Tampa, FL, USA; 3Pharmacy Services, Parallon, HCA West Florida, Largo, FL, USA Purpose: To measure, evaluate and improve effectiveness of PN administration within the 16 hospitals of the West Florida Division of HCA.

Background: The 2009 Guidelines for the Provision and assessment of Nutrition Support Therapy in the Adult Critically Ill Patient, published by the Society of Critical Care Medicine (SCCM) and ASPEN are used in our facilities as a guide for evaluating nutrition support therapy. Over the past few years, the HCA West Florida Division clinical nutrition directors and pharmacy clinical director identified excessive use of parenteral nutrition (PN), frequently not meeting ASPEN evidence-based inclusion criteria. Several of our facilities had been monitoring appropriateness of PN therapy for a number of years without adequate sustained improvement. In January, 2014, the team initiated a division wide quality improvement project. As a result of initial review, patient safety issues and potential cost savings were identified.


Methods: Initial division baseline data collection for fourth quarter of 2013 included number of PN initiations and average duration of PN therapy. Beginning January 2014 the following data were collected in each hospital: number of PN initiations; duration of PN therapy, number of PNs meeting ASPEN inclusion criteria; and dollar cost of PN formulations. In an effort to standardize the PN process, a division wide policy and procedure was developed with input and engagement from dietitians, pharmacists and nursing leadership throughout the division. The new policy was also presented to the West Florida Division clinical nutrition council. All facility Chief Medical Officers (CMOs) were educated about the new policy via conference calls with the division CMO. The new policy was implemented at each facility in the division at different times during the first 3 quarters of 2014. Each hospital established a nutrition support team consisting of pharmacists and dietitians with distinct team roles for nurses, physicians and the IV team defined in the standardized policy. Toolkits were developed for training clinicians about the new policy and processes. According to the PN and nutrition support team policy, the physician may consult the nutrition support team to manage PN or write the orders him/herself. Daily collaboration between dietitians and pharmacists in each facility is expected regarding all PN cases, with nursing and physician involvement as needed. If PN does not meet the inclusion criteria, dietitian or pharmacist consults with the ordering physician to discuss the case and offer alternatives. Together with division Information Technology and Services (IT&S), a computerized physician order entry (CPOE) Nutrition Support Team Consult and PN order set was also developed. Nutrition support team consults and PN PICC line orders are concurrently transmitted to the clinical dietitians and pharmacists to evaluate PN compliance with ASPEN inclusion criteria prior to PN initiation. Decision support was built into CPOE order sets for patient safety enhancement. Facility by facility implementation of the CPOE order set began in early 2015.

Results: Over 18 months from January 2014 to June 2015, the West Florida Division reduced PN utilization by 33% (see Figure 1), while increasing adherence to evidence based criteria by 47% (see Figure 2) and enhancing patient safety. Supply cost savings of nearly $150,000 were realized (see Table 1), not including savings for labor and avoidance of central line infections and other metabolic complications.

Conclusion: The West Florida Division of HCA improved patient care and safety while reducing supply cost using a consolidated, collaborative standardized approach to PN management.


Table 1. PN data.


Figure 1. PN Usage

Figure 2. PN meeting evidence-based criteria.


S6—A Retrospective Review of Short-Term Parenteral Nutrition: Patient Characteristics and Outcomes Brittany Ray, PharmD1; Michelle Meyer, PharmD, BCPS, BCNSP1; Jennifer Hartwell, MD, FACS2 1Pharmacy, Grant Medical Center, Columbus, OH, USA; 2Trauma Services, Grant Medical Center, Columbus, OH, USA Purpose: Parental nutrition (PN) is not recommended for treatment that is anticipated to be less than five days in patients without severe malnutrition. However, many times this guideline is not followed and patients receive PN inappropriately, increasing their risk of complications such as catheter-related bloodstream infections and hyperglycemia. The goal of this study is to identify characteristics of patients that have received inappropriate PN therapy for ≤4 days compared to patients who have received PN therapy for ≥7 days with the hope of increasing ability to estimate duration of treatment and assist in deciding which patients to initiate on PN. Those patients that received PN therapy for 5–6 days will also be compared to determine if it differs from the ≤4 days study group and the ≥7 days study group.

Methods: The electronic medical record was reviewed for all adult patients ≥18 years of age who were admitted to an OhioHealth facility (Doctors Hospital/DH, Grant Medical Center/GMC, Riverside Methodist Hospital/RMH, Dublin Methodist Hospital/DMH, and Grady Memorial Hospital/GMH) for any reason, and received PN from July 1, 2013 through June 30, 2014. The following data was collected: age, gender, weight, height, body mass index, APACHE II score, albumin, pre-albumin, hospital site, date of admission, length of stay, days on ventilator, days of peripheral line insertion, days of central line insertion, clinical indication for PN (small bowel obstruction, ileus, GI fistula, inflammatory bowel disease, malabsorption, and other), day of admission that PN was initiated, duration of PN, need for PN resolved, central venous access solely for PN, enteral access contraindicated, enteral access attempted prior to PN, surgical procedures performed, hyperglycemia, hypertriglyceridemia, catheter-related bloodstream were then split into the following groups: those who received PN for ≤4 days (group 1), those who received PN for 5-6 days (group 2), and those who received PN for ≥7 days (group 3). The groups were evaluated to identify differences and/or commonalities between them.

Results: Results are preliminary for 250/368 patients. Table 1 shows that the outcomes and characteristics studied were not statistically significant between the groups except the two outcomes that were statistically significant: patients who experienced hyperglycemia and patients who received corrective insulin. This outcome showed that more patients in the ≥7 days group experienced more hyperglycemia and received more corrective insulin than the ≤4 days group. This demonstrates that the longer a patient receives PN the more at risk they are for experiencing hyperglycemia and receiving insulin to correct their hyperglycemia. Two graphs are displayed; Figure 1 shows the amount of PNs started at each study site within each group and Figure 2 shows the number of PNs for each indication divided between the study groups. Riverside Methodist Hospital is the largest hospital within OhioHealth explaining why they might have the most PNs at their site. However the amount of patients that received PN for 4 days or less is something that can be improved upon. In Figure 2, "other" is the largest category in all three groups with "small bowel obstruction" with the second most.

Conclusions: Parenteral nutrition puts patients at risk for complications like hyperglycemia requiring correction with insulin as seen in this study. The results reflect that many patients who receive PN experience complications of therapy and this should not go unnoted. When initiating practitioners ordering PN should reflect on the indication for PN and the risk and complications that patients are subjected to before initiating PN.



Table 1. Results of Study Categories Compared Among Study Groups.


Poster of Distinction

S7—To Swim or Not to Swim: Survey on Swimming Practices With a Central Venous Catheter Carol Cheney, RN, CNSC1; Kristi Griggs, RN, CNSC1; Stacie Brandt, RN1; Kelly Sutich, RN, CRNI1; Lloydette Zieman, RN, CRNI1; Deborah Pfister, MS, RD, CNSC1 1ThriveRx, Cincinnati, OH, USA

Purpose: Quality-of-life issues are an important concern for those who are on long-term home parenteral nutrition (HPN). One issue verbalized by consumers involves swimming while having a central venous catheter (CVC) in place. Requests for guidance on safe procedures for this activity are directed to this provider of long-term nutrition support. As a result of very limited published research and the lack of evidence-based recommendations in this area, a survey was designed to identify attitudes and practices of the HPN consumer surrounding swimming with a CVC.

Methods: To determine if a correlation exists between swimming and the incidence of line infection, an anonymous electronic survey was distributed randomly to the HPN community via social media. The survey consisted of 11 questions pertaining to the practice of swimming with a CVC and specific practices of CVC care. The survey also queried regarding practice awareness by the consumer’s medical team.


Results: Of the 182 consumers with a CVC who were surveyed, 158 (87%) reported that swimming was important to them, and 101 (55%) reported actually swimming with a CVC. Of the CVC swimmers, 71 reported swimming in a pool; 24 reported swimming in the ocean; 5 reported swimming in a lake; and 11 reported swimming in all 3 media. As for line care practice, 62 (61%) of the CVC swimmers wore CVC-protective covering during swimming, with 16 of the 62 wearing wetsuits. Thirty-six (36%) reported not wearing any protective covering, and 2 did not respond to the question. Sixty-one (60%) CVC swimmers reported changing their dressing immediately after swimming; 36 (36%) reported changing the dressing as needed; and 2 did not respond to the question. Ninety-four (93%) CVC swimmers reported having 0 CVC infections after swimming; 6 (6%) reported 1 CVC infection; and 1 (1%) reported 2 CVC infections. Of those reporting infections, all but 1 reported wearing a CVC-protective covering, and 6 reported changing their CVC dressing immediately after swimming. Of the 101 CVC swimmers, 63 (62%) reported that their medical team knew that they swam, with 43 (68%) of these informed teams allowing swimming and 20 (32%) of these teams prohibiting swimming. Thirty-eight (38%) CVC swimmers reported that their team was not aware of the activity.

Conclusions: Consumer practice of swimming with a CVC varies by swimming location, care of the CVC during and after the activity, and notification of the medical team regarding swimming practice. The ability to swim was reported as being an important contribution to quality of life in the HPN consumer. Further research to explore documented CVC infections relative to swimming practice will assist in development of standardized practice.

Financial support: BioRx.

S8—Effect of Manganese Removal From Home Parenteral Nutrition Solutions on Whole Blood Levels and Magnetic Resonance Imaging of the Brain: A 5-Year Cohort Study Jennifer Jin, MD, FRCPC1; Olivia Saqui, RN, BScN, MAEd1; Johane P. Allard, MD, FRCPC1 1Gastroenterology, University of Toronto, Toronto, Canada

Purpose: Manganese (Mn) is included in many commercial pre-mixed multiple trace element (TE) additives for home parenteral nutrition (HPN). However, there is a risk of over-supplementation because of inherent levels of Mn due to contamination of many parenteral nutrient additives. Over-supplementation can produce Mn toxicity with neurologic symptoms and basal ganglia deposits on brain magnetic resonance imaging (MRI) and could theoretically contribute to PN associated liver disease (PNALD). In 2009 we reported that whole blood Mn levels were above the upper limit of normal (ULN) in a sample of 16 HPN patients with 81% having MRI findings. Subsequently, we removed Mn supplementation from all our HPN patients and performed yearly monitoring. We present a 5-year follow-up here.

Methods: This is a prospective cohort study on 11 of the 16 patients who remained alive on HPN. All patients had Mn removed from their PN over the 5 years and had yearly monitoring of blood Mn levels. Eight of these patients had a repeat MRI to compare with the original imaging for resolution of basal ganglia deposits. Patient demography, clinical history and bloodwork were recorded.

Results: Main results are shown in Table 1. Five out of 6 patients who initially had elevated Mn levels had normal levels 5 years after Mn was removed from PN. All patients who had Mn levels measured serially had a decrease in levels following Mn removal. The mean percent decrease of Mn was 38.1% (range 10.1%–53.8%). Two patients had Mn levels above the ULN despite the absence of Mn supplementation in their PN prescription. Six out of 8 patients who had repeat MRIs had complete resolution of basal ganglia deposition. Of the 2 patients with persistent MRI abnormalities, both had current Mn levels within the upper limits of normal. One patient had an undetectable Mn level during annual monitoring in the intervening years; Mn was re-added in PN and removed again the following year when levels became close to the ULN. There was a trend in the decrease of ALT and ALP but this did not reach significance.

Conclusions: Removal of Mn as additive in HPN solutions resulted in resolution of MRI abnormalities in most patients. Liver enzymes also improved. Over 5 years, all patients except for one, maintained normal blood Mn levels. Therefore, Mn levels should be monitored regularly and supplementation should be individualized.



Table 1. Results.

Patient Number

MRI findings

Mn level in

2009 (nmol/L)

Mn level in

2015 (nmol/L)

Date of Mn

Removal

Mn Level Percent

Change (%)

ALT in 2009(U/L)

ALT in 2015 (U/L)

ALP in 2009 (U/L)

ALP in 2015 (U/L)

1 n/a n/a 537 Jan

2013 n/a 248 24 164 102

2 n/a 286 n/a Jan

2013 n/a 29 35 169 109

3 Resolved 353 213 Jul

2009 -39.7 34 24 210 174

4 Resolved 320 187 Jan

2010 -41.6 33 18 245 223

5 Resolved 435 218 Jan

2010 -49.9 14 17 86 87

6 Abnormal 386 233 Jan

2010 -39.6 64 35 170 69

7 Resolved 251 116 Mar 2012

-53.8 13 19 113 110

8 Resolved 109 98 Oct

2013 -10.1 118 36 203 131

9 n/a 326 197 Mar 2012

-39.6 43 13 263 166

10 Resolved 1365 644 Nov 2009

-52.8 17 23 253 334

11 Abnormal 222 187 Jan

2013 -15.8 24 25 80 106

n/a: result not available.

S9—The Profile of Adult Patients Receiving Home Parenteral Nutrition (HPN) and Evaluation of HPN Use in Relationship to Current Nutrition Support Guidelines Kelly Kinnare, MS, RD, LDN, CNSC1; Mary Englert, MS, RD, CSO, LDN, CNSC1; Megan Paul, RD, LDN, CNSC1; Hannah Hilado, RD, LDN, CNSC1; Lindsay Eisenhut, MPH, RD, LDN, CNSC1 1Optioncare, Wood Dale, IL, USA Purpose: Parenteral Nutrition (PN) is the intravenous provision of nutrients given to patients who cannot be fed via the gastrointestinal (GI) tract. In 2002, the American Society for Parenteral and Enteral Nutrition (A.S.P.E.N.) published evidence-based guidelines for the use of PN to assist clinicians on decisions about the appropriate use of PN. According to these guidelines, PN should be used when the GI tract is nonfunctional and enteral nutrition (EN) is not feasible. Typically, home parenteral nutrition (HPN) is prescribed and monitored by nutrition support certified clinicians, who utilize A.S.P.E.N. guidelines as the foundation for decision making. However, PN is often initiated in the inpatient setting by non-nutrition support certified professionals, which leads to concerns for higher inappropriate HPN use. The purpose of this study was to evaluate the profile of adult HPN patients and determine whether current reasons for HPN correlate with A.S.P.E.N. guidelines among patient receiving HPN from a large infusion center over a five year period.

Methods: Patients 21 years of age and older and started on HPN between January 1, 2010 and December 31, 2014 were evaluated. Data were collected on patient demographics, HPN clinician management type, total number of PN days, location of HPN initiation, insurance type and reason for PN. Disease states in agreement with guidelines


include: bowel dysmotility/perforation/rest, fistula (unable to feed via enteral route), intolerance to EN/no enteral access, intractable vomiting/diarrhea, malabsorption, bowel obstruction and short bowel syndrome (SBS). Conditions considered not in agreement included: cancer/malnutrition, hyperemesis, inadequate oral intake/weight loss, pancreatitis and partial bowel obstruction/fistula with ability to feed via enteral route. Comparisons of PN use in relationship to A.S.P.E.N. guidelines were analyzed over the five year study period.

Results: A total of 505 patients were started on PN during the review period. The average number of days a patient required PN decreased by 33%. Patient’s HPN were managed by a Registered Dietitian (RD) 91% of the time in 2014, which represented an 18% increase from 2010. There was an observed decrease of 14% of private insurance and an increase of 14% in public insurance (Table 1). The most common medical conditions requiring HPN were obstruction (18%), inadequate oral intake/weight loss (13%) and fistula with inability to feed via enteral route (12%). A decrease in HPN use was seen over the years in patients with inadequate oral intake/weight loss, intractable vomiting/diarrhea and pancreatitis while an increase in use was observed in patients with intolerance to EN/no enteral access, hyperemesis, and partial bowel obstructions/fistulas with ability to feed via enteral route (Table 2). The percentage of patients on HPN that met A.S.P.E.N. guidelines increased from 63% to 72% from 2010 to 2013; however, this trend was not observed to continue into year 2014. Overall, an average of 65% of cases met A.S.P.E.N. guidelines while 35% did not meet A.S.P.E.N. guidelines (Figure 1).

Conclusions: The total days on HPN decreased as RD management of HPN increased. Also noted was an increased shift towards public insurance which warrants compliance with guidelines for reimbursement. Several conditions not in agreement with A.S.P.E.N. guidelines decreased while others increased. Over the five year review there was an increase in patients receiving HPN in agreement with A.S.P.E.N. guidelines, however greater than one third of all patients on HPN could benefit from EN. There is a need for further research on utilizing EN over PN for several medical conditions and education for non-nutrition support professionals on the validation of HPN use.



S10—Medicare Part B and Home Parenteral Nutrition Coverage: How Often Is Objective Evidence Available? Lisa M. Kinder, RD, CNSC; Holly Flax, MS, RD, CNSC1 1Soleo Health, Sharon Hill, PA, USA Purpose: Infusion providers are challenged with determining home parenteral nutrition (HPN) coverage for patients with Medicare Part B as payor. Medicare Part B covers HPN therapy under the prosthetic device benefit. Along with the test of permanence, Medicare requires specific diagnoses with documented objective evidence within the medical record to support these diagnoses, in order to meet coverage criteria. Objective evidence can include diagnostic tests and procedures. The infusion providers are responsible to determine if the objective evidence clearly supports the diagnosis to meet coverage criteria. The purpose of this review was to summarize how often tests and procedures were performed that supported the diagnoses to meet Medicare HPN coverage criteria.

Methods: This home infusion provider reviewed medical record documentation for HPN orders received between January 1, 2015 and August 15, 2015, with Medicare Part B as payor. The following data was gathered: 1. Number of patients that had diagnoses within Medicare coverage criteria A through H. 2. Number of patients that had diagnoses within the criteria that also had an objective test performed. 3. Number of patients that had diagnoses within the criteria, objective tests performed, with test results clearly supporting the diagnosis.

Results: Medical records for 21 patients were reviewed. Twelve (57%) had diagnoses within the criteria, while nine (43%) did not. Of the twelve medical records that documented diagnoses within the criteria, seven had objective tests, with three providing clear evidence of diagnoses (Figure 1): criteria A (massive small bowel resection), and criteria C (regional enteritis and pancreatitis). The operative report for the massive small bowel resection clearly indicated that the length of small bowel remaining beyond the ligament of Treitz was less than five feet. The exploratory laparotomy for regional enteritis clearly documented a severely inflamed small bowel. The CT scan for pancreatitis clearly concluded moderately severe pancreatitis. The other four medical records that had objective tests did not have results to clearly support the diagnoses. CT scans for these four patients failed to support the diagnoses of pancreatitis, complete small bowel obstruction (two patients), and enterocutaneous fistula. In total, three (14%) of the 21 medical records reviewed provided clear objective evidence to support the diagnoses within the Medicare HPN coverage criteria. These patients also met the other required criteria and were considered covered for HPN therapy. Trends of the nine medical records that did not have diagnoses within the criteria included colitis, PEG tube leak, abdominal bloating, small bowel pneumatosis, dysphagia, and four cancer diagnoses (Figure 2).

Conclusions: One of the challenges in qualifying patients for HPN coverage within Medicare Part B criteria is the need for objective evidence to support the diagnoses. Forty-two percent of the medical records reviewed did not include objective tests, although the diagnoses were within the criteria. This can result in no HPN coverage, and may potentially lead to an extended hospital length of stay and/or costly long-term facility admission for continuation of TPN therapy. This infusion provider promotes awareness of the need for objective tests by providing the care team with: criteria reference lists detailing required documentation, a criteria focused continuing education program, and ongoing education to all clinicians involved with HPN coordination. If objective tests are available, infusion providers will be better equipped to determine HPN coverage, which may help reduce costs related to extended hospital stays for therapies that can be appropriate to administer in the home, such as parenteral nutrition. Financial support: None.


Figure 1. Results of medical record review.

Figure 2. Trends of criteria and other diagnoses.

S11—Excessive Potassium Requirements in Parenteral Nutrition: High Need or Low Compliance? Lynda N. Newman, MSN, RN, ACNS-BC, NP-C1; Sandra I. Austhof, MSN, RD, CNSC2; Denise M. Konrad, RD, CNSC2; Ronelle Mitchell, MSN, RD, CNSC2; Jamie Davila, RD, CNSC2; Donald F. Kirby, MD, FACP, FACN, FACG, AGAF, CNSC, CPNS1 1Digestive Disease Institute, Cleveland Clinic, Cleveland, OH, USA; 2Center for Human Nutrition, Cleveland Clinic, Cleveland, OH, USA Purpose: Introduction: Our Home Parenteral Nutrition (PN) program located in a large urban quaternary care center and serving patients across the US, experienced a number of incidences where the potassium requirements in a patient’s formula exceeded the 10 mEq/hour threshold of the serving pharmacy (120 mEq total in a 12 hour run). Other pharmacies will allow up to 20 mEq/hour (240 mEq total in a 12 hour run) but express concern when the requirement is considerably above 2 mEq /kg. We decided to determine how commonly the issue of excessive potassium requirement occurs.

Methods: We reviewed PN prescriptions of all active patients that were in our program between December 1, 2014 and January 31, 2015. The formulas of 264 patients were analyzed for: the diagnosis for PN; volume of fluid needed; hours of infusion; number of infusion days per week; and the total doses of potassium, magnesium, and octreotide doses. The diagnoses for PN were categorized as: fistula (10%), GI dysmotility (16%), intestinal malabsorption (16%), obstruction (15%), and short bowel syndrome (43%).

Results: Of the 264 patients whose formulas were reviewed, 79% used less than 10 mEq/hr of potassium, 38 (14%) used 10-14 mEq/hr; and 13(5%) used 15-19 mEq/hr. Only 5 patients (2%) infused at a KCL of 20 mEq/hr or greater, but 4 of those people were infusing for less than 12 hours per night. Lengthening the infusion time to 12 hours moved all but one of those patients into the category of 16-19 mEq/hr. The patients with diagnoses of short bowel


syndrome, obstruction and intestinal malabsorption were more likely to use > 10mEq/hr of KCL (27%, 25%, and 20 % respectively). Exploration of each patient’s situation revealed that some cases of high potassium requirement were related to high ostomy output and likely high potassium losses. Some of the cases of high potassium requirements were found to be related to inconsistent infusion, as documented by pump reports. Pump reports had been requested on only 1 of the patients in the > 20mEq/hr KCL/hr range and that patient was revealed to have only been infusing 48% of the time. Pump reports had been requested on 4 of the 13 patients who required between 15 and 19 mEq/hr of KCL. Two of those patients were infusing 95-100% of the time. The other 2 were infusing 50-78% of the time. Fifty patients (19%) had Octreotide in their prescriptions. Most patients who used octreotide (72%) were on PN for short bowel syndrome or fistula. 82% of the patients using octreotide required < 10mEq/hr of potassium. Magnesium requirements never came close to exceeding the 1 gm/hour safety threshold in any patients.

Conclusions: Excessive potassium use can be a red flag for poor compliance with the PN regimen. Because pump reports are time-consuming to review, it is important to have criteria for suspecting non-infusion. However, there are patients who truly have high losses and need high potassium replacement. Attention should be paid to the hours of infusion to ensure that the infusion time is long enough to keep potassium to < 20 mEq/hr. Future research to quantify potassium losses in high output ostomies could be useful in determining which patients have actual high potassium needs. In addition, it was notable that most high output patients whose prescriptions included octreotide did not have high potassium needs, but establishment of a causal relationship would require further investigations. Financial support: None

S12—Central Venous Catheter Life in Very Long-Term Home Parenteral Nutrition Patients Marianne Opilla, RN, CNSC1; Rodney Okamoto, RPh1 1Nutrishare, Elk Grove, CA, USA Purpose: Patients with debilitating gastrointestinal conditions may require home parenteral nutrition (HPN) for a lifetime. HPN is infused through a central venous catheter (CVC). HPN complications associated with the CVC often result in removal and replacement of the device, which over time, may lead to devastating loss of venous access sites. The CVC is considered a lifeline, and maintaining a trouble-free CVC is an important goal for HPN patients. The aim of this study was to examine a group of very long-term HPN patients and report their longest catheter life history.

Methods: A retrospective chart review for CVC history was completed for adult and pediatric HPN patients from one home infusion pharmacy. Data collection included age, gender, diagnosis for HPN, years on HPN, CVC type, and the longest number of days a CVC remained in place for each patient. Patients on HPN for 3 years or more, and with a CVC in place for 2 years or more were included in the cohort.

Results: A total of 103 patients were identified: 91 adults and 12 pediatric (<18 years). Females represented at 60% and the average age was 49 years (range 9-90 years). The primary diagnosis was short bowel syndrome (70%) followed by chronic intestinal pseudo-obstruction (27%) and radiation enteritis (3%). The total HPN years for all patients was 1995, with an average of about 19 years (range 4 -39 years) on HPN. The combined total CVC days for each patient's longest lasting CVC were 291,744 with an average of 2832 days/7.7 years for the entire group. Adults had 270,330 CVC days (range 832-7490 days) averaging 2970 days/8 years. Pediatric patients had 21,414 CVC days (range 813-4686 days) averaging 1784 days/4.88 years. There were 81 tunneled CVCs (251,115 days). The majority of the tunneled CVCs were single lumen, open ended, silicone devices (88%). There were 16 implanted infusion ports (31,236 days) and 6 peripherally inserted central catheters (9393 days). Patients were further divided into 3 groups based on CVC years. Twelve tunneled CVCs lasted 15 + years, 13 tunneled CVCs and 1 implanted infusion port 10-15 years, and 27 tunneled CVCs, 6 implanted ports, and 1 peripherally inserted central catheter lasted 5-10 years. The 25 tunneled CVCs lasting 10 years or greater were all single lumen, open ended, silicone devices.

Conclusions: In this cohort of adult and pediatric HPN patients we have shown that CVCs can be safely maintained for many years without serious complications that require removal and replacement. Tunneled CVCs, particularly the single lumen, open ended, silicone design, last significantly longer than implanted ports and peripherally inserted central catheters, and should strongly be considered for very long-term HPN patients. Pediatric CVC complication rates are generally reported as higher than adult's, but in this group the pediatric patients experienced good CVC


longevity. Through education on CVC care and clinical support, long-term HPN patients can successfully preserve their lifeline for many years. Financial support: None.

S13—Creation of a Virtual Nutrition Support Team to Focus on Patient Safety and Improve Quality of Care in Parenteral Nutrition Patients Michelle Meyer, PharmD, BCPS, BCNSP1; Jennifer Hartwell, MD, FACS5; Theresa Cattell, RD, CNSC4; David Ashman, RD7; Brad Babines, RD, CNSC4; Amy Beatty, PharmD, BCPS9; Roopinder Bhullar, RPh8; Cato Coleman, RD4; Kaitlyn Donaldson9; Lois Frame, RD2; Penny Masters-Boes, RD, CNSC4; Julie McGlone, RPh, MS9; Gretchen Roeger, PharmD, BCPS9; Abby Tyson, PharmD, BCPS3; Gina Wowk, PharmD1; Tammy Young, PharmD6 1Pharmacy, Grant Medical Center, Columbus, OH, USA; 2Nutrition Services, Grant Medical Center, Columbus, OH, USA; 3Pharmacy, Riverside Methodist Hospital, Columbus, OH, USA; 4Nutrition Services, Riverside Methodist Hospital, Columbus, OH, USA; 5Surgery, Grant Medical Center, Columbus, OH, USA; 6Pharmacy, Doctor's Hospital, Columbus, OH, USA; 7Nutrition Services, Doctor's Hospital, Columbus, OH, USA; 8Pharmacy, Dublin Methodist Hospital, Dublin, OH; 9Corporate Pharmacy Services, OhioHealth, Dublin, OH, USA Purpose: OhioHealth is a large multi-center health system located in central Ohio that consists of acute care and tertiary care hospitals. The parenteral nutrition (PN) management lacked consistency in clinical indication and competency in clinicians responsible for PN at each care site. In February 2015 the system implemented a new electronic health record (EHR). A need was identified to standardize care across the system with experienced clinicians who would be responsible for daily PN ordering. This presented an opportunity to create the first virtual patient care team for parenteral Nutrition Support.

Methods: Developed a model for Nutrition Support Team (NST) to manage the parenteral nutrition for OhioHealth patients using the EHR to virtually care for patients at different care sites. The NST identified gaps in knowledge and skills and provided education to all clinicians to ensure clinical competency throughout the system, using board-certified clinicians as the core of the team. Competency was assessed by the NST medical director for all team members. Developed practice guidelines to provide clinicians with consistent methods for carrying out patient care activities related to ordering and providing PN. All PN orders are entered by NST members only and transmitted directly to pharmacy. NST manages electrolyte and acid-base disorders related to PN in consultation with the primary team. Team members monitor for continued clinical indication and progress toward transition to enteral nutrition (EN) and/or oral diet. The NST is responsible for conducting chart audits, educating physicians and other clinicians, achieving goals of cost control, managing limited resources and improving patient outcomes.

Results: A multi-disciplinary team consisting of physician, dietitians, and pharmacists was created and clinical practice guidelines and policies were approved by Pharmacy and Therapeutics and Medical Executives Committees. A.S.P.E.N. best practice clinical and safety recommendations were implemented if not already in place. Each day one dietitian and one pharmacist are responsible for monitoring and ordering parenteral nutrition for patients at all hospitals within the EHR system. Physician input is available from the medical director as needed. Daily conferences between dietitian and pharmacist were instituted to decide on therapeutic plans. Care was taken to preserve communication between NST clinicians and the unit clinicians when physical presence was beneficial to patient care. Discharge orders were standardized to improve continuity of care and maintained in the EHR. Data across 6 hospitals was collected and included in analysis for appropriateness of PN therapy (designated by dietitian review in concurrence with A.S.P.E.N. recommendations) and maintaining acceptable glucose control (identified as maintain blood sugars 65-180mg/dL in at least 80% of glucose checks). Baseline results demonstrated orders were deemed appropriate by indication in 68% of patients, and 62% of patients met the standard for acceptable blood glucose control. After implementation of the NST, orders were deemed appropriate by indication for 97% of patients, and 74% of patients met the standard for acceptable blood sugar control.

Conclusions: A virtual team concept was applied to management of PN patients in a six hospital health care system to optimize care of PN patients by reducing inappropriately prescribed therapy and improving blood glucose control.



S14—Dietitian Collaboration With Vascular Access Nurses Prior to Peripherally Inserted Central Catheter Placement for Parenteral Nutrition: A Partnership That Can Prevent Unnecessary Central Lines Sharon Tate, MS, RD, LD1; Michele Creech, RD, LD, CNSC1; Pamela Perkins, RD, LD, CNSC1 1Food and Nutrition Services, Maine Medical Center, Portland, ME, USA Purpose: Peripherally inserted central catheters (PICCs) are essential in the delivery of life saving Parenteral Nutrition (PN); however, central lines carry a high risk of infection leading to increased mortality, length of stay and associated costs. Historically, the physician orders a consult for PICC placement as well as nutrition consult, however there has been no coordination as to the order in which these consults occur. Ideally, the nutrition consult should occur prior to PICC placement as Registered Dietitians (RDs) are best equipped to apply best-practice guidelines to determine if PN is indicated.

Methods: A quality improvement project was initiated in which the Vascular Access Team worked in collaboration with Registered Dietitians (RDs) to ensure that nutrition consults were completed by RDs prior to PICC placement. All adult (18 years of age or older) patients with an order for PICC line insertion were triaged by the Vascular Access Registered Nurse (RN), if the PICC was ordered for the sole purpose of PN, the RN assured that a nutrition consult was ordered and alerted the RD. The RD, using an algorithm developed with the Nutrition Support Team and based upon best-practice guidelines, then assessed the appropriateness of PN. If the RD found that the criteria for PN were not met at that time, a call was placed to the referring physician to discuss alternative nutrition recommendations. A note was also written in the electronic medical record regarding appropriateness of PN and communicated to the Vascular Access RN. The decision to proceed with the PICC line and initiation of PN was then left to the physician.

Results: The four month project resulted in 37 patients with consults for PICC line for the purpose of PN only. In all 37 patients, the appropriateness of PN was assessed prior to PICC placement and within our goal of less than four hours of consult to avoid delay in patient care. In total, seven patients (19%) were deemed inappropriate for PN. Of these seven patients, six had a functioning GI tract and one patient who was well-nourished on admission had been NPO for less than seven days. PN was not initiated and a PICC line was not placed in six of the seven patients deemed inappropriate by the RD, a PICC was placed in the seventh patient on the following day when the indication was changed to include IV antibiotics.

Conclusions: Registered Dietitians are uniquely qualified to assess the appropriateness of PN and thus play a key role in decreasing the placement of unnecessary central lines; collaboration between the RD and the Vascular Access RN is essential in this process.


S15—Improvement in Quality Metrics at a Large Academic Medical Center After the Implementation of an Education Program and the Nutrition Intervention Team Stephanie B. Cutrell, MS, RDN, LDN, CNSC1; Dena McGhee, RDN, LDN, CNSC2; Niko Nuguid, PharmD, CNSC1; Monica McPherson, RPh, CNSC1; Anissa Cavaliere, PharmD1; Laura E. Matarese, PhD, RDN, LDN, FADA, CNSC, FASPEN3 1Pharmacy, Vidant Medical Center, Greenville, NC, USA; 2Nutrition, Vidant Medical Center, Greenville, NC, USA; 3Internal Medicine at Brody School of Medicine, East Carolina University, Greenville, NC, USA Purpose: Vidant Medical Center (VMC) is a 900 bed academic medical facility providing tertiary health care in acute, intermediate, rehabilitation and outpatient services. In 2013 we identified several areas for improvement regarding the initiation and compounding of parenteral nutrition (PN). The objectives of this study were to: 1) measure improvements in the nutrition support practice after in-depth nutrition education and the implementation of the Nutrition Intervention Team (NIT), 2) assess compliance with the A.S.P.E.N. Clinical Guidelines, and 3) identify areas for quality improvement initiatives and interventions.

Methods: An IRB-approved prospective study was conducted over a 3-month time period in 2014-2015, where data relating to nutrition support practice were collected in adults started on PN. Decisions which required clinician judgment were guided by the A.S.P.E.N. clinical guidelines. Study results were then compared to a previous study conducted in 2013 which assessed the same variables prior to the implementation of an educational program and the NIT.


Results: Compared to 2013, the appropriate use of PN, based on indication, increased by 25% from 71% to 89%. PN was indicated most often for GI Ileus (25%, n=22) and GI obstruction (19%, n=17). PN was administered in (17%, n=15) patients due to inability to gain enteral (EN) access in potential EN candidates. The length of time on PN decreased from 11 days to 9.9 days. Hospital length of stay (LOS) was reduced from an average of 27.45 days (range=2-186) to 20.4 days (range=3-98). PN continued to be prescribed most often by Surgery. Of the 88 patients started on PN, 25% (n=22) were transitioned to EN, 16% (n=14) were discharged from the hospital on PN and 11% (n=10) expired. PN was ordered 80% (n=70) of the time by a clinician certified in nutrition support and 20% (n=18) by a non-certified clinician. Compliance with the Safe Practice Guidelines for Compounding PN was achieved 100% of the time.

Conclusions: As a result of our quality improvement initiatives, including a robust educational series and implementation of the NIT, the appropriate use of PN increased and compliance with the Safe Practice Guidelines was achieved. A new area for quality improvement was identified regarding the need for gaining EN access in all patients who are EN candidates. This further highlights the importance of continuous quality improvement initiatives.


S16—The Prevalence and Factors Associated With Line Sepsis in Hospitalized Patients Receiving Parenteral Nutrition: A Single Site Observational Study Sultan Alenezi, RD, MSc1; Daniela Adjemian, MD2; Johane P. Allard, MD, FRCPC2 1Nutritional Sciences, University of Toronto, Toronto, Canada; 2Medicine, University Health Network, Toronto, Canada Purpose: To investigate the prevalence and risk factors for line sepsis in patients receiving PN during their stay at the Toronto General Hospital.

Methods: This is a prospective cohort study on 149 hospitalized surgical, medical and ICU patients who had their clinical data collected in an ongoing inpatient PN registry. Patients who required PN as part of their clinical care, were followed from the start of PN until PN was discontinued. Data collected included information such as patient demography, type of ward, estimated energy requirements, total energy received from PN, PN related laboratory results, and catheter-related infections. Descriptive statistics were performed, such as rate per 1000 catheter-days, frequency tables, median, means, and standard deviations. The infection rates for 1000 catheter-days were calculated for categorical variables. Poisson regression was used to investigate factors associated with line sepsis. Statistical analysis was conducted using SAS version 9.4. A p value <.5 was considered statistically significant.

Results: The mean age was 55.9±16.5 years (n=148) with males and females representing about 52% and 48% of patients, respectively. About 17% of the patients were admitted in medical wards, whereas 60% and 23% of the patients were in surgical wards and ICUs, respectively. Almost 10% of patients (n=14) had confirmed line sepsis on PN while an additional 40% had suspected line sepsis anywhere from 1 to 6 episodes per patient. Approximately 16% of the patients had single lumen and 84% had multiple lumens (2 or more) in their central venous catheters. The mean PN days was 19.22±20.57 days (n=143) with an average amount of PN dextrose of 185.04±60.80 grams (n=141) and PN lipids of 40.4±9.34 grams (n=141). The mean blood glucose level was 7.60±3.62 mmol/L (n=104). The infection rates of either confirmed or suspected line sepsis with respect to some of the factors that were assessed are shown in Table 1 and Table 2.

There was no statistically significant association between confirmed line sepsis and all factors that were assessed including age, sex, number of lumens, hospital wards and average blood glucose. Similarly, all of these factors, except average blood glucose, were not statistically associated with suspected line sepsis. The results revealed that average blood glucose was significantly associated with suspected line sepsis (p=.004).

Conclusions: In this small population sample, a total of about 50% of patients had, at some point in time during their PN, a least one line sepsis episode that was either confirmed (10%) or suspected (40%). The line sepsis prevalence was 5.11 per 1000 catheter-days. The only factor associated with suspected line sepsis was average blood glucose. This is a known factor associated with line sepsis that needs to be monitored more closely in our hospital environment. We continue to collect data prospectively and plan to repeat this analysis on a larger number of patients.

Financial support: Baxter Canada.


Table 1. The infection rates of confirmed line sepsis.

Variable Infection Rate

(per 1000 catheter-days) P Value

Sex Male 6.45 0.318 Female 3.70

Number of Lumens Single 8.18 0.233 Multiple 4.20

Hospital Wards Medical or ICU 4.67 0.773 Surgical 5.46

Average Blood Glucosea Normal (< 6.0 mmol/L) 6.66 0.217 Elevated (> 6.0 mmol/L) 3.21

Age < 65 years 4.85 0.449 ≥ 65 years 5.61

aAverage blood glucose was calculated by taking the mean of patients’ blood glucose levels that were recorded once a week during their hospitalization, according to the study protocol. Table 2. The infection rates of suspected line sepsis.

Variable Infection Rate

(per 1000 catheter-days) P Value

Sex Male 41.54 0.150 Female 31.04

Number of Lumens Single 27.82 0.211 Multiple 38.82

Hospital Wards Medical or ICU 41.28 0.209 Surgical 32.08

Average Blood Glucosea Normal (< 6.0 mmol/L) 26.63 0.004 Elevated (> 6.0 mmol/L) 48.12

Age < 65 years 34.47 0.794 ≥ 65 years 40.36

aAverage blood glucose was calculated by taking the mean of patients’ blood glucose levels that were recorded once a week during their hospitalization, according to the study protocol.

S17—Analysis of Current Parenteral and Enteral Monitoring System Used by a Multidisciplinary Home Infusion Nutrition Support Team Results in Positive Outcomes and Further Process Improvement Maura O'Neill, MBA, RD, CNSC1; Noreen Luszcz, MBA, RD, CNSC1 1Nutrition Support Program, Option Care, Deerfield, IL, USA Purpose: It has been well documented that the multidisciplinary team approach to care including registered dietitians is the most beneficial practice model for management of nutrition support patients. This model has proven to decrease cost, increase quality of life, and improve outcomes. It is critical to analyze clinical practices in the home infusion arena since it is often customary for the caring physician to rely on the home nutrition support team (HNST) for clinical management decisions. This study analyzed current process and tools used for home parenteral and enteral nutrition (HPEN) monitoring by a national home infusion company that employs the multidisciplinary team model.

Methods: A retrospective study examining data collected by home infusion dietitians seeks to measure patient outcomes and identify areas for improvement. Dietitians were asked to submit information gathered from two monitoring tools; one designed for home parenteral nutrition (HPN) and the other for home enteral nutrition (HEN). Patients 21 and older were randomly selected from those on service from February 1 – July 31, 2015. Data collected included diagnosis, gender, age, primary nutrition goal, whether goal was achieved to date, hospitalizations and reason for hospitalizations, length of therapy and clinical interventions performed.


Results: 63 dietitians submitted data on 124 HPN and 126 HEN patients over a 6 month period. HPN patients’ average age was 54; 44% on service for greater than 180 days, 90% were meeting or progressing toward their goal (most common goal was weight repletion 44%), 11 patients were not meeting or progressing to goal: 4 off service before goal met, 1due to illness, 2 noncompliant, 4 other. Half were been hospitalized at least once: 42% disease related, 35% unrelated to therapy, 19% nutrition therapy related, 3% unknown. HEN patients’ average age was 63; 86% on service for greater than 180 days, 92% meeting or progressing to goal (most common goal was tolerating tube feeding 46%), 10 patients were not meeting or progressing to goal: 1 illness, 1 unrealistic goal, 1 intolerance, 3 noncompliance, 4 other. 33% of patients hospitalized at least once: 32% disease related, 39% unrelated to therapy, 9% nutrition therapy related, 20% unknown. Four clinical interventions were tracked for each therapy. 390 HPEN clinical interventions were reported resulting in an estimated 957 hospital days avoided equating to $1,914,000 in healthcare dollars saved.

Conclusions: Current monitoring tools appear to be capturing critical information for on-going clinical management however when collecting outcomes, limitations were noted It is difficult to capture detailed information in a tool, with much of the details ending up in progress notes. For example when reporters have to choose a response around patient status, individual patient complexities cause “other” or “unknown” to be chosen. This study did point out a few essential modifications that need to occur for better translation of data. The use of reliable forms and tools is essential for evaluation of HNST performance, patient outcomes and process improvement. Most significantly noted, systematic monitoring with responsive clinical interventions correlates with reduced hospitalizations and healthcare dollar savings.

Financial support: Option Care.

S18—Where Are We Today With the 2014 PN Safety Consensus Recommendations: A Snapshot of PN Order Forms Across the US Angelina Mason, RD, LDN, CNSC1; Penny L. Allen, RD, CNSC1 1Parenteral Nutrition, AxelaCare, Lenexa, KS, USA Purpose: Focus on safe ordering practices for parenteral nutrition (PN), a high alert medication, is not a new topic. The A.S.P.E.N. National Advisory Group (NAG) published Safe Practices for PN Formulations in 1997 as a result of common errors related to PN ordering, compounding and administration, with specific guidelines on the standardization of PN labeling. Guidelines have been updated several times since 1997. Boullata et al. published the results of a survey in 2013 demonstrating that little progress had been made in adopting NAG guidelines for safe PN ordering. In 2014, A.S.P.E.N. published updated safety recommendations and guidelines that specifically address safe prescribing practices and a standardized PN process for facilities and practitioners. The impact of drug shortages over the last 5 years has altered prescribing practices and the actual PN order forms being used by institutions who have not yet adopted the A.S.P.E.N. Parenteral Nutrition Safety Consensus Recommendations.

Methods: A total of 60 adult hospital PN order forms were collected for a three month time period (June 2015 through August 2015) from a variety of medical institutions across the United States that refer home PN patients to a national infusion provider. Pediatric order forms were excluded, along with adult forms that were incomplete. Components of the order form were tracked, analyzed and assessed for compliance against the new 2014 recommendations.

Results: Most notable observations were: -90% of institutions utilized standard PN order forms but 50% were handwritten, not electronic, with twelve percent (12%) not clear and/or easily understood.-60% did not order macronutrients in grams per day as recommended.-58% of institutions order PN as a 2:1 solution, with 6% of lipid orders having an infusion time >12 hours.-32% of institutions do not order electrolytes in amount per day, with 37% ordering individual ions vs. complete salt as recommended.-30% of institutions do not list all parenteral ingredients on their order forms. Frequently omitted were calcium, phosphorus, MVI and trace elements, assumedly as a response to shortage status.-86.7% order multivitamins (MVI) daily with 60% ordering a full dose of 10ml of MVI daily. -11.7% order MVI per “dose” (volume unspecified). -10% ordered 5ml of MVI daily (50% of the recommended dose).–8.7% did not list MVI at all as an option to order.

Conclusions: Ongoing efforts to educate, adopt and implement standardized PN processes and current A.S.P.E.N. Safe Practices is clearly needed in many institutions where guidelines are not recognized. Clinicians ordering PN should be well versed in the field of nutrition support and receive ongoing training especially as parenteral product shortages resolve (specifically regarding appropriate dosing of micronutrients such as calcium, phosphorus, MVI,


and trace elements in times of non-shortage). The risks and safety issues associated with PN will continue to be an issue without the guidance, education and push for changes to institutional policies including out of date, incomplete PN order forms. Home infusion companies with nutrition support teams may be at a unique advantage to identify gaps and inaccuracies with the PN order as the patient is transferred to the alternate site and assist with education to the referring institution and community.



S19—Thirty-Day Readmission Rates Are High in Hospitalized Patients Discharged on Home Parenteral Support

Denise M. Konrad, RD, CNSC1; Hiren Vallabh, MD1; Robert DeChicco, MS, RD, CNSC1; Rocio Lopez, MS, MPH1; Ezra Steiger, MD, FACS1; Donald F. Kirby, MD, FACP, FACN, FACG, AGAF, CNSC, CPNS1

1Cleveland Clinic Foundation, Cleveland, OH, USA

Purpose: Clinicians, payers, and policy makers have increased focus on reducing hospital readmissions to reduce healthcare costs and improve quality of care. There are no published studies examining the 30-day readmission rate in patients discharged on home parenteral support (HPS; parenteral nutrition or intravenous fluids) and which risk factors are associated with increased readmission risk. The objective was to determine 30-day readmission rates for patients discharged on HPS and if various factors had an associated effect on readmission risk. We hypothesized that patients with a higher degree of malnutrition would be more likely to be readmitted within 30 days of being discharged on HPS.

Methods: A retrospective case series was completed using an IRB-approved database to identify patients followed by the home nutrition support service (HNSS) who were discharged on HPS for the first time from a large tertiary care hospital between July 1, 2013, and June 30, 2014. Data collected included HPS indication, presence of an enterostomy, HPS formula volume, comorbidities, inflammatory markers, and presence and degree of malnutrition at discharge among other variables. A comprehensive nutrition assessment was performed by trained clinicians on all patients using a combination of the Academy/A.S.P.E.N. guidelines and our institution’s guidelines for diagnosing adult malnutrition.

Results: There were 239 patients included, with an average age of 50.4 years. Twenty-seven percent of patients had cancer; 25% had inflammatory bowel disease (IBD); and the remaining had other underlying diseases. Indications for HPS were short bowel syndrome (28.6%), motility disorder (26.1%), fistula (21%), obstruction (20.6%), mucosal disorder (2.9%), and pancreatitis (0.8%). Thirty-six percent (n = 86) of patients were readmitted between 1 and 3 times within 30 days of initial hospital discharge. Of the 86 initial readmissions, 14% (n = 12) were HPS-related, and 86% (n = 74) were not. Of the HPS-related readmissions, 6 were related to metabolic reasons; 5 were catheter-related bloodstream infections (CRBSIs); and 1 was a mechanical complication. All patients readmitted with CRBSIs were prescribed ethanol lock. The majority of patients (84.5%) were diagnosed with some degree of malnutrition, including 13.4% with mild, 24.2% with moderate, and 46.9% with severe, but the presence or degree of malnutrition was not associated with readmission rates (P = .27; Figure 1). On univariable analysis, patients with IBD had a 45% lower hazard of 30-day readmission than those without IBD (HR = 0.55, P = .44). Compared with subjects that underwent small bowel resections, those with the entire small bowel had twice the risk of 30-day readmission (HR = 2.2, P = .006). Patients with an enterostomy had 39% lower risk of readmission than those with bowel in continuity (HR = 0.61, P = .046). Last, for every 100-mL increase in HPS volume, the hazard of readmission decreased by 5% (HR = 0.95, P < .001).

Conclusions: Patients discharged with HPS for the first time who were cared for by an experienced HNSS have a high 30-day readmission rate. Surprisingly, the presence and degree of malnutrition were not associated with readmission rates. Also unexpectedly, challenging patients to manage (eg, those with IBD, small bowel resections, enterostomies, and patients with higher HPS volumes at discharge) had lower rates of 30-day readmissions.



Figure 1. Cumulative 30-day readmission rates by malnutrition diagnosis: Kaplan-Meier plot.

S20—Severe Dermatitis Associated With Zinc Deficiency in a Parenteral Nutrition–Dependent Patient Due to National Shortage of Trace Minerals and a Monitoring Tool to Improve Care Patricia Brown, RD,CNSC1; Peter Cornell, MSN, CRNP1; Frederic E. Eckhauser, MD1; Renay Tyler, RN, ACNP, CNSN, DNP1 1Surgery, The Johns Hopkins Hospital, Baltimore, MD, USA Purpose: To report a case of severe, misdiagnosed acrodermatitis resulting from zinc deficiency in a patient receiving central parenteral nutrition (CPN), and suggest a monitoring tool to increase patient safety, improve communication and decrease lab draws.

Case Summary: An 83 year old woman presented with malnutrition, multiple enteric fistulas and an extensive, non-painful excoriating skin rash. The rash was initially diagnosed and treated as Herpes Simplex without improvement. The patient remained on complete bowel rest and CPN for four months with no closure of the fistulas. During this time she was transferred to multiple facilities. After transfer to our facility, a Dermatology consultation was obtained and established a correct diagnosis of zinc deficiency. A profoundly low plasma zinc level of 11 ug/dl (normal 60-130 ug/dl) confirmed the diagnosis. The CPN was supplemented with 20 mg zinc daily with rapid improvement in the rash and complete resolution after several weeks. Shortly thereafter, the patient underwent successful closure of her enteric fistulas.

Discussion: Several factors contributed to this patient’s malnutrition, zinc deficiency, and missed diagnosis; including poor communication between the treating facilities and lack of awareness of the clinical signs and symptoms of zinc deficiency. One facility had not provided trace elements to the patients CPN due to national shortages and were unaware of the impact of zinc deficiency on wound healing and immune competency. Frequent transfers between facilities further complicated the matter, and made it difficult to establish and communicate clear-cut nutritional goals, including the need to monitor and replete trace elements on a regular basis.

Conclusion: The use of a patient specific online tool or pocket card to monitor and record nutritional status and goals and recent laboratory values at regular prescribed intervals may be a valuable adjunct to improve nutritional support and avoid errors and confusion when a patient is being monitored in a hospital or home health care setting or when he or she is transferred between facilities (see Figures 1 and 2). We have designed a tool with a suggested list


of data items critical for handoffs that we plan to pilot. Insuring awareness of nutritional goals and possible trace element deficiencies in any patient on PN is critical to provide safe and cost-effective nutritional care.


Figure 1. Pocket card (front).

Figure 2. Pocket card (back).

Patient Name Duodenum Jejunum ILeum CPN Additives IVF Weight FE Infusion

Goals of Therapy: Intact: Intact: Intact: Base: Levocarnitine: Regimen: Current Wt: Dates:

CM: CM: CM: AA: Ranitidine: Goal Wt:

Central Access: Dexi: Famotidine:

Po Diet: Lipid: MVI:

Total Volume: MLs:

ORS/day:

Cycle: MTE type:

Antimotility Meds: MLs:

Zinc deficiency: Dermatitis,alopecia, anorexia, taste sensitivity, poor night vision,immune compromise, and impaired wound healing.

Copper deficiency: Hypochromic, microcytic anemia and neutropenia are common findings. Hypercholesterolemia may be observed.

Selenium deficiency:cardiomyopathy,myalgias, myositis, hemolysis, and impaired cellular immunity.

Chromium deficiency: Glucose intolerance, hyperlipidemia, peripheral neuropathy,and encephalopathy

Manganese deficiency: Weight loss, transient dermatitis, and, occasionally, nausea and vomiting. (1)

ZN CU SE CR MN VIT A Vit D B12 Vit E Prealb 24HrUUN Alk Ph AST ALT Tbili CRP PT INR Fe Ferrit

Date

Date

Date

Date

Date

Date

Date

(1) A.S.P.E.N. Clinical Practice Committee Shortage Subcommittee. A.S.P.E.N. Parenteral Nutrition Trace Element Product Shortage Considerations. Nutr.

Clin. Pract. 2014;29:249‐251.


S21—Medicare Parenteral Nutrition Policy 20 Years Later: Need for Change When 9 Out of 10 Do Not Qualify for Benefits Penny L. Allen, RD, CNSC1 1Parenteral Nutrition, AxelaCare, Lenexa, KS, USA Purpose: For greater than a decade, the National Home Infusion Association (NHIA) along with organizations like A.S.P.E.N. have lobbied the Centers for Medicare and Medicaid Services (CMS) to change the existing policies for home infusion therapy. Current CMS coverage for home infusion lags behind all other payers so it is with hope that the latest bill, the Medicare Home Infusion Site of Care Act S. 275/H.R. 605 will pass, allowing patients to receive infusion treatments such as anti-infectives and parenteral nutrition (PN) in the home setting. Current policy for home PN severely limits access to services with few patients meeting criteria due to the test of permanence (documented need for PN for 90 days or longer), or non-qualifying conditions for PN such as complications due to cancer treatment or bariatric surgery. In a previous abstract presented at Clinical Nutrition Week in 2007 by this author, it was reported that only 16% of Medicare PN referrals received by a national infusion provider met CMS criteria. Almost 10 years later, even fewer Medicare beneficiaries requiring PN appear to be meeting this restrictive criteria leaving very few options other than to receive therapy in a more costly setting such as the hospital or skilled nursing facility (SNF), pay for therapy out of pocket, or go without treatment.

Methods: All Medicare PN referrals received by a national infusion provider were manually tracked for a 60 day period in 2015. Ninety five referrals for PN with Medicare as primary payer were sent to a centralized email review group comprised of a multidisciplinary team of clinicians and Medicare billing specialists who review each case upon referral for determination of coverage. A checklist based on current CMS Local Coverage Determination policy for PN is used to assess documentation received and to provide guidance regarding objective testing required if qualification is possible. A final determination of coverage is provided along with suggestions if the situation does not meet the criteria (a secondary insurance, SNF, self-pay, hospital pay, etc.).

Results: Ten years after an initial report describing limited home PN coverage for Medicare beneficiaries, a recent snapshot of geographically and clinically diverse PN referrals reveals only 10.5 percent of referrals now meeting the Medicare criteria for home PN. The Medicare fee-for-service program is the only payer in the United States currently not recognizing the clinical and cost benefits of providing infusion in the home setting. The PN policy is 20 years old and is not reflective of current practice patterns and appropriate utilization of PN today, often requiring testing that could be harmful such as a fecal fat test or an enteral trial in patients that may not tolerate. Conclusions: With 9 out of 10 Medicare beneficiaries not qualifying for home PN coverage when medically necessary, the Medicare Home Infusion Site of Care Act would ensure that beneficiaries could receive the infusion therapy they require in a setting that is desirable, convenient, and by far the most cost effective.


S22—Maintaining Patient Safety When Transitioning From Hospital to Home: Case Report of an Unsafe Referral for Home Parenteral Nutrition Shirley Au, RD, CNSC1; Penny Allen, RD, CNSC1 1Parenteral Nutrition Services, AxelaCare, Lenexa, KS, USA Purpose: Healthcare professionals are often faced with decisions regarding the safety and appropriate use of nutrition support. This is particularly true when determining clinically whether a patient is an appropriate candidate for home parenteral nutrition. The following is a case study of patient with longstanding anorexia nervosa who presented with refeeding syndrome after initiating parenteral nutrition in the hospital and was subsequently referred for home parenteral nutrition.

Methods: 62-year-old female weighing 57 lbs (25.9 kg) and 4’11” tall with a 30+ year history of anorexia nervosa presented to the ER of a community hospital with recurrent hypoglycemia and chronic hypotension. Patient also complained of sore mouth and throat, swollen tongue, and swelling of bilateral lower extremities, most likely a result of vitamin B deficiencies and hypoalbuminemia. After admission, she requested a PICC line and initiation of PN, since she had been on PN 3 years earlier and felt it successfully improved her nutritional status. The physician initiated the hospital “standard” PN solution (dextrose 20%, AA 5%) at 20 ml/hr, with goal to advance to 40 ml/hr continuous infusion at home with lipids 3 days per week. The standard PN solution included electrolytes and multivitamins, but no additional thiamine supplementation. After 3 days of PN, the patient was referred to a home PN provider to continue therapy. Clinical documents were reviewed by the home PN clinical team, which consisted


of nutrition support dietitians, nurses, and pharmacists. Per the records, electrolytes were normal at the start of PN, and although blood glucose began to normalize once PN was started, potassium and phosphorus significantly decreased as PN was increased. Potassium fell from 4.3 mmol/dL to 3.5 mmol/dL and phosphorus decreased significantly from 3.2 mg/dL to 1.9 mg/dL, indicating the onset of refeeding syndrome (Table 1). During the referral process, it was also revealed that the patient had a long history of non-compliance in the past: refusing diagnostic studies, turning off infusions while in the hospital, and leaving a hospital stay AMA earlier in the month.

Results: Upon review of the patient’s case and history and a number of discussions with the patient and healthcare providers, the home infusion PN team deemed that the patient was an inappropriate home PN candidate. The risk of refeeding syndrome in a severely malnourished patient combined with a significant history of non-compliance were major factors, along with the concern that the hospital team did not understand the severity of the situation. The home infusion team, in efforts to advocate for the patient’s overall safety, recommended the hospital reassess the discharge plan and consider transferring her to a specialized inpatient facility for eating disorders. The hospital discharged the patient home that evening without PN. She presented to the ER next day with a blood glucose below 20 mg/dL and was readmitted with a subsequent referral to the same home PN provider, with electrolytes and overall status remaining unstable. The patient was not accepted for home PN services due to concerns for patient safety. Conclusions: Patients with severe eating disorders that require nutrition support should be followed in specialized inpatient facilities where strict monitoring of clinical status and adherence to therapy is available. Home PN therapy requires adherence to defined standards of practice for admission criteria in order to provide for patient safety in the alternate site. Ongoing education is needed regarding the risk and benefits of nutrition support for patients with eating disorders.

Financial Support received from: AxelaCare for abstract submission fee

Table 1. Lab Values Through Hospital Stay.

Hospital Day Day 1 Day 2 Day 4

PN at 10 ml/hr PN at 20 ml/hr

Glucose (mg/dL) 32 47 62

Albumin (g/dL) 3.1 2.8 2.7

Potassium (mmol/L) 4.9 4.3 3.5

Magnesium (mg/dL) 1.9 2.0 1.8

Phosphorus (mg/dL) 3.3 3.2 1.9

S23—Secular Trends in Parenteral Nutrition Use: A 15-year Experience Jennifer Isautier, APD1; Troy J. Cross, PhD1; Erin M. Nystrom, PharmD2; Randy F. Wendt, BS3; Daniel L. Hurley, MD1; Molly McMahon, MD1; John M. Miles, MD1 1Endocrinology, Diabetes, Metabolism and Nutrition, Mayo Clinic, Rochester, MN, USA; 2Pharmacy, Mayo Clinic, Rochester, MN, USA; 3Information Technology, Mayo Clinic, Rochester, MN, USA Purpose: For over 20 years, clinical practice guidelines for administration of parenteral nutrition (PN) have emphasized selective use because of associated risks. However, relatively little information is available regarding the impact of these guidelines on clinical practice. Specifically, there are very few studies that report long-term trends in PN utilization, especially in relation to the use of enteral nutrition (EN) support.

Methods: We undertook a review of PN use in two large tertiary care hospitals (2059 beds) from 2000 through 2014 to determine its frequency and composition as well as survival in the PN patient population. PN macronutrient and energy content was analyzed in detail. Records were also reviewed to determine the number of patients receiving EN; these data were available only beginning in the year 2005. In the case of multiple hospital admissions, only the first admission was included in the analysis. The data were analyzed using one-way analyses of variance. One year mortality rate was assessed with Kaplan-Meier Survival curves.


Results: A total of 14,998 PN patients were included in the analysis. The number of patients receiving PN declined from 1140/y to 700/y over the 15 year observation period. An increase in the use of EN from 1566/y to 2086/y occurred between 2005 and 2014. In the PN population, there was no significant change in age (59 ± 1 y), body mass index (26.0 ± 0.2 kg/m2) or hospital length of stay (25 ± 1 days) over the 15-year study period. However, the duration of time on PN increased significantly (7 ± 1 to 11 ± 1 days; P<0.05). There was a significant decrease in daily energy supply from PN (Δ = -57 kcal/day) and carbohydrate content of PN (Δ = -20 g/day), both p<0.05. Protein content increased significantly (Δ = +11 g/day, p<0.05) but there was no change in lipid supply. The average blood glucose concentration during PN decreased (Δ = -28 mg/dL, p< 0.05). There was no significant change in triglyceride concentrations during PN (Δ = -14 mg/dl, P =NS). The one-year survival rate improved significantly from 65% to 76%, (log rank χ2 = 89.6, P<0.01).

Conclusions: These data indicate decreased PN use and a reciprocal increase in EN use in a large tertiary care center. Although the total number of patients receiving PN decreased, the average duration of PN increased. This observation may reflect an influence of the clinical guidelines, which emphasize that short-duration PN should be avoided. Of note, one-year survival greatly improved over the 15 years of observation. There was also an improvement in hyperglycemia, potentially due to a reduction in energy and carbohydrate supply. It cannot be determined from these results whether changes in PN composition and/or improved glycemic control were causal factors mediating the improvement in survival. Financial support: Mayo Foundation.

S24—Understanding the Clinical Characteristics, Indications, and Nutrient Provision in Obese Adult Patients Who Received Home Parenteral Nutrition: A Preliminary Analysis From the Sustain™ Registry Jisun Park, BS1; Shirley Paski, MD4; Charlene Compher, PhD, RD, CNSC, LDN3; Lingtak-Neander Chan, PharmD, BCNSP2,1 1Graduate Program in Nutritional Sciences, University of Washington, Seattle, WA, USA; 2Pharmacy, University of Washington, Seattle, WA, USA; 3Nursing, University of Pennsylvania, Philadelphia, PA, USA; 4Medicine, University of Washington, Seattle, WA, USA Purpose: Historically, the goal of HPN is to promote weight gain and achieve weight maintenance since patients who received long-term PN are usually underweight and malnourished. With obesity becoming a common comorbidity, the prevalence of obesity among HPN patients is unknown. The research of HPN use among obese patients is essentially nonexistent and there is no established consensus on how to provide HPN in obese patients. The objective of this study was to establish baseline knowledge regarding obese patients requiring HPN. The aims were: 1) to identify the prevalence and clinical characteristics of obese patients receiving HPN, and 2) to determine the indications and goals of HPN in obese patients.

Methods: This is a descriptive analysis of data from the HPN patients enrolled in the Sustain™ Registry between August 2011 and February 2014. Adult, obese patients, defined as having a body mass index (BMI) ≥30.0 kg/m2 were included. The following data were included for analysis: patient anthropometric and demographic information, medical history, indications and goals for HPN, nutrient provision information, and other pertinent patient laboratory results. Descriptive statistics was used for data analysis.

Results: Of the 1251 patients included in the database, 1064 were adults at the time of enrollment. A total of 154 (14.5%) of these adults were obese. The median age of the obese cohort was 52.8 years (range 25.0 to 86.2). The majority of these patients were Caucasian (70%), and female (75%). The median baseline weight was 95 kg (range 70 to 140) with a median BMI of 34.1 kg/m2 (range 30.0 to 57.1). The details of their weight distribution at baseline are summarized in Table 1. The most significant past medical/surgical history associated with the gastrointestinal tract (GI) is summarized in Table 2. Bariatric surgery and GI tract cancer were among the most commonly reported categories. The leading indications for HPN were GI fistula (26%), followed by others (24%), bowel obstruction (13%), short bowel syndrome (12%), bowel dysmotility (10%), and intractable vomiting (10%). The average values of the liver functions tests at baseline were within the normal range. Baseline serum album was 3.0 g/dL. The mean daily calorie prescribed was 20.4 kcal/kg/day. The mean protein and fat prescribed was 1.21 and 0.68 g/kg/day (using actual body weight), respectively. The majority of the PNs were administered via PICC line (66%), followed by a tunneled catheter (28%). Ninety-five percent of these patients received HPN daily, with 17% receiving oral diet ad libitum. With regard to the goal of HPN, weight loss was the most common treatment goal (66%), followed by weight maintenance (18%), and weight gain (2%). In 14% of the patients, the goal of HPN was not specified.


Conclusions: The prevalence of obesity among this cohort of HPN patients is 14.5%. The leading underlying GI-related complications were bariatric surgery and GI tract cancer. The most common treatment goal for HPN in obese patient was for weight loss. We are performing ongoing analysis to assess the safety, effectiveness, and impact of costs associated with HPN in obese patients.


S25—Accuracy of Using Intravenous Electrocardiography to Confirm Placement of Peripherally Inserted Central Catheter for Total Parenteral Nutrition Manpreet Mundi, MD1; Jithin Edakkanambeth-Varayil, MD2; Megan McMahon, PA-C2; Nishanth Vallumsetla, MBBS1; Akiko Okano, CNP2; Sara Bonnes, MD2; Ryan Hurt, MD2 1Division of Endocrinology, Mayo Clinic, Rochester, MN, USA; 2Division of General Internal Medicine, Mayo Clinic, Rochester, MN, USA Purpose: Total parenteral nutrition (TPN) is a life-saving therapy for patients with intestinal failure. Safe delivery of the hyperosmotic solution of PN requires a central venous catheter (CVC) with tip in the lower third of the superior vena cava, at the atrio-caval junction, or in the upper portion of the right atrium. Malposition of CVC can lead to thrombosis, pericardial tamponade, cardiac perforation, and even death making it paramount that tip location be verified prior to initiation of TPN. Typically, this confirmation occurs with use of anterior/posterior chest radiography, which does have the drawback of small radiation exposure as well as a delay given the time required to take the x-ray and have the tip location confirmed by radiology prior to use. Due to this, other techniques such as intravenous electrocardiogram (ECG) are being more readily utilized to confirm tip location at time of placement.


Methods: Usage of intravenous ECG was implemented at our institution on December 17, 2014. Records for all patients consulted for Home Parenteral Nutrition (HPN) from December 17, 2014 to June 16, 2015 were reviewed to assess for research authorization, availability of intravenous ECG confirmation and chest x-ray following PICC line placement. Data regarding patient demographics, diagnosis leading to HPN initiation, as well as ECG confirmation and chest x-ray reports were obtained. Chest X-rays were subsequently reviewed by Radiologist to reassess location of CVC tip and identify those that should be adjusted. Osmolarity of TPN formulas at the time of CXR was also obtained.

Results: A total of 73 patients were eligible to be included. Research authorization was not available in 5 patients. 51 patients were excluded because post placement chest x-ray was not available. 17 patients (30% male) with average age at time of catheter placement of 54 +/- 14 years were reviewed (Table 1). Enterocutaneous fistula was the most common indication for initiation of TPN (5 patients). Other indications included small bowel obstruction (3 patients), enteric/anastomotic leak (3 patients), short bowel/high ileostomy output (3 patients), mesenteric ischemia (1 patient), chylous leak (1 patient), and enterocutaneous fistula combined with small bowel obstruction (1 patient). For all patients, post-placement intravascular ECG reading stated tip in SVC. According to radiologic evidence, the location of catheter tip was satisfactory (low SVC or SVC/RA junction) in 10 patients (59%). Remaining patients had tip in unsatisfactory location; mid to high SVC in 4 patients (24%), and brachiocephalic vein in 3 patients (18%). The mean osmolality of TPN in patients with unsatisfactory location was 1932 +/- 352, with an average duration of 20.6 +/- 4.0 hours.

Conclusions: Due to the high osmolality of TPN, CVC tip location is of paramount importance. After radiology review of CXR, we noted that 7 out of 17 (41%) PICC lines were in unsatisfactory position despite Intravenous ECG confirmation. With current data available, Intravenous ECG confirmation should not be used as the sole source of tip confirmation in patients receiving TPN.



Table 1. Description of 17 Cases With HPN Formula Osmolality, Duration, and Comparison of ECG Confirmation Reading versus Chest X-Ray Reading.

S26—IV Lipids May Improve Glycemic Control in Patients Receiving Total Parenteral Nutrition Allison K. Parker, MS, RD, CD, CNSC1; Ruth Thomas, RD CD1 1Nutrition Services, Harborview Medical Center, Seattle, WA, USA Purpose: Patients on total parenteral nutrition (TPN) receive IV lipids that are vulnerable to shortages, primarily due to their small number of manufacturers, manufacturing complexity, and discontinuations. Practice at Harborview Medical Center (HMC) is to reduce provision of 250ml 20% Intralipid from daily to three times per week during times of shortage. However, little is known about how this lipid rationing affects TPN-fed patients. We hypothesized that lipid rationing due to shortages would cause patients to receive more kilocalories (kcal) from carbohydrate and protein in their TPN, and less kcal from lipid, and that this would correlate with impaired glycemic control. Methods: A review of TPN-fed patients at a level 1 trauma center was undertaken. All patients that received TPN while HMC was affected by the lipid shortage (September 2013–March 2014) were enrolled in the study. Initial enrollment included 61 patients. Patients were then excluded according to the following criteria, designed to


eliminate potential confounding effects on glycemic control: Received PO (n=21) or enteral nutrition (n=6) during TPN; were in ICUs during TPN (n=16); were aged <18 years (n=6); had very poor baseline glycemic control as defined by an HbA1C of >10% (n=2). This resulted in ten shortage-affected patients, which were matched with ten control patients that were unaffected by the shortage and thus received daily lipids. Final analysis included 20 (ten case, ten control) patients comprising 139 patient days. Data was analyzed using STATA software, with researchers blinded to shortage/non-shortage status during data collection and analysis. As shown in Table 1, groups were similar with regard to age, baseline HbA1C, BMI, and sex. Groups received nearly identical total kcals in their TPN, but their macronutrient distribution ratios were shifted due to the shortage, as was the subject of our study. Shortage-affected patients received significantly less kcal from lipid, and more from carbohydrate and protein.

Results: Consistent with our hypothesis, glycemic control was impaired in the lipid shortage-affected group that received less kcal from lipid and more kcal from carbohydrate and protein. Mean blood glucose was higher in the shortage-affected group (148.6 ± 43.5 mg/dl) compared to the unaffected group (131.8 ± 45.4 mg/dl), as seen in Figure 1 (p=0.0322). Insulin usage did not relate significantly to shortage status (p=0.3604). Interestingly, within the shortage-affected patients that received lipids three times per week, mean blood glucose was higher on the days that they did not receive lipids (152.9 ± 40.5 mg/dl) than on days that they did (135.1 ± 46.0 mg/dl), even though the carbohydrate and protein portions of their TPN are identical each day (p=0.0497). Insulin usage was unaffected by whether patients received lipids or not each day (p=0.3999). Potentially contributing to this impaired glycemic control among short-age affected patients, Figures 2 and 3 show that kcal from carbohydrate (dextrose) correlated positively with mean blood glucose (p=0.0031, R2 = .0652) and total insulin usage (p=0.0002, R2 = .0943). Kcal from protein (amino acids) did not significantly relate to blood glucose (p=0.0979) nor insulin usage (p=0.4827).

Conclusions: Patients affected by the national shortage of IV lipids received less kcal from lipid in their TPN, and more kcal from carbohydrate and protein. This correlated with impaired glycemic control as evidenced by significantly higher mean blood glucose and insulin usage. These findings suggest that IV lipids may be beneficial in improving glycemic control in TPN-fed patients.


Table 1. Subject and TPN Characteristics.

Shortage (n=10) No shortage (n=10)

Mean SD Mean SD p value

Age (years) 59.9 11.1 57.8 13.8 0.7126

HbA1C (%) 6.5 1.3 6.2 1.2 0.6129

BMI 27.0 7.9 28.0 7.6 0.7744

TPN components Total kcal 1902 609 1875 466 0.7734

Kcal from lipid 230 251 500 0 0.0000

Kcal from dextrose 1154 448 929 355 0.0012

Kcal from amino acids 518 151 446 139 0.0041

Sex Male n=6 n=6

Female n=4 n=4


Figure 1. Blood glucose and IV lipids.

Figure 2. Blood glucose and Kcal from dextrose in TPN.


Figure 3. Insulin requirement and Kcal from dextrose in TPN.

S27—Can the Oral Caloric Contribution to Overall Energy Balance in Long-Term Home Parenteral Nutrition Consumers Be Estimated? Carol Ireton-Jones, PhD, RDN, CNSC, FASPEN1; Marianne Opilla, RN1; Reid Nishikawa, Pharm D, BCNSP1; Rod Okamoto, RPh1 1Nutrishare, Carrollton, TX, USA Purpose: Energy expenditure (EE) of long-term home parenteral nutrition (HPN) consumers may be estimated using standard equations. Many long-term HPN consumers also have oral intake. Although oral intake is encouraged and may reduce liver dysfunction, actual oral calorie intake is rarely reported in this population because of the challenges in obtaining an accurate diet history as well as knowing how much is absorbed. The purpose of this analysis was to estimate the oral calories that contribute to the overall energy balance in long-term HPN consumers.

Methods: Data was collected on 8 long-term (>1 year) HPN consumers from one home infusion provider. Demographics including length of time on HPN, age, sex, weight, and BMI were recorded. Laboratory data collected included: Albumin (alb), BUN, creatinine (Cr), and total bilirubin (T bili). HPN data collected included amino acid, lipid, dextrose and total caloric intake as well as HPN frequency in days per week. The Mifflin-St Jeor (MSJ) equation increased by a factor of 1.3 or 1.7 was used to estimate EE. HPN caloric intake was compared to the EE to estimate the oral calories needed to meet the EE for energy balance. Actual caloric intake was obtained through a 3-day diet history reported by the consumer to estimate absorption. No one was receiving enteral nutrition.


Results: Six females and 2 males with intestinal failure requiring HPN for 23 + 15.2 (mean +SD) years to maintain nutritional status with an average age of 51+ 20 years (mean+SD) and weight of 58.6+ 13 kg (BMI = 22.1, IBW 56.2 kg) were evaluated. All consumers were weight stable (+/- 5%) for the past 12 months and had labs within range for long-term HPN (mean values: alb - 3.98 g/dL; BUN – 23 mg/dL, Cr – 0.98 mg/dL, and T bili – 0.74 mg/dL.). On average, these consumers received HPN 6.6 d/wk (range 4 to 7 d/week). The HPN nutrition provision is reported here as daily intake averaged over a 7 day period. Mean HPN intake was: amino acids = 64.7+16.6g/d (1.1 g/kg/d), IVFE = 171+171 kcal/d (0.3 g/kg/d), dextrose =203 + 103 g/d (3.5 g/kg/d) with total of 1118+ 372 kcal/day (19.2 kcal/kg/d). Mean EE was 1671+ 446 kcal/d. The oral caloric intake to meet EE (EE – HPN) was estimated to be 558+ 354 kcal/day while actual oral intake from diet history was 1745+ 1285 kcal/d (range 25 – 3753 kcal/day). Absorption of oral calories was an average of 47 + 35%. The calculation of the difference between HPN intake to EE did not provide an estimate of oral calories required for energy balance. In these patients, actual oral intake and absorption varied tremendously between individuals (see Table 1).

Conclusions: Oral calories contribute to overall energy balance; however, oral caloric need cannot be accurately estimated in long-term HPN consumers because the variable of absorption is unknown. There are many challenges in obtaining an accurate diet history in long-term HPN consumers as intake may vary from day to day based on activity, disease state and overall health status. All consumers desired to consume oral nutrients even if it was a very small amount daily, as has been previously reported. Therefore, serial, stable weights (assessed bi-monthly or more often) may be the most useful indicator of energy balance. More research is needed in this population to determine if the type of oral diet also plays a part in the absorption and oral caloric contribution to EE. Financial support: None.

Table 1. Caloric Comparisons in HPN Consumers.

Consumer ID 1 2 3 4 5 6 7 8

Estimated oral calorie needs, kcal/d 798 421 1132 268 833 20 412 577

Actual oral calorie intake, kcal/d 2047 878 1420 25 3753 1723 3364 750

Estimated absorption of oral caloric intake (%) 39 48 79 100 22 2 12 77

S28—Incidence of Complicated Liver Disease During the Administration of Long-Term Home Parenteral Nutrition Patients Davina Ngo, PharmD, BCNSP1; Craig Petersen, RD, CNSD1; Nabil Musallam, PharmD, MBA1;Rodney Okamoto, RPh1 1Nutrishare, Inc., Elk Grove, CA, USA Purpose: Liver dysfunction is a potential complication of long-term home parenteral nutrition (HPN) in adult and pediatric patients. The reported prevalence of liver dysfunction ranges from 10% to 70% in long-term HPN patients. Complicated liver disease (CLD) such as cholestasis may be life threatening and can lead to liver transplantation or death. Previous investigators have defined CLD as the presence of portal hypertension, portal fibrosis or cirrhosis on liver biopsy, total serum bilirubin concentration >3.5 mg/dL for at least 1 month, ascites, hepatic encephalopathy, variceal hemorrhage, or a factor V concentration <50%. Multiple factors have been associated with the development of CLD, including dose of intravenous fat emulsion (IVFE) >1gm/kg/day, soy based IVFE, lack of enteral/oral feedings, < 50cm small bowel length, duration of HPN use, overfeeding and other factors. Recent practices have changed in response to a better understanding of these causative factors. We wanted to determine the incidence of CLD in patients receiving long-term HPN from a single home infusion provider by analyzing more recent data than previous reports.

Methods: Clinical records were reviewed for all patients receiving long-term HPN from one home care provider from 12/1/2012 to 11/30/2014. All patients who were on long-term HPN for more than one year and expected to remain on HPN for greater than one year were included. Patients who had other documented risk factors for liver disease were excluded. We collected patient demographics and all total bilirubin results, documented liver biopsy reports, and documented signs/symptoms of liver failure possibly related to HPN therapy. Patients were considered to have CLD, if they had any of the following: portal hypertension, portal fibrosis or cirrhosis on biopsy, total serum


bilirubin concentration >3.5 mg/dL for at least 1 month, ascites, hepatic encephalopathy, variceal hemorrhage, or a factor V concentration <50%.

Results: There were 128 patients who qualified for the study. The mean age was 47 years with a range of 7 to 89 years. Seventy-six patients (59%) were female. Of these 128 patients, 89 patients (70%) had short bowel syndrome, 39 patients (30%) had <50 cm of small bowel, 58 patients (45%) had no ileo-cecal valve, 39 patients (30%) had no colon, and 2 patients (1.6%) who didn't eat or receive enteral feedings. There were 8 patients (6%) who met the criteria for CLD. Of these 8 patients, 2 patients had consistent total bilirubin levels of >3.5 mg/dL, which eventually lead to death. The other 6 patients had CLD documented by liver biopsy reports.

Conclusions: In this cohort of 128 long-term HPN patients, only 6% experienced CLD. Previous reports have ranged from 10 to 70%. Due to a better understanding of causes leading to CLD, including cholestasis, prescribing and other treatment practices may have changed to contribute to this change in incidence. Total bilirubin is normally used to monitor for cholestasis in this population and therapy may be adjusted using this marker. Surprisingly, only 25% of the affected patients had documented bilirubin greater than 3.5gm/dl. The other patients (75%) were diagnosed with CLD with liver biopsies. More work is needed to determine if the incidence of CLD in long-term HPN patients has changed with specific changes in prescribing patterns. Also, more investigation is required to evaluate whether a liver biopsy is warranted in all long-term HPN patients, to identify CLD and whether this will better guide their therapy. Financial support: None.


S29— Rifampicin Treatment Reduces Parenteral Nutrition–Associated Liver Disease in Neonatal Piglets Gregory J. Guthrie, PhD1; Barbara Stoll, PhD1; Douglas Burrin, PhD1 1Pediatrics, USDA ARS Children’s Nutrition Research Center, Baylor College of Medicine, Houston, TX, USA

Purpose: Preterm infants with complications due to gastrointestinal immaturity can receive parenteral nutrition (PN) for periods greater than 2 weeks. These infants are at risk for a progressive increase in cholestatic liver disease marked by elevated serum bilirubin and bile acid levels and, in severe cases, steatosis and liver damage (PNALD). The elevated levels of bile acids in the liver due to cholestasis can precipitate this liver injury. Modalities to minimize bile acids in the liver may be beneficial for preventing liver damage. Treatment with the antibiotic rifampicin has been used in adults to treat cholestatic pruritus. Treatment with rifampicin has been shown to increase clearance of bile acids through activation of the nuclear hormone receptor pregnane x receptor (PXR). PXR target genes are enzymes involved in phase I and phase II drug detoxification, which lead to increased 6-alpha hydroxylation of bile acids and their subsequent glucuronidation, increasing overall hydrophilicity and clearance. Our aim was to test whether rifampicin treatment can suppress the development of PNALD in neonatal PN-fed piglets

Methods: Newborn pigs (4 days old) were implanted with jugular catheters and received PN for 14 days. Pigs received daily PN consisting of 240 mL/kg of fluid, 240 kcal/kg, 25 g/kg of carbohydrate, 14 g/kg of protein, and 10 g/kg of lipid; the parental lipid used was Intralipid. Pigs (n = 8/group) were infused intravenously with either saline (IL) or 600 mg/kg of Rifadin IV (RIF), each day for 14 days.

Results: Treatment with RIF led to a 3-fold increase in the PXR target phase I enzyme Cyp3a29 gene expression compared with IL. RIF increased protein but not mRNA expression of the phase II glucuronidation gene Ugt1a6. Pigs in the IL vs RIF group had higher levels of serum direct bilirubin (2.62 ± 0.33 vs 1.33 ± 0.23 mg/dL) and total bile acids (34.05 ± 6.69 vs 17.06 ± 2.50 µM). However, the liver and gall bladder bile acid pool did not differ between groups. RIF treatment did not adversely affect the liver, as liver weight was significantly lower in the RIF group compared with the IL group (65.99 ± 3.70 vs 55.02 ± 2.21 g/kg). A marker of liver injury, serum ALT levels were not different between IL (15.75 ± 1.92 U/L) and RIF (14.75 ± 1.53 U/L) groups and within the normal range.

Conclusions: Treatment with rifampicin was effective in suppressing serum direct bilirubin and bile acid levels. The clearance of bile acids did not appear to occur via increased biliary secretion into the gall bladder but likely through increased clearance in the urine. Rifampicin did not appear to be hepatotoxic to neonatal pigs at the dosage level used. We conclude that short-term rifampicin treatment can reduce cholestasis associated with PNALD.

Financial support: USDA Agricultural Research Service under Cooperative Agreement Number 58-6250-6-001.


S30—Serum Linoleic Acid as a Marker of Intestinal Fat Absorption: A Case Report Duy T. Dao, MD1; Paul Mitchell, MSc2; Kathleen M. Gura, PharmD3; Alexis K. Potemkin, RN1; Meredith A. Baker, MD1; Gillian L. Fell, MD, PhD1; Lorenzo Anez-Bustillos, MD1; Bennet S. Cho, BA1; Alison O'Loughlin, MEd1; Mark Puder, MD, PhD1 1Vascular Biology Program and Department of Surgery, Boston Children's Hospital, Boston, MA, USA; 2Clinical Research Center, Biostatistics Core, Boston Children's Hospital, Boston, MA, USA; 3Department of Pharmacy, Boston Children's Hospital, Boston, MA, USA

Purpose: Introduction: Short bowel syndrome (SBS) results from extensive loss of small intestine either due to surgical resection or underlying gastrointestinal diseases. Patients with SBS require parenteral nutrition (PN) to maintain adequate fluid, electrolyte, and calorie intake. The diagnosis of parenteral nutrition associated liver disease (PNALD) is usually made in PN-dependent children with a direct bilirubin >2 mg/dL. Soybean oil, which contains a high omega-6:omega-3 fatty acid ratio, is the standard intravenous lipid emulsion and is thought to be a major contributor to PNALD. We have previously shown how substitution of omega-3-rich fish oil (FO) emulsion can reverse cholestasis, accompanied by a drop in the serum levels of linoleic acid (LA), an omega-6 fatty acid. Essential fatty acid deficiency (EFAD) is avoided due to the presence of LA’s downstream metabolites in the emulsion. In patients with SBS maintained on FO therapy, which contains only a trace amount of LA, we hypothesize that the rise in serum LA levels can serve as a surrogate marker of improved intestinal fat absorption.

Description: We report the case of an ex-25 week preterm baby girl who was transferred to Boston Children’s Hospital (BCH) at 8 months of age for the treatment of failure to thrive and PNALD. She had undergone extensive bowel resection for necrotizing enterocolitis at the referring hospital, leaving a total of 20-30 cm of small intestine. FO therapy was started shortly after arrival. Her cholestasis reversed on hospital day (HD) 119 and bowel continuity was subsequently re-established on HD 144. She initially relied on PN for all of her caloric intake and began enteral feedings five months after her admission. Her diet was advanced as tolerance of enteral feedings improved. At the time of submission of this abstract, the patient was tolerating tube feeds and continued to receive PN for supplemental caloric intake.

Results: During the first month, the serum level of linoleic acid dropped rapidly from 3863 to 275 umol/mL with the substitution of Intralipid for FO emulsion. This level remained below 500 umol/mL (mean 334 umol/mL) as she relied exclusively on PN during the first few months at BCH, and increased between HD 174 and 293 to a high of 661 umol/mL (mean 485 umol/mL) when enteral feeding was introduced (Figure 1). Tube feeds were decreased on HD 293 due to mild colitis and held on HD 349 for persistent leakage around the gastrostomy tube that required surgical repositioning. During this time, LA decreased to 292 umol/mL, the lowest since enteral feeding was introduced. After these events, she resumed tube feeds and once again LA levels increased to a high of 950 umol/mL (mean 826 umol/mL). Additionally, there was an inverse relationship between serum LA levels and percentage of calories obtained from PN. Throughout the duration of her treatment, the triene:tetraene ratio remained below 0.2, indicating the absence of biochemical EFAD.

Conclusions: In SBS children who are maintained on FO therapy for the treatment of PNALD, serum linoleic acid levels tracked over time may serve as a surrogate marker for enteral fat absorption. This may be a useful tool for evaluating intestinal adaptation while transitioning these patients to enteral feedings. Financial support: None.


Figure 1. Linoleic acid levels versus enteral and parenteral nutritional intake.


S31—Risk of Postprocedural Bleeding in Patients on Intravenous Fish Oil: Is it Real?

Lorenzo Anez-Bustillos, MD1; Prathima Nandivada, MD2; Alison O’Loughlin, MEd1; Paul Mitchell, MSc3; Meredith A. Baker, MD1; Duy T. Dao, MD1; Gillian Fell, MD, PhD1; Bennet S. Cho, BA1; Alexis K. Potemkin, RN, BSN1; Kathleen M. Gura, PharmD4; Mark Puder, MD, PhD1

1Vascular Biology Program and Department of Surgery, Boston Children’s Hospital, Boston, MA, USA; 2Department of Surgery, Beth Israel Deaconess Medical Center, Boston, MA, USA; 3Clinical Research Center, Biostatistics Core, Boston Children’s Hospital, Boston, MA, USA; 4Department of Pharmacy, Boston Children’s Hospital, Boston, MA, USA

Purpose: Omega-3 fatty acids are known to affect hemostatic pathways. Despite this knowledge, an abundance of clinical evidence fails to translate these effects into an increased risk of bleeding. Surgeons, however, continue to be concerned about bleeding, and many recommend discontinuing the use of omega-3 fatty acids before invasive procedures. In the pediatric population, intravenous fish oil lipid emulsion (FOLE) is used for management of intestinal failure–associated liver disease (IFALD) in parenteral nutrition–dependent patients. The publication of a recent isolated case report has raised concern for an increased risk of postprocedural bleeding while on FOLE. The objective of this study is to describe the incidence of clinically significant postprocedural bleeding (CSPPB) based on our experience using fish oil monotherapy among a large cohort of patients with IFALD.

Methods: We performed a retrospective chart review of prospectively enrolled patients at Boston Children’s Hospital treated with the FOLE Omegaven® for IFALD between September 2004 and July 2015 and subjected to any type of invasive procedure, defined as any intervention that involves violation of a vascularized and epithelialized surface. Clinically significant bleeding was defined as the presence of at least 1 of the following up to 1 month postprocedure: (1) reexploration or return to operating room for bleeding, (2) need for antibleeding therapy, (3) readmission for bleeding, or (4) death due to bleeding. Preprocedure data—including complete blood count,


coagulation profile, liver function tests, and fatty acid profile—were collected, as was duration of therapy at the time of procedure.

Results: Charts from 219 patients treated with FOLE were reviewed, for whom 180 (82%) underwent at least 1 invasive procedure for a total of 728. The patient sample was 58% male and had a median age of 10.1 months (interquartile range, 5.2–24.0) at the time of procedure. Forty-three percent of these were related to vascular access; 28% were abdominal; and 14% pertained to invasive endoscopic procedures. Only 4 of 728 (0.55%, 95% CI: 0.15%–1.40%) procedures resulted in CSPPB (Table 1). Three were of gastrointestinal origin and required return to the operating room. The first one followed a small bowel resection and returned for esophagogastroduodenoscopy, which failed to reveal the source of bleeding; it was later found to be from a vessel at the ileostomy site, which was controlled at bedside. The second one occurred after an ostomy takedown and required laparotomy for push enteroscopy, which did not identify a true source of bleeding but allowed for 2 questionable areas at the anastomosis to be oversewn. The third one occurred in a coagulopathic patient with portal hypertension, which presented with persistent bleeding from the stomal site and failed initial cauterization and suture ligation. A fourth case followed placement of a central venous line, which required neck reexploration upon patient awakening with an expanding hematoma. An oozing defect in the sidewall of a small vessel feeding into the internal jugular vein was repaired. Of note, FOLE therapy was never interrupted in any of the events. No distinct pattern among those that developed CSPPB (Figure 1) nor deaths were seen.

Conclusions: While hemostatic changes with fish oil therapy exist, we hereby show that these may not translate into an increased risk of significant bleeding. Our data suggest that fish oil monotherapy is safe, with a postprocedural bleeding risk that does not exceed the 2%–8% that is generally reported. FOLE may thus continue to be offered as a lifesaving therapy for parenteral nutrition–dependent children with IFALD and should not be held in preparation for invasive procedures, let alone interrupted in the event of CSPPB.


Table 1. Patient Characteristics and Procedure Information.

Subjectsa Statistic

N 180

Male sex 105 (58.3)

NEC diagnosis 84 (46.7)

Age at Omegaven initiation, d 91 (56, 176)

Ever on anticoagulant 19 (10.6)

Ever postoperative bleeding 4 (2.2)

Procedures

N 728

Type

Vascular accessb 315 (43.3)

Abdominal 206 (28.3)

Endoscopic 103 (14.2)

Skin and soft tissue 50 (6.9)

Thoracic 23 (3.2)

ENT 15 (2.1)

Ophthalmologic 8 (1.1)

Orthopedic 3 (0.4)


Neurosurgical 3 (0.4)

Other 2 (0.3)

Age at time of procedure, mo 10.1 (5.2, 24.0)

Omegaven exposure, mo 4.6 (1.5, 13.0)

Anticoagulant prior to procedure 50 (6.9)

Preoperative labs

Direct bilirubin, mg/dL 0.3 (0.1, 2.4)

ALT, U/L 53 (32, 101)

Albumin, g/dL 3.2 (2.8, 3.7)

PTT, s 31.6 (28.5, 37.7)

haPTT, s 32.6 (29.9, 36.4)

INR 1.17 (1.09, 1.28)

Platelets (per uL) 246,000 (150,000, 316,000)

Hemoglobin, g/dL 10.1 (8.8, 11.3)

Hematocrit, % 30.7 (27.1, 34.1)

ARA, µmol/L 379 (272, 532)

DHA, µmol/L 867 (556, 1319)

EPA, µmol/L 907 (596, 1247)

ALA, µmol/L 37 (25, 57)

LA, µmol/L 1093 (712, 1531)

T:T ratio 0.020 (0.014, 0.027)

EFAD 2 (0.3)

Postoperative labs

Hemoglobin, g/dL 10.2 (8.8, 11.4)

Hematocrit, % 30.6 (27.3, 34.5)

Postoperative bleeding 4 (0.55)

ALA, alpha-linolenic acid; ALT, alanine aminotransferase; ARA, arachidonic acid; DHA, docosahexaenoic acid; EFAD, essential fatty acid deficiency; ENT, ear, nose, and throat; EPA, eicosapentaenoic acid; haPTT, partial thromboplastin time, heparin absorbent; INR, international normalized ratio; LA, linoleic acid; NEC, necrotizing enterocolitis; PTT, partial thromboplastin time; T:T, triene:tetraene.

aStatistic: n, n (%), or median (IQR).

bIncluded only those that were tunneled or required cut-down technique for insertion.


Omegaven exposure (mos)

n=727/728 (99.9%)

Direct bilirubin (mg/dL)

n=699/728 (96.0%)

ALT (U/L) n=702/728 (96.4%)

Platelets (uL) n=714/728 (98.1%)

INR n=692/728 (95.1%)

Median (IQR)

4.6 (1.5, 13.0) mos Median (IQR) 0.3 (0.1, 2.4)

Median (IQR) 53 (32, 101)

Median (IQR) 246000 (150000,

316000)

Median (IQR) 1.17 (1.09, 1.28)

Pt 1 = 57.0 mos (1734d)

Pt 1 = 0.4 Pt 1 = 82 Pt 1 = 68000 Pt 1 = 1.23

Pt 2 = 0.4 mos (12d)

Pt 2 = 9.0 Pt 2 = 262 Pt 2 = 239000 Pt 2 = 1.05

Pt 3 = 6.2 mos (188d)

Pt 3 = 16.0 Pt 3 = 254 Pt 3 = 148000 Pt 3 = 1.60

Pt 4 = 8.1 mos (247d)

Pt 4 = 0.1 Pt 4 = 49 Pt 4 =262000 Pt 4 = 1.28

Figure 1. Distribution of time of exposure of Omegaven® at the time of surgery and preoperative data.

S32—Parenteral Nutrition Utilization and Identifying the Correct Population for Clinical Trials of Hepatoprotective Lipid Emulsions Timothy Sentongo, MD1; Jonathan Cordova, DO1; Dana Weinstein, RD1; Melanie Purser, RD1; Ellen Newton, RN1; Stefano Guandalini, MD1 1Pediatrics, University of Chicago, Chicago, IL, USA Purpose: Current Parenteral Nutrition (PN) with soy oil based lipid emulsions (SOLE) is a life-saving mode of nutrition support; however, it is associated with parenteral nutrition associated liver disease (PNALD) thus prompting novel alternatives. There is a need for appropriate subject selection in clinical trials examining hepatoprotective effects of novel PN solutions. This study examined the utilization patterns and prevalence of metabolic abnormalities and PNALD in a large population of hospitalized infants receiving PN therapy with SOLE.


Methods: This was a retrospective study of children hospitalized at a tertiary care hospital who received PN support during a 12 month period. Parameters assessed included inpatient unit, birth weight (g), duration on PN therapy, occurrence of metabolic abnormalities (metabolic acidosis, hyponatremia and hyperglycemia) and conjugated bilirubin >1.5 mg/dL (PNALD) during PN therapy. Metabolic acidosis was defined as HCO3 <22 mmol/L; hyponatremia: Na <135 mmol/L; hyperglycemia: glucose >125 mg/dL. Data was summarized into medians and interquartile ranges. Associations between prevalence of metabolic abnormalities, PNALD and duration of PN therapy were examined.

Results: The neonatal intensive care unit (NICU) accounted for 70% of all hospitalized pediatric patients on PN therapy. The median (interquartile range) birth weight of infants on PN therapy (n = 348) was 1618 g (1070 to 2365 g), duration on PN 1.3 weeks (0.6 to 2.9 weeks), and 60% weaned off PN in ≤ 2 weeks. One hundred thirty-two infants (38%) received PN for ≥ 2 weeks; 84/348 (24%) received PN for ≥ 3 weeks and; 54/348 (16%) received PN for ≥ 4 weeks. Thus among the 132 infants who received PN for ≥ 2 weeks, 54 (41%) received it for at least 4 weeks. During the first 2 weeks on PN therapy the prevalence of metabolic acidosis was 80%; hyponatremia 45%; hyperglycemia 46% and; PNALD 5%. Among infants on PN therapy ≥3 weeks (n = 82), the prevalence of PNALD was 46%. Occurrence of PNALD was associated with duration on PN therapy >2 weeks (p<0.001) but not gestation age or birth weight. Occurrence of hyperglycemia during the 1st week on PN therapy was associated with a 5-fold risk for remaining on PN longer than 3 weeks OR 4.63 (95% CI, 2.91, 7.39).

Conclusions: In this large sample of hospitalized pediatric patients, the highest utilization of PN was in preterm and sick newborn infants; however, most got weaned off therapy within 2 weeks. Electrolyte abnormalities and glucose control were the most frequent problems in infants on PN ≤ 2 weeks while PNALD was most prevalent after ≥3 weeks on therapy. This pattern suggests that clinical trials examining hepatoprotective efficacy of novel PN solutions should target enrollment of infants at 2 – 3 weeks on PN therapy Financial support: None.

S33—Parenteral Nutrition Guidelines Reduce Non-Established Use Heather E. Skillman, MS, RD, CSP, CNSC1; Amy K. Schwartz, RD, CSP, CNSC1; Whitney D. Gillies, RD, CNSC1; Joseph A. Albietz III, MD2; Timothy R. Stidham, MD2; Nancy F. Krebs, MD1,3 1Clinical Nutrition, Children's Hospital Colorado, Aurora, CO, USA; 2Pediatrics, Section of Critical Care, University of Colorado School of Medicine, Aurora, CO, USA; 3Pediatrics, Section of Nutrition, University of Colorado School of Medicine, Aurora, CO, USA Purpose: Guidelines for critically ill children recommend enteral nutrition (EN) as the preferred route of nutrition support to optimize patient outcomes. Use of parenteral nutrition (PN) is indicated only when EN is not feasible or is unable to meet nutrition needs. We implemented a quality improvement project to reduce the use of non-established PN in the Pediatric Intensive Care Unit (PICU).

Methods: Baseline data on PN use in PICU patients was collected and evaluated. The literature was reviewed to determine evidenced based indications for PN. PICU PN Guidelines were created, revised, approved by consensus, and implemented (see Table 1). In-services, fliers, emails, interdisciplinary rounds, and one-on-one discussions facilitated guideline dissemination. Consecutive patients receiving PN were tracked before and after guideline implementation. To balance the goal of reducing non-established PN, patients receiving EN were monitored to ensure that enteral feeding practices were maintained or improved. Comparisons between baseline and post-guideline implementation data were performed using one-tailed Student’s T-Tests, and Levene’s test was used to assess for equality of variance. SPSS® Statistics 21 was used with significance set at P <.05 for all analyses. Descriptive statistics are reported as mean +/- standard deviation. Children's Hospital Colorado’s Organizational Research Risk and Quality Improvement Review Panel approved this as a quality improvement project; therefore, institutional board review was not required.

Results: During the baseline data collection in 2013, 33 PICU patients received PN; of which 9 did not meet established guidelines (see Table 2, Figure 1). Variances from the indications included patients receiving PN for established conditions such as small bowel obstruction and ileus, but for less than the designated duration of at least 3 or 5 days according to age and nutrition status. After implementation of PICU PN guidelines, only 2/33 patients receiving PN did not meet criteria (P = .000 vs. baseline). Both patients received 1 day of PN; one patient received 1


day of PN for a mild gastrointestinal bleed and concern for pneumatosis; the other received 1 day of PN followed by 2 days of intravenous fat emulsion for inadequate EN (see Table 2). During the 2013 baseline data collection, 45% +/- 6% of patients who received EN initiated feeds in less than 24 hours from admission, and 17% +/- 2% of these met less than 50% calorie goal throughout the PICU stay. In 2015, 53% +/- 4% of patients receiving EN were fed in less than 24 hours (P = .027 vs. baseline), and 8% +/- 1% met less than 50% calorie goal (P = .000 vs. baseline). See Figure 1.

Conclusions: Adherence to evidenced based PICU PN Guidelines significantly reduces non-established PN use without a negative impact on EN delivery. Ongoing surveillance indicates the improved PN practices have been sustained. Efforts to optimize enteral feeding of critically ill children are essential to reduce nutrition deficits and reliance on PN to meet calorie and protein goals. Financial support: None.

Critical Care Posters S34–S82 S34Weight Loss During Hospitalization of HIV+ Patients With Respiratory Diseases Is Associated With Patients’ Body Mass and CD4+ Cell Count Adriana Aguilar Vargas, MSc1; Christopher E. Ormsby, MSc1; Angelica Leon Tellez Giron, MD1; Andrea Villazon de la Rosa1; Giselle Cheskin1; Marco Antonio Melgarejo Hernandez, Nutritionist2; Gustavo Reyes-Teran, MD1 1Departamento de Investigación en Enfermedades Infecciosas, Instituto Nacional de Enfermedades Respiratorias, Mexico City, Mexico; 2Centro de Atencion Integral del Paciente con Diabetes, Instituto Nacional de Ciencias Medicas y Nutricion Salvador Zubiran, Mexico City, Mexico Purpose: In Mexico most of the patients living with HIV do not know their diagnosis until they are hospitalized of an opportunistic infection and weight loss. Malnutrition is associated with poor prognosis, clinical complications and longer hospital or intensive care unit stay. The aim of the study is to determine the factors that are associated with poor clinical outcomes in patients living with HIV+ who are first admitted to the hospital. Methods: We evaluated retrospective data from hospitalized patients from the Instituto Nacional de Enfermedades Respiratorias in Mexico City (from 2011 to 2014). We analyzed data from 298 participants if they were over 18 years old, with documented HIV infection. Data collected included: age, weight, height, triceps skinfold thickness, and mid arm, calf and abdominal circumferences, opportunistic infections, CD4+ count, viral load, gastrointestinal symptoms, food supplements during hospitalization and clinic laboratories from admission to hospital discharge. We carried out variable reduction by principal component analysis and then used the selected variables in a multivariate logistic regression model. Results: We included 298 HIV+ patients (9.7% female) with median CD4+ count of 63 (interquartile range, IQR, 21-178.5). Median age was 34 years (IQR 24-48) and median hospital stay was 15 days (IQR 10-25). 29.5% were hospitalized with pneumonia (22.8% by pneumocystis), 22.1% with tuberculosis, 24.5% with other infectious agents and 9.1% with chronic diseases, and the rest with other pathologies such as cancers. 21.5% required mechanical ventilation, and 15.4% died. We selected from 89 clinical and laboratory variables those that explained the same variance through a principal component analysis, reducing them to weight at hospitalization (WH, predicting other weight measures), mechanical ventilator use (MV, predicting clinical outcome), albumin (Alb), gender, creatinine, diarrhea, plasma electrolytes, diagnosis, CD4+ count, and food supplements during hospitalization (FS), which were then used in a multivariate logistic regression for hospitalization weight loss. MV was independently associated with the percentage of weight loss/gain (p=0.012), together with Alb (p=0.005) and FS (p<0.001). Weight loss was in turn significantly predicted by WH (p<0.001) and CD4+ count (p=0.019). The other weight associated variables (calf, arm and abdominal circumferences and BMI) were associated univariately with weight loss (p’s<0.05), but did not independently predict it in the multivariate analysis. Conclusions: We conclude that as expected, avoiding weight loss during hospitalization is associated with better clinical outcome. However, weight gain is determined by CD4+ counts at hospitalization and by the patients’ absolute mass more than relative body composition variables, diagnosis or laboratory values. Financial Support received from: none.

S35—Pan Hypopituitarism by Manganese Brain Intoxication During Long-Term Parenteral Nutrition Marcia Tanaka, RD1; Maria T. Zanella, MD, PhD2; Adriano O. Seixas, MD3; Juliana B. Barban, RD1; Silvia M. Piovacari, RD1; Ana Paula N. Barrere, RD1; Fabiana Lucio, RD1; Juliana M. Nabarrete, RD1; Nelson Hamerschlak, MD, PhD4; Andrea Z. Pereira, MD, PhD4,2 1Nutrition, Hospital Israelita Albert Einstein, Sao Paulo, Brazil; 2Endocrinology, UNIFESP, Sao Paulo, Brazil; 3Endocrinology, Hospital Israelita Albert Einstein, Sao Paulo, Brazil; 4Oncology, Hospital Israelita Albert Einstein, Sao Paulo, Brazil Purpose: Manganese (Mn) is an essential trace element, a constituent of parenteral nutrition (PN). However, long-term PN administration can be toxic to the brain, and there are few studies about Mn intoxication. Most of them has shown that brain deposition and some neurologic symptoms, mainly parkinsonism, as a result of Mn administration in PN. It can be seen by Magnetic Resonance Imaging (MRI). Methods: We report one case of an adult patient, male, 34 years old, receiving long-term parenteral nutrition with 0.4 mg Mn /day, from August 2014 to March 2015. He had severe Crohn's disease, XIAP Mutation. HSCT was considered as a treatment due to severity of his disease. In March 2015, the patient had several episodes of


hypoglycemia during the day. His serum level of cortisol, testosterone, thyroid hormones were low. Brain MRI showed high spontaneous signal on T1 of the globus pallidus and substantia nigra more tenuous, midbrain tectum, higher brain pedundulcos, pontine integument, dentate nucleus and anterior pituitary, bilateral and symmetrical. Results: The patient was diagnosed of Pan Hypopituitarism caused for MN brain intoxication. The Mn was excluded of PN. He was undergoing HSCT in April 2015, in Bone Marrow Transplantation Division, S.Paulo, Brazil. In addition, he didn’t have any HSCT complications. Conclusions: Although Mn brain intoxication is not an usual disorder in long-term parenteral nutrition, we must remember and monitor it in our patients. Financial Support received from: None

Encore abstract: Presented at NACF (October, 2015 North American Cystic Fibrosis Conference). Not published at request of the authors.

S36—Adherence With Pancreatic Enzyme Replacement Therapy Is Associated With Reduced Hospitalization in Adults With Cystic Fibrosis Bruce Trapnell, MD1; Traci Jerkins, RD, LD2; Su Chen, PhD3; Rupal Khurmi, MD3; Mark Haupt, MD3 1Cincinnati Children’s Hospital Medical Center, Cincinnati, OH, USA; 2University of Cincinnati Medical Center, Cincinnati, OH, USA; 3AbbVie Inc., North Chicago, IL, USA Financial Support received from: AbbVie Inc. funded the research and writing of this abstract.

S37—Effects of Altered Serum Phosphorus on Mortality of Critically Ill Elderly Patients Receiving Artificial Nutrition Diana B. Dock-Nascimento1; Amanda C. Ribeiro1; José E. de Aguilar-Nascimento2 1Nutrition and Dietetics, Federal University of Mato Grosso, Cuiaba, Brazil; 2Medical School, UNIVAG - Centro Universitário, Cuiabá, Brazil Purpose: Phosphorus is an essential mineral normally found in nature combined with oxygen. Normal range of serum phosphorus is 2.5-4.5 mg/dL. In critically ill patients phosphorus is essential for the formation of adenosine, glucose-6-phosphatus and other metabolic-energetic compounds. The well recognized re-feeding syndrome is partially explained by hypophosphatemia. On the other hand, sequential organ failure is associated with hyperphosphatemia. Elderly critically ill patients are most prone to mortality in critical care settings. We wonder whether mortality in elderly patients in a tertiary critical unit is affected by altered levels of serum phosphorus. The aim of this study was to correlate mortality in critically ill elderly patients with levels of serum phosphorus during the first week of artificial nutrition implementation. Methods: A retrospective cohort study including 1,027 adult critically ill patients (650 elderly patients) receiving artificial nutrition (EN, PN, or EN+PN). Blood samples for serum phosphorus assays were collected in all cases before (P1) the initiation of early nutrition (<48h) and 48-92h after (P2) implementation of artificial nutrition. After hemodynamic stability nutrition was initiated as early as possible reaching a target of 20-25 kcal/kg/ day and 1.2-1.5 g of proteins/kg/day. When found, either hypo- or hyperphosphatemia was managed according to the ICU protocol. Mortality was correlated with the level of serum phosphorus at both P1 and P2 Results: In this cohort, 650 (60,6%) elderly (mean age =74y/o; 343 (52,8%) females) patients were included in the study. Mortality was 24% (n=156). An altered value for initial serum phosphorus was found in 269 (41.4%) patients: 176 (27.1) with hypo- and 93(14.3%) with hyperphosfatemia. Hypophosphatemia rate increased to 36% (n=234) and hyperphosfatemia to 20% (n=130) of the cases despite treatment at P2. The odds ratio (OR) for mortality was 1.34 (95%CI 0.89-2.02; p=0.16) at P1 and 1.12 (95%CI 0.85-1.48; p= 0.45) at P2 for hypophosphatemia in comparison to normal phosphatemia. However, the OR for hyperphosphatemia was 2.5 (95% CI 1.6-4.07; p=0.000) at P1, and 15.6 (95% CI 8.1-30.1; p=0,000) at P2 versus normal serum phosphorus. Cumulative survival using K-M tables during the 1st week was similar to patients with hypophosphatemia and normal phosphatemia. However, hyper-phosphatemia survival rate at P1 was lesser compared to both hypo- (9.7% vs. 63% vs, p=0.001) and normal phosphatemia (9.7% vs. 80.8%, p=0.000). The figures were very similar for P2 (65.5% vs 19.4%, p=0.000 for hypo vs. hyperphosphatemia, and 73.9% vs 19.4%, p=0.000 for normal to hyperphosphatemia). Conclusions: Hyperphosphatemia is associated with greater mortality than hypo- or normal phosphatemia in critically ill elderly patients receiving artificial nutrition.

Financial Support received from: None


S38—Magnesium Oxide Does Not Worsen Outcomes of Mice With Carcinomatous Peritonitis Eiji Tominaga, PhD1,2; Kazuhiko Fukatsu, MD, PhD2; Yutaka Nishihara, PhD1; Satoshi Murakoshi, MD, PhD2; Motonari Ri, MD3; Tomoki Watanabe, MD2; Midori Noguchi, BD2; Natsuko Yamada, BD1; Miyuki Ohnawa, BD1; Hiroshi Yasuhara, MD, PhD2 1Planning, Coordination & Information Division, Yoshida Pharmaceutical Co., Ltd, Tokyo, Japan; 2Surgical Center, The University of Tokyo Hospital, Tokyo, Japan; 3Department of Gastrointestinal Surgery, The University of Tokyo Hospital, Tokyo, Japan Purpose: Clinically, carcinomatous peritonitis (CP) frequently causes constipation due to malignant ascites formation, opioid use and/or GI tract obstruction. Magnesium oxide (MgO) is often and readily prescribed for constipation to CP patients in Japan. However, MgO might worsen the outcomes of these patients because it increases fluid extraction into the gut lumen and may thereby cause dehydration. Herein, we examined influences of MgO on the survival times and body weights (BW) of mice with PancO2-induced CP. Methods: Male C57BL/6J mice (8 weeks old, n=20) were inoculated intraperitoneally with the PancO2 cell line, generated from pancreatic cancer cells derived from C57BL/6 mice, and were randomized to the normal diet (ND) or the normal diet with 1,500 ppm of MgO (MD) group. Mice received the respective diets for 50 days. Survival, BW, weight change, and food intake amount were recorded during this period. Results: All mice had died by day 50 due to CP progression. There was no significant difference in survival times between the ND and MD groups (log-rank test, p=0.65). BW and food intake were also similar in the two groups during the observation period (t-test). Conclusions: MgO treatment does not exert detrimental effects on survival or BW in the murine CP model used in this study. Financial Support received from: Yoshida Pharmaceutical Co., Ltd.

Figure 1. Survival curves after PancO2 inoculation. Table 1. Body weight and food intake. MD ND

days initial 9 20 30 40 50 initial 9 20 30 40 50 No. of alive

mice 10 10 10 9 6 0 10 10 10 10 7 0

BW (g) 23.6±0.3 25.0±0.4

26.2±0.5

23.2±0.8

23.6±1.1 - 23.6±0

.3 24.9±0

.3 26.3±0

.3 23.4±0

.7 24.1±1

.0 -

Food intake (g)/day 2.5±0.1 3.1±0.1 2.7±0.

1 2.4±0.

4 2.0±0.

2 - 2.5±0.1

3.0±0.1

2.7±0.1

2.4±0.4

1.7±0.3 -

Date: means±SE


S39—Tolerability of a Volume-Based Enteral Nutrition Protocol in Critically Ill Patients Receiving Intravenous Vasopressor Therapy Erin L. Dickert, PharmD, BCPS1; Phillip L. Mohorn, PharmD, BCPS2,1; Jessica A. Justice, RDN, LD, CNSC1; Tanya O. Ezekiel, PharmD, BCPS1; Jeremy M. Reeves, MD, FACS3; Christopher M. Watson, MD, FACS3 1Department of Pharmacy and Clinical Nutrition, Palmetto Health Richland, Columbia, SC, USA; 2South Carolina College of Pharmacy, Columbia, SC, USA; 3Palmetto Health Surgical Specialists, Columbia, SC, USA Purpose: Current enteral nutrition recommendations advocate for the initiation of feeds within 24 to 48 hours of intensive care unit admission provided appropriate resuscitation and hemodynamic stability; however, hemodynamic stability is not well defined. Protein-energy provision volume-based feeding is an innovative enteral nutrition protocol that involves daily volume-based feed goals, liberalized gastric residual volume thresholds, and increased protein and calorie intake. Recent literature has reported safety and tolerability of enteral nutrition in patients with vasopressor support of 12.5 micrograms per minute (mcg/min) of norepinephrine equivalents or less. The purpose of this study is to determine the tolerability of a volume-based enteral nutrition protocol in medical and surgical/trauma intensive care unit (ICU) patients receiving intravenous vasopressor support. Methods: This was a single center, retrospective, observational, cohort study for adult patients admitted to a teaching hospital medical or surgical/trauma ICU initiated on volume-based enteral nutrition from December 1, 2013 to October 1, 2014. Patients were then subdivided into groups based on presence or absence of vasopressor therapy, with further categorization based on the individual vasopressor agent. Patient demographics, pertinent laboratory values, tube feed rates and residual volumes were collected. Vasopressor therapy, including dose and duration, was collected and converted to norepinephrine equivalents (mcg/min). The primary objective was to determine the tolerability of enteral nutrition in the presence of vasopressor therapy. Tolerance was defined as an absence of gastric residuals greater than 300mL in addition to documented episodes of emesis, positive findings on abdominal imaging, or evidence of ileus, bowel ischemia or perforation. Categorical and continuous variables were analyzed using Chi Square and Student’s t-test, respectively. This study was approved by the Institutional Review Board. Results: Fifty patients met criteria for study inclusion, with 12 patients (24%) who concomitantly received enteral nutrition and vasopressor therapy. All patients included achieved goal feeding rates based on caloric calculations by a registered dietitian. Baseline characteristics were similar between the two groups. Patients in the intolerable group had a higher actual body weight on admission compared to the tolerable group (median (IQR): 114 (37.5) vs. 83 (32), p=0.055) as well as a longer length of stay in the ICU (median (IQR): 18.5 (9.4) vs, 7.9 (18.5), p=0.048). Norepinephrine was the most common vasopressor utilized, with the most common indications for vasopressor therapy similar between septic shock and permissive hypertension (25%, respectively). Norepinephrine equivalents between the two groups were similar with a median (IQR) of 8 (9.75) mcg/min in the tolerable group versus 9 (19) mcg/min in the intolerable group. Of the three patients (6%) who experienced intolerance, two cases were confirmed with positive radiographic imaging and one patient developed a clinically noted ileus. Two intolerant patients (66%) required vasopressor therapy. Conclusions: Administration of volume-based enteral nutrition at goal to patients receiving IV vasopressor therapy appears to be reasonably safe, with a low incidence of intolerance. No serious adverse events were observed in either group. More research is needed to determine the tolerability of volume-based enteral feeds in the presence and absence of vasopressor therapy in critically ill patients. Financial Support received from: None

S40—Estimation of Energy Expenditure in Mexican People Living With HIV: The Cieni Equation Ivan A. Osuna-Padilla, MD1; Adriana Aguilar-Vargas, MD1; Andrea Villazón-De la Rosa, RD1; Yuria Ablanedo-Terrazas, MD1; Christopher E. Ormsby, PhD1; Gustavo Reyes-Teran, PhD1 1Centro de Investigación en Enfermedades Infecciosas, Instituto Nacional de Enfermedades Respiratorias, Distrito Federal, Mexico Purpose: The knowledge of Resting Energy Expenditure (REE) in People Living with HIV (PLHIV) is important for the prescription of Medical Nutrition Therapy. Indirect calorimetry (IC) is considered the gold standard, however due to its cost is not available everywhere. Predictive equations are useful for a practical and low-cost estimation of REE. Few predictive equations are developed for PLHIV, and some authors have reported an increment in REE in this population linked to the HIV infection, antiretroviral therapy and a chronic inflammation. The aim of the study


was to develop a predictive equation for resting energy expenditure in Mexican PLHIV. Methods: This is an observational, cross-sectional study. The sample consisted of PLHIV who initiate antiretroviral treatment with CD4 count <200 cells/mL. Anthropometric measurements included body weight, height, and circumferences (waist, hip and abdominal). Body composition was measured using whole-body DXA imaging, which measures fat free mass and fat mass. REE was measured using an IC monitor (VMAX ENCORE PFT System) and conditions for the measurements were accomplished as protocol indicates. Multivariate analyzes were applied to obtain a prediction equation for REE. T test, Intraclass correlation coefficient and Bland-Altman plot were used to compare the differences. The statistical significance (p) was set at 5%. Results: We included data from 65 patients (85% male). The mean age was 33.6 ± 8.6, and the mean BMI was 22.2 kg/m2 ± 4.0. REE mean was 1533.4 kcal ±223.4. All participants lost weight before the measurement of IC, with a mean of 11.7±10.5% in the last six months. (Table 1) The REE equation that resulted from the statistical model had an R2=0.17 and a bias of ±137.8 kcal. When we compared the REE obtained from IC and the REE estimated by the CIENI equation, there was no significant difference. We developed the following equation: REE= 464.814 + (Weight in kg x 5.345) + (Fat Free Mass in kg x 16.569). Conclusions: This study is the first to elaborate an equation to predict REE in Mexican PLHIV. The new formula has an acceptable correlation with IC. This equation may represent a useful tool for health care professionals who do not have access to IC technology. The relevance of the study is to have a better understanding of the nutritional intervention in order to improve the nutritional status of our HIV/AIDS population who has a severe weight loss before the first nutritional counseling session. Further research is needed for the validation of this equation. Financial Support received from: None


S41—Identifying and Treating Malnutrition: A Case Study of a Patient with Stage IV-A Head and Neck Cancer with a Percutaneous Endoscopic Gastrostomy (PEG) Jennifer R. Bridenbaugh, MS, RDN, CNSC1; L. Jeanne Gee, RDN, CSO, CDE2; Jane Ziegler, DCN, RDN, LDN 1 1Nutritional Sciences, Rutgers University, Newark, NJ, USA; 2Hunterdon Regional Cancer Center, Flemington, NJ, USA Purpose: Identification and treatment of malnutrition is essential in the care of patients with cancer. Patients with Head and Neck Cancer (HNC) have a higher risk of malnutrition than those with other cancers. Weight loss is a predictor of malnutrition and overall morbidity and mortality in the HNC population. Close weight monitoring is essential in the treatment of HNC. Interventions should include specific medical nutrition therapy and possibly enteral nutrition (EN) support with goals to improve weight stability to potentially improve morbidity, mortality, infections, and hospital readmission.

Description: 61 y/o male Diagnosis: Squamous Cell Cancer (SCCA) right tonsil cT2N2bM0 stage IV-A; former smoker. Treatment regimen: Concurrent chemoradiation (Cisplatin). During treatment the patient experienced fatigue, significant xerostomia, dysguesia, odynophagia and thick oral secretions. Other findings included abdominal pain, nausea, constipation, diarrhea, and decreased appetite. Initial Nutrition Focused Physical Exam (NFPE) – no significant findings Pre-treatment weight: 207#/94 kg Height: 5’10” /180 cm Pre-treatment Body Mass Index (BMI): 30. Other Medications: Diflucan, Magic Mouthwash (Benadryl/viscous lidocaine/Maalox), Omeprazole, Prochlorperazine Maleate, Nystatin, Sancuso patch Nutrition Problems/Diagnosis: Malnutrition related to significant decline in oral intake and weight as evidenced by oral intake <50% estimated energy requirement for >5 d and 6.7% weight loss in 1 month. Unintentional weight loss related to cancer treatment symptoms: xerostomia, dysguesia, nausea as evidenced by overall 7.4% weight loss during cancer treatment. Interventions: Initially over the first month of treatment calorically dense foods and supplements were encouraged along with frequent and ongoing nutrition education with patient and family. A PEG was placed on treatment day (TD) 30 due to progressive weight loss, odynophagia, poor oral intake, and increased dehydration. With PEG placement the patient received 1200cc of a 2 calorie/cc formula providing 2400 calories and 100 g protein. Additional free water (1800cc daily) was also prescribed to maintain hydration. Monitoring/Evaluation: Prior to the PEG (TD 1 to TD 30) 1-2% weight loss per week was observed (6.7% total) with oral intake <50% estimated energy needs. After the PEG (TD 30 to TD 73) the patient tolerated the EN without any reported complications and weight remained stable (192#/87kg) and progressive weight loss was stopped. Oral intake was trialed when symptoms improved and as oral intake increased (>50% estimated energy needs) the EN was decreased by approximately 240cc EN/week. Once treatment was finished and weight remained stable without the use of PEG feedings, the PEG was removed and the patient was maintained on oral intake. The PEG was successful in providing weight stabilization for the patient. Malnutrition was attenuated through initiation of EN while the patient was in active treatment. Changes in weight and oral intake are only 2 parameters that can be used in identifying the presence of malnutrition. Use of a hand grip dynamometer and nutrition focused physical exam findings obtained by the Registered Dietitian Nutritionist can also assist in identifying and classifying malnutrition. Financial Support received from: none


S42—Early Oral Feeding as Compared to Traditional Timing of Oral Feeding After Upper Gastrointestinal Surgery: A Systematic Review and Meta-Analysis Kate Willcutts, DCN, RD, CNSC1,4; Mei Chun Chung, PhD, MPH2; Cheryl Erenberg, MLS3; Kristen Finn, DCN, RD4; Bruce Schirmer, MD5; Laura Byham-Gray, PhD, RD4 1Patient Care Services, University of Virginia Medical Center, Charlottesville, VA, USA; 2Department of Public Health and Community Medicine, School of Medicine, Tufts University, Boston, MA, USA; 3George F. Smith Library of the Health Sciences, Rutgers University, Newark, NJ, USA; 4Department of Nutritional Sciences, Rutgers University, School of Health Professions, Newark, NJ, USA, USA; 5Surgery, University of Virginia Medical Center, Charlottesville, VA, USA Purpose: Early oral feeding after all types of surgeries is becoming more common, particularly as part of multimodal or fast track perioperative protocols. However, there are still concerns about the safety of early initiation of oral feeding after upper GI surgery. The purpose of this study is to compare the effects of early oral feeding to traditional (or late) timing of oral feeding after upper gastrointestinal (GI) surgery on clinical outcomes. Methods: Comprehensive literature searches were conducted across five databases from January 1980 until June 2015 without language or study design restriction. Risk of bias of included studies were appraised and random effects model meta-analyses were performed to synthesize results of anastomotic leaks, pneumonia, need for nasogastric tube reinsertion, reoperation, mortality and readmission. The selected studies were randomized controlled trials, cohorts with historical controls, and a retrospective cohort study in adult patients. A random-effects model for meta-analyses was used for pooled analysis. Results: Fifteen studies including a total of 2112 adult patients met all the inclusion criteria. Mean hospital stay was significantly shorter in the early fed (EF) group than in the late fed (LF) group (weighted mean difference = -1.72


days; 95% CI, -1.25 to -2.20; P < 0.01) and when reported as the postoperative LOS, it was also significantly shorter (weighted mean difference = -1.44 days; 95% CI, -0.68 to -2.20; P < 0.01). There was no significant difference in risk of anastomotic leak in just the randomized controlled trials (RCT)s or the RCTs pooled with the non-RCTs. Nor were there significant differences in the risk of pneumonia, need for nasogastric tube (NGT) reinsertion, reoperation, readmission or mortality in the RCTs. The pooled RCT + non-RCT results however, showed a significantly lower risk of pneumonia in EF as compared to LF (OR = 0.6; 95% CI 0.41 to 0.89; P = 0.01). Conclusions: Early postoperative oral feeding as compared to traditional (or late) timing of postoperative oral feeding results in a statistically significant shorter LOS and without an increase in anastomotic leaks, pneumonia, need for NGT re-insertion, reoperation, readmission or mortality. Financial Support received from: none

Encore: Presented at NASPGHAN 2105

S43—Procalcitonin as a Predictor of Bacteremia in Children Presenting With Fever from Possible Central Line Associated Blood Stream Infections Kelly B. Haas, MD1; Patricio Arias, MD1; Kt Park, MD1; Jennifer Damman, MD1; Colleen Nespor, RN1; John Kerner, MD1 1Pediatric Gastroenterology, Stanford, Palo Alto, CA, USA Purpose: Unnecessary use of broad spectrum antibiotics among hospitalized children with possible central line associated bloodstream infections (CLABSIs) leading to bacterial resistance to broad spectrum antibiotics is a major concern among children who require long term central venous access. Our current protocol for a patient who is dependent on parenteral nutrition with a central line and fever is to obtain a blood culture from the central line and start inpatient broad spectrum antibiotics pending results of the blood culture. An early biomarker for predicting bacteremia in a pediatric patient with a fever and a central line has the potential to reduce unnecessary antibiotic exposure and the development of multi-drug resistant organisms in this population. This study aims to determine if Procalcitonin (PCT) is a useful biomarker for early prediction of bacteremia in children who present with a fever and a central line, and to determine the best cutoff point of PCT (ng/ml) for identifying bacteremic patients. Methods: A retrospective chart review at Lucile Packard Children’s Hospital (August 2013-August 2014) revealed that 45% of children admitted with fever and a central line had a negative blood culture. We evaluated PCT in 31 patients who presented to the Emergency Room or clinic with fever and a central line from August 2014 to May 2015. Seventeen patients (54%) had a positive blood culture (16 bacterial and 1 fungal). The range of PCT levels was found to be between <0.1 and >20 ng/ml. The mean value of PCT in bacteremic patients was 6.45ng/ml. Higher cutoff PCT values were associated with a more severe clinical presentation and faster growth of positive blood cultures. Two patients with positive blood cultures for coagulase-negative staphlococci thought to be contaminates given very late growth were excluded from the study. Both excluded patients had a PCT value of < 0.1. Results: A PCT level of > 0.3ng/ml produced a sensitivity of 93.3% and positive predictive value of 87.5% for detecting bacteremia. The specificity of a PCT level ≤0.3 ng/ml was 85.7% with a negative predictive value of 92.3%. Given the small sample size, further studies are needed to assess the clinical utility of PCT for the early detecting of CLABSIs. Conclusions: One patient with PCT level <0.1 ng/ml yielded a positive fungal culture result at 48 hours; the use of PCT in assessing fungemia remains unclear. In conclusion, PCT represents a promising biomarker in the early detection of bacteremia in children with a fever and a central line, although further studies are needed to validate sensitivity and specificity and to determine an optimal PCT cutoff value for highest diagnostic yield. Financial Support received from: None


S44—Current Nutrition Practices in the ICU: Results of an International Quality Improvement Initiative Daren Heyland, MD, FRCPC1,2; Margot Lemieux, RD3; Xuran Jiang, MSc3 1Critical Care Medicine, Queen’s University, Kingston, Canada; 2Public Health Sciences, Queen's University, Kingston, Canada; 3Clinical Evaluation Research Unit, Kingston General Hospital, Kingston, Canada Purpose: Numerous barriers to adequate feeding exist resulting in variable and suboptimal nutrition delivery. Insights can be gained by identifying sites that achieve a high degree of compliance with nutrition guidelines. The purpose of this study was to describe current nutrition practices in Intensive Care Units (ICUs), describe variability in ICU nutrition performance, and to identify the ‘Best of the Best’ ICUs with respect to nutrition practices. Methods: We conducted an international, prospective, observational, cohort study conducted September 2014-June 2015. Participating sites enrolled mechanically ventilated adult patients that stayed in the ICU for at least 72 hours. Data on nutrition practices was collected from ICU admission to ICU discharge or a maximum of 12 days. Relative to the Canadian Clinical Practice Guidelines (CPG), we report average, best, and worst site performance on key nutrition practices. Previously reported nutrition performance indicators (NPIs) were used to identify the ‘Best of the Best’ nutrition performing sites. Results: 183 adult intensive care units from 19 countries contributed 3859 patients to this analysis. Average time to start of EN was 38.2 hours (site average range: 4.9-157.4 hours). The average use of motility agents and small bowel feeding in patients with high gastric residual volumes was 65.2% (site average range: 0-100%) and 12.4% (site average range: 0-100%) respectively. The proportion of blood glucose levels greater than 180 mg/dl was 15.4% (site range: 2.9-42.1%). Average nutritional adequacy was 48.0% (site average range 7.8-93.6%) for energy and 44.8% (site average range 7.0-99.0%) for protein. Using the NPIs, we were able to identify top and recognize top performing ICUs around the world. Conclusions: Large gaps exist between the evidence based recommendations and actual practice in ICUs, and consequently nutrition therapy is sub-optimal. We have identified top performing ICUs and ‘best achievable’ practices which can serve as targets for future quality improvement initiatives Financial Support received from: Abbott Nutrition

S45—Influences of Vagotomy on Splenic Immunity in a Murine Gut Ischemia-Reperfusion Model Motonari Ri, MD1; Kazuhiko Fukatsu, PhD2; Taiki Miyakuni, MD3; Satoshi Murakoshi, PhD2; Hiroshi Yasuhara, PhD2; Yasuyuki Seto, PhD1 1Department of Gastrointestinal Surgery, the University of Tokyo, Tokyo, Japan; 2Surgical Center, the University of Tokyo, Tokyo, Japan; 3Department of Digestive Surgery, Nihon University School of Medicine, Tokyo, Japan Purpose: The vagal reflex pathway has been demonstrated to regulate inflammatory responses through a nicotinic anti-inflammatory pathway. Using mice, we previously demonstrated that vagotomy, a standard procedure for lymph node dissection in upper GI tract malignancies, worsens survival after gut ischemia-reperfusion (I/R) with increased IL-6 and MCP-1 levels in plasma and the gut, which might be associated with a stronger postoperative inflammatory response and complications. Recent studies have suggested that this pathway works through modulation of immune cell function in the spleen as well as in the gut. We examined whether vagotomy affects splenic immunity in gut I/R mice by determining 1) spleen cytokine levels and 2) splenic macrophage numbers and activation. Methods: Experiment 1) Six-week-old male Institute of Cancer Research (ICR) mice (n=63) were divided into 4 groups (Sham: n=10, Vago: n=10, I/R: n=22, Vago+I/R: n=21). The I/R or Vago+I/R group underwent 45-minute superior mesenteric artery (SMA) occlusion with or without vagotomy, while the Vago group received vagotomy only. Mice were killed at 3h or 6h after reperfusion. We measured cytokine levels (IL-6, TNF-α, IFN-γ, MCP-1, IL-10) in the spleen using the Cytokine Beads Array method.Experiment 2) Mice (n=80) underwent the same procedures as in Ex.1 (Sham: n=13, Vago: n=13, I/R: n=27, Vago+I/R: n=27). Splenic cells were isolated at 3h or 6h after reperfusion. Splenic macrophage numbers and activities were analyzed by F4/80 and CD11b antibody using Flow Cytometry, respectively. Results: Ex.1) Gut I/R insult increased splenic pro-inflammatory cytokine levels without significant differences between the I/R and Vago+I/R groups. (Table 1.) Ex.2) There were no significant differences in macrophage number or CD11b expression between the I/R and Vago+I/R groups. (Table 2.) Conclusions: Vagotomy did not alter the splenic cytokine milieu or macrophage profile after gut I/R. Other mechanisms appear to account for the poor survival of gut I/R mice with vagotomy.


Financial Support received from: No financial support

Table1. Spleen cytokine levels (pg/ml) Sham Vago I/R Vago+I/R

IL-6 (3h/6h) 10.2±41.5 /11.3±34.8

8.67±41.5 /19.5±34.8

83.0±28.0 /78.0±23.4

89.9±28.0 /108.7±23.4*†

TNF-α (3h/6h) 7.72±32.3 /7.55±0.96

6.96±32.3 /6.79±0.96

47.0±21.8 /9.69±0.65†

8.56±21.8 /10.5±0.65*†

IFN-γ (3h/6h) 0.42±0.11 /0.47±0.12

0.41±0.11 /0.55±0.12

0.83±0.07*† /0.80±0.08*

0.69±0.07 /0.80±0.08*

MCP-1 (3h/6h) 28.4±38.2 /85.2±39.6

12.7±38.2 /91.3±39.6

126.2±25.8*† /217±26.7*†

72.8±25.8 /250±26.7*†

IL-10 (3h/6h) 0.80±1.35 /2.77±1.45

2.25±1.35 /1.16±1.45

3.40±0.91 /3.73±0.98

2.35±0.91 /3.17±0.98

Means±SE, *p<0.05 vs. Sham group, †<0.05 vs. Vago group. Table2. Splenic cell profiles

Sham Vago I/R Vago+I/R

Cell number (×10*7) (3h/6h) 3.59±0.41 /3.89±0.62

3.18±0.41 /3.64±0.62

2.69±0.28 /4.77±0.44

2.50±0.28* /4.85±0.44

F4/80 positive (%) (3h/6h) 2.50±0.48 /3.33±0.59

2.75±0.45 /2.58±0.65

3.20±0.32 /3.83±0.41

3.92±0.32*† /4.48†±0.41

Macrophage number (×10*5) (3h/6h) 8.58±2.01 /14.2±4.22

8.82±1.86 /9.73±4.62

8.67±1.32 /19.1±2.98

9.78±1.32 /22.4†±2.98

CD11b expression (×10*5 MFI) (3h/6h) 6.04±2.51 /4.25±2.90

4.08±2.32 /3.13±3.17

2.67±1.64 /7.21±2.05

2.70±1.64 /4.01±2.05

Means±SE, * p<0.05 vs. Sham group, †<0.05 vs. Vago group.

S46—Predicting Elevated Glutamine Levels in Critically Ill Patients: Role of LDH, Blood Urea Nitrogen, and Need for Dialysis Paul Wischmeyer, MD1; Christine Baird, BS1; Olav Rooijackers, PhD3; Daren Heyland, MD2 1Anesthesiology and Pediatrics (Nutrition Section), University of Colorado School of Medicine, Aurora, CO, USA; 2Critical Care Medicine, Queens University-Public Health, Kingston, Canada; 3Anesthesiology and Intensive Care, Karolinska Institutet and University Hospital, Huddinge, Sweden Purpose: Purpose: Glutamine (GLN) given in standard, accepted doses (0.3-0.5 g/kg/d) has demonstrated reduced mortality and improved outcomes in systematic reviews and recent meta-analysis when administered as a component of Total Parenteral Nutrition (TPN) However, in trials (REDOXs) of patients with admission multi-organ failure (MOF) high-dose GLN administration (> 0.5 g/kg/d) is associated with a signal of increased mortality. Secondary analysis indicates much of this increased mortality appears to have occurred in patients with renal failure patients at ICU admission who did not undergoing dialysis. Unexpectedly, we observed elevated GLN levels > 930 μmol /L in REDOXs in ~ 10% of MOF patients at ICU admit. Previous research by other groups and our REDOXs data suggest admission GLN levels > 930 μmol/L may be associated with increased mortality. GLN is stored in large quantities in muscle and immune cells, and is actively metabolized by the kidney. Thus, we hypothesize that markers of admission renal failure and cellular injury may be able to predict patients with increased GLN levels. As existing analytic methods to analyze plasma GLN are not suited for routine clinical practice, utilization of easy obtained clinical and laboratory data may help guide clinicians decide who should or should not receive GLN. Methods: Methods: A laboratory substudy in 68 patients of the multi-center REDOXS trial of GLN and


Antioxidants given independent of nutrition as a pharmacologic intervention in critically ill patients was performed to evaluate clinical and laboratory markers that may correlate with elevated plasma GLN levels. In brief, the REDOXs trial studied patients with MOF at ICU admit receiving high dose GLN (> 0.5 g/kg/d), antioxidants, both or neither. Results: Results: Increased admission LDH was found to meaningfully correlate with increased plasma GLN (> 930 μmol/L) (Pearson Correlation Coef.-0.72 (p < 0.0001). Less strong, yet statistically significant correlations were found for admission creatinine, admission APACHE II score and study day 4 Blood Urea Nitrogen (BUN) (Pearson Correlation-0.31, 0.32, and 0.32 respectively) predicting elevated GLN levels. No correlation was found for Admission Bilirubin, AST, ALT, CPK, cortisol, age, BMI, SOFA score, or vasopressor use. A combination of an elevated LDH > 271 U/L, a BUN > 40 mg/dl (14 mmol/L), and an indication leading to dialysis within 48 h of admission identified 95% of patients with elevated admission GLN levels. If BUN cutoff was lowered to >30 mg/dl (10 mmol/L) this marker combination identified 100% of patients with an elevated GLN. (See Figure 1). Conclusions: Conclusions: LDH, Elevated BUN, and an indication leading to dialysis within 48 h of ICU admit may be clinically useful predictors of elevated GLN levels. LDH is largely contained in immune cells and it’s elevation in patients without myocardial infarction may be associated with lymphocyte apoptosis, which is known to correlate with death from sepsis and critical illness. Measurement of LDH isoforms specific to immune cells could be used to explore this hypothesis. Renal dysfunction may lead to impaired GLN use and metabolism and elevation of GLN levels. An additional hypothesis for the observation of increased GLN levels in severely ill MOF patients is some patients may exhibit “GLN resistance” (an inability to utilize GLN), perhaps like the inability to utilize glucose in insulin resistant critically ill patients. These hypotheses deserve further validation. In conclusion, measures such as LDH, BUN, and need for dialysis may be used to assist clinicians in predicting which critically ill patients are at risk for elevated GLN levels and which patients are candidates for GLN treatment as part of TPN. Financial Support received from: CIHR Clinical Trial Funding


S47—A Randomized Trial of Supplemental Parenteral Nutrition in Under and Over Weight Critically Ill Patients: The TOP UP Pilot Trial Paul Wischmeyer, MD1; Michel Hasselmann, MD2; Christine Kummerlen, MD2; Rosemary Kozar, MD, PhD3; Demetrios Kutsogiannis, MD, MHS4; Constantine Karvellas, MD5; Beth Besecker,MD6; David C. Evans, MD7; Jean-Charles Preiser, MD, PhD8; Leah Gramlich, MD4; Kursheed Jeejeebhoy, MD9; Rupinder Dhaliwal, RD10; Xuran Jiang,MSc11; Andrew Day, MSc11,12; Daren Heyland, MD, MSc, FRCPC11 1Anesthesiology and Pediatrics (Nutrition Section), University of Colorado School of Medicine, Aurora, CO, USA; 2Faculté de Médecine de l'Université de Strasbourg, Centre Hospitalier Universitaire de Strasbourg Nouvel Hospital Civil, Strasbourg Cedex, France; 3Department of Surgery, Shock Trauma Center, University of Maryland, Baltimore, MD, USA; 4Department of Medicine, University of Alberta, Royal Alexandra Hospital, Edmonton, Canada; 56. Divisions of Gastroenterology and Critical Care Medicine, University of Alberta, University of Alberta Hospital, Edmonton, Canada; 6Division of Pulmonary, Critical Care, Allergy & Sleep Medicine, The Ohio State University, The Ohio State University Medical Center, Columbus, OH, USA; 7Department of Surgery, The Ohio State University, The Ohio State University Medical Center, Columbus, OH, USA; 8Department of Intensive Care, Universite Libre de Bruxelles, Erasme University Hospital, Brussels, Belgium; 9University of Toronto, Toronto, Canada; 10Clinical Evaluation Research Unit, Kingston General Hospital, Kingston, Canada; 11Department of Medicine, Queen’s University, Kingston General Hospital, Queens University, Kingston, Canada; 12Department of Community Health and Epidemiology, Queens University, Kingston, Canada Purpose: Controversy exists around optimal amount and route of feeding in ICU. Nutritional guidelines differ in recommendations on role of Parenteral Nutrition (PN) and existing clinical trial data is inconclusive. However, our recent observational data showed the amount of energy and protein received early in ICU stay impacts patient mortality, independent of delivery route. More specifically, mortality is highest among patients at extremes of BMI (BMI<25 and BMI>35) who receive lowest intake of energy and protein. Thus, we hypothesized that increased calorie and protein delivery via supplemental PN + enteral nutrition (EN) to underweight and obese, high nutrition risk ICU patients (Body Mass Index [BMI] <25 or >35) will improve 60-day survival and Quality of life compared to usual care. Prior to launching a full trial, a multicenter randomized pilot trial was conducted to assess feasibility of large trial. Methods: We conducted a block randomized pilot trial in 11 centers in Belgium, Canada, France and U.S to demonstrate feasibility of trial protocol. We included adult ICU patients expected to require mechanical ventilation for > 72 h with admission BMI < 25 or >35, based on actual/estimated dry weight. Key exclusion criteria included >72 h from ICU admission to consent. Patients randomized to receive EN only or EN+PN. Nutrition goals were 25-30 kcal/kg/d and 1.5 g/kg/d protein. Adjusted body weight was used in patients with BMI > 35. EN group received no additional PN in first 7 d post- randomization unless a contraindication to EN developed. EN+PN group received additional PN to achieve calorie goal using 3-in-1 PN solution (Olimel N9, Baxter Inc). Amount of PN+EN adjusted to attempt to administer 100% of prescribed calories daily. A Feeding protocol to standardize the provision of EN and PN was provided. The primary aim of pilot study was to attempt to achieve a 30% difference in calorie/protein delivery between two groups. Clinical outcomes including 60-day mortality, functional status at hospital discharge, lean body mass U/S, and health-related quality of life were also collected. Results: We recruited 125 patients at 11 centers with characteristics shown in Table 1. On average, participating sites enrolled 0.80 patients/site/month (range: 0.30 patients/m to 1.9 patients/m). There was moderate compliance with study procedures (Table 2). Over first 8 ICU days, EN+PN group received 18.6% more of their prescribed calories versus EN only arm (70.4±29.0% vs. 51.8±25.3% of goal respectively, P < <0.001). See Table 3 for full nutrition delivery data. In this pilot trial, underpowered for clinical outcomes, no significant differences in major clinical outcomes were observed (Table 4). Potentially interesting trends to improved hospital discharge Barthel Functional Index (p = 0.08) and an improvement in change from baseline to 6-month SF-36 vitality subscore (p=0.05) was observed in PN+EN group. However, these findings are limited due to lack of power and significant loss to follow up at 6 months. Conclusions: Conclusions: In this pilot trial of supplemental PN versus EN alone over first week of ICU stay in high nutrition risk ICU patients despite an increase in calorie and protein delivery in PN+EN group, we did not achieve our pre-determined target of a 30% difference in calorie delivery. Further, achieving goal calories delivery in the PN+EN group was more challenging then expected and compliance with study procedures around calorie


delivery and post-ICU follow up were more challenging for sites then expected. This trial will help to inform future trial design of supplemental PN in high nutrition risk ICU patients. Financial Support received from: NIH R34 HL109369 to Paul Wischmeyer and Daren Heyland


Encore Abstract: Previously Presented at ACCO May 2015. Previously Published in ASCO Meeting Abstracts May 18, 2015:6066. Abstract not re-printed at the Author’s request.

S48—Double Blind Multicenter Phase III GORTEC Trial Evaluating the Efficacy of Oral Immune Modulating Formulae Therapy During Adjuvant Radiochemotherapy in Head and Neck Squamous Cell Carcinoma (HNSCC) Pierre F. Senesse, MD, PhD1; Pierre Graff-Cailleaud, MD2; Christian Sire, MD3; Olivier Gallocher, MD4; Sandrine Lavau-Denes, MD5; Muriel Garcia-Raimrez, MD6; Marc Alfonsi, MD7; Didier Cupissol, MD8; Marie-Christine Kaminsky, MD9; Stéphanie Fiore, RDN1; Nicoals Flori, MD1; Hélène De Forges, PhD12; Renaud Garrel, MD, PhD10; Chloé Janiszewski, MSc12; Simon Thézénas, MSc13; Pierre Boisselier, MD11 1Clinical Nutrition and Gastroenterology, Montpellier Cancer Institute, Montpellier, France; 2Radiotherapy, Lorraine Anticancer Centre, Vandeouvre-Lès-Nancy, France; 3Radiotherapy, Bretagne Sud Hospital, Lorient, France; 4Radiotherapy, High Energy Centre, Toulouse, France; 5Radiotherapy, Limoges University Hospital, Limoges, France; 6Radiotherapy, Libourne Hospital, Libourne, France; 7Radiotherapy, Sainte Catherine Institute, Avignon, France; 8Oncology, Montpellier Cancer Institute, Montpellier, France; 9Oncology, Lorraine Anticancer Centre, Vandeouvre-Lès-Nancy, France; 12Research unit, Montpellier Cancer Institute, Montpellier, France; 10Surgery, Montpellier University Hospital, Montpellier, France; 13Biostatistics unit, Montpellier Cancer Institute, Montpellier, France; 11Radiotherapy, Montpellier Cancer Institute, Montpellier, France Financial Support received from: Fund provided by the National LIgue against Cancer Products provided by Nestlé International

S49—Chronic Critical Illness: Emerging Post-Acute Care Nutrition Practice Population for CNSCs Sharron Diot, MS, RD, CNSC2; Meg Stearn-Hassenpflug, MS, RD, FCCM1; Jillisa Steckart, Med, PsyD3,1 1Research, Barlow Respiratory Hospital, Los Angeles, CA, USA; 2Clinical Nutrition Services, Barlow Respiratory Hospital, Los Angeles, CA, USA; 3VA-GLAHS, Los Angeles, CA, USA Purpose: To identify the long-term acute care (LTAC) hospital as an emerging practice venue for Certified Nutrition Support Clinicians (CNSC) by characterizing the population of chronically critically ill (CCI) patients admitted to the post-acute care (PAC) setting of a LTAC regional weaning center. Barlow Respiratory Hospital (BRH) is a 105-bed LTAC hospital network that serves as a regional weaning center, accepting patients transferred from the ICUs of hospitals in southern California for weaning from prolonged mechanical ventilation (PMV). Methods: Selected patient admission characteristics and weaning outcome data were queried from the Ventilation Outcomes Database (VOD), a performance improvement database, for patients discharged from 1/1/2014-6/30/2015. VOD data were merged with those from selected electronic health record (EHR) queries constructed for department specific (clinical nutrition, pharmacy, infection control, laboratory) performance improvement measures. Results: From 1/1/2014-6/30/2015, 421 patients admitted for weaning from PMV were discharged from BRH. Patient demographics and characteristics on admission are presented in Table 1. Per pharmacy reports, patients were receiving 20 medications on average at time of admission. Outcomes: length of stay 35 [1-296] days; weaning outcomes scored at discharge: 59% weaned, 33% ventilator-dependent, 8% died. Discharge disposition: home 6%, ECF/ acute rehab 77%, short term acute care 17%. For patients discharged alive, 376/386 (97%) retained their enteral feeding tube. Conclusions: This is a descriptive report of the population of CCI patients admitted to the PAC venue of long-term


acute care for weaning from PMV. The population is largely elderly, and arrives at the LTAC on enteral nutrition support after roughly three weeks in the transferring hospital ICU; transfer documents are often with incomplete records of nutrition assessment, alimentation provided, and progress. The great majority of patients remain on enteral nutrition support throughout the LTAC stay. These are complex patients presenting with multiple co-morbid conditions that challenge the nutrition support clinician to assess, design, and implement alimentation profiles considering: fluid imbalance, glycemic control, electrolyte management, anemia, polypharmacy, chronic kidney disease/hemodialysis, infectious complications, aspiration risk, dermal failure, and altered integrity of the gut microbiota. Although patients are no longer in the ICU, it is evident that nutrition support services in the PAC venue of long-term acute care should be provided at a level of expertise appropriate to provide disease-specific nutrition assessment, diagnosis, intervention, monitoring and evaluation in a patient population where multiple system organ dysfunction and failure is the norm rather than the exception. Financial Support received from: None

Table 1. Characteristics of 421 Chronically Critically Ill Patients Admitted for Weaning from Prolonged Mechanical Ventilation

Characteristics Data Age, years 71.9 [21-100]

Gender, male 57.2% LOS transferring facility 21.0 [1-384]

Glasgow Coma Score 14 [3-15] Serum albumin (g/dl) 2.4 ± .62

Hematocrit (%) 28.6 ± 4.3 BUN (mg/dl) 37.6 ± 25.5

Serum creatinine (mg/dl) 1.2 ± 1.3 Serum glucose (mg/dl) 141.8 ± 50.5

25-OH Vitamin D total (n=34) 30.32 ± 14.6, 26 [9-71] Body mass index (BMI) 25.9 [12-72]

Enteral feeding tube (NG, G, J, GJ) 98.6% Central line 60.5% TPN/PPN 1%

Pressure ulceration≥ stage II 54.7% Multiple pressure ulcers 21.9%

Clostridium difficile 6.8% Tracheostomy tube 95.0%

NPO status 90.5% Renal replacement therapy (hemodialysis) 12.4%

Diabetes mellitus 55.6% Congestive heart failure 41.2%


S50—Chronic Critical Illness: Association Between Vitamin D Status and Weaning From Prolonged Mechanical Ventilation Sharron Diot, MS, RD, CNSC3; Meg Stearn-Hassenpflug, MS, RD, FCCM1; Jillisa Steckart, Med, PsyD2,1 1Research, Barlow Respiratory Hospital, Los Angeles, CA, USA; 2VA-GLAHS, Los Angeles, CA, USA; 3Clinical Nutrition Services, Barlow Respiratory Hospital, Los Angeles, CA, USA Purpose: There have been numerous studies looking at vitamin D status in association with various lung diseases focusing on asthma, chronic obstructive pulmonary disease (COPD), cystic fibrosis (CF), and respiratory infections. These studies have demonstrated a high prevalence of vitamin D deficiency in their participants. Our purpose was to examine the association between 25-hydroxyvitamin D (25(OH)D) concentration and outcomes of weaning from prolonged mechanical ventilation (PMV) in a cohort of ICU survivors admitted to a long-term acute care (LTAC) hospital for continued weaning attempts. Barlow Respiratory Hospital (BRH) is a 105-bed LTAC hospital network that serves as a regional weaning center accepting chronically critically ill (CCI) patients transferred from ICUs of hospitals in southern California for weaning from PMV. Methods: This was a retrospective cohort study of ICU survivors admitted to a LTAC hospital for weaning from PMV. Selected patient admission characteristics and weaning outcome data were queried from the Ventilation Outcomes Database (VOD), a performance improvement database, for patients discharged 1/1/2012-6/30/2015 with 25(OH)D concentrations drawn on admission. Subjects were divided into two groups based on their 25(OH)D concentrations (deficient, < 20 ng/mL; not deficient, ≥ 20 ng/mL), with comparison analysis performed between the two cohorts. Results: 541 patients admitted for weaning from PMV were discharged from 1/1/2012-6/30/2015 (Table 1). 68/541 (13%) had 25(OH)D concentrations drawn on admission; 23/68 (34%) were found to be deficient. There was no significant difference in weaning outcomes between the two groups. Of the deficient group, 17/23 (74%) received vitamin D supplementation at the LTAC hospital. Conclusions: Vitamin D deficiency was not as common in the BRH cohort (34%) in contrast to a study by Verceles et al (Respir Care, 2015 Jul; 60(7):1033-9), where 61% of CCI patients were found to be deficient. But as in the aforementioned report, we were not able to demonstrate a significant relationship between admission 25(OH)D concentrations and successful weaning from PMV, although substantial data exist that demonstrate an association between vitamin D insufficiency/deficiency and mortality/clinical outcomes of critically ill patients. We do not know what effect vitamin D supplementation in 74% of the deficient group may have had on outcomes. The statistically significant differences that were realized between the two cohorts were not surprising, given the known relationships between vitamin D deficiency and obesity, chronic kidney disease (CKD), and hypoalbuminemia. Limitations of our study include the fact that it was observational in nature, with relatively small numbers in each cohort. Future studies related to vitamin D status in CCI would be of interest to further our understanding of the potential underlying biological mechanisms by which vitamin D optimization may improve outcomes in these survivors of critical illness. Supplementation of vitamin D will increase the serum concentrations of this vitamin; however the optimal dose needs to be identified along with an assessment of clinical outcome. Financial Support received from: None


Table 1. Admission Characteristics and Weaning Outcomes of Chronically Critically Ill ICU Survivor Cohort by 25(OH)D Status

Variable 25(OH)D Concentration (n=68) p

Deficient < 20 ng/mL Not Deficient ≥ 20 ng/mL Subjects, n (%) 23 (34%) 45 (66%)

Age, years 72.2 [34-95] 73.8 [36-99] ns Gender, male 9 (39%) 27 (60%) ns

Race, Caucasian 16 (70%) 26 (58%) ns Premorbid location, Home 17 (74%) 29 (64%) ns Premorbid location, ECF 6 (26%) 16 (36%) ns

Premorbid Zubrod (functional status) ns "Good" functional status 17 (74%) 23 (51%) "Poor" functional status 6 (26%) 22 (49%)

LOS transferring facility, days 27 [4-64] 19 [1-138] .02 Pressure ulceration ≥ stage II 14 (61%) 29 (64%) ns

Multiple pressure ulcers 9 (40%) 12 (27%) ns APACHE III© APS 47 [28-85] 47 [15-73] ns

BMI (kg/m2 ) 29 [20-39] 24 [15-40] .007 Lab Values Mean (SD) Mean (SD)

25(OH)D (ng/mL) 15.5 (3.2) 33.9 (13.9) * Serum albumin (g/dl) 2.3 (.76) 2.5 (.49) .02

Hematocrit (%) 28.7 (4.2) 29.4 (3.6) ns BUN (mg/dl) 49.9 (34.5) 36.6 (23.2) .05

Creatinine (mg/dl) 1.4 (.93) 1.3 (1.6) ns Serum glucose, (mg/dl) 139.0 (34.8) 133.2 (40.2) ns

Enteral feeding tube (NG, G, J, GJ) 22 (96%) 45 (100%) ns Hemodialysis 8 (35%) 10 (22%) ns

Weaning Outcomes ns Weaned 12 (52%) 25 (56%)

Vent Dependent 7 (30%) 14 (31%) Died 4 (17%) 6 (13%)

LOS, days 41 [25-134] 36 [10-223] ns


Encore Abstract: Previously presented at 2015 Neurocritical Care Society Annual Conference. Publication pending.

S51—A Survey of Current Practices Regarding Fasting Times in the Intubated, Neurocritically Ill Patient Stephanie D. Dobak, MS, RD, LDN, CNSC1; M. Kamran Athar, MD2 1Nutrition and Dietetics, Thomas Jefferson University Hospital, Philadelphia, PA, USA; 2Neurosurgery, Thomas Jefferson University, Philadelphia, PA, USA Purpose: Currently, pre-operative fasting guidelines exist for healthy patients undergoing elective procedures. Due to the lack of guidelines and related clinical trials regarding fasting times within the critically ill population, many institutions order enteral nutrition (EN) to be held at midnight for procedures/surgery or possible tracheal extubation. This may result in unnecessarily long periods without nutrition. Suboptimal EN is associated with increased septic complications, increased days on mechanical ventilation and in hospital, and a trend toward increased mortality. This survey sought to evaluate current practices among neurocritical care practitioners regarding pre-operative and pre-extubation fasting times in the intubated patient. Methods: A nine question survey was distributed through email (1296 members of Neurocritical Care Society) and available for participation October 19-31, 2014. Descriptive statistical analyses were utilized to review data. Results: Out of 102 respondents, 71 (70%) order EN to be held at midnight pre-operatively; 64 (63%) order EN to be held 4-6 hours prior to extubation; 45 (44%) do not order EN to be suctioned out of the gastrically fed patient prior to operation or extubation. Feeding tube bore size and location do not play a significant role in determining fasting times. Decisions dictating current fasting times are strongly influenced by institutional practice guidelines (73%) and concern for aspiration risk (58%). Conclusions: Current practices regarding fasting times in the intubated, neurocritical care population vary widely. The majority of practitioners in this setting order EN to be held at midnight prior to surgery and 4-6 hours prior to extubation. Frequent and prolonged cessations of EN result in inadequate calorie and protein provision, which in turn may lead to poorer clinical outcomes. We recommend clinical trials be undertaken to assess the safety of shorter fasting times for this patient population. In addition, practice guidelines are needed to define fasting times within the critically ill, intubated population. Financial Support received from: None

S52—Predictive Equations Unreliable in Determining Resting Energy Expenditure in ICH Patients Undergoing Therapeutic Hypothermia Stephanie D. Dobak, MS, RD, LDN, CNSC1; Fred Rincon, MD, MSc, FCCM, FNCS2 1Nutrition & Dietetics, Thomas Jefferson University Hospital, Philadelphia, PA, USA; 2Neurosurgery, Thomas Jefferson University, Philadelphia, PA, USA Purpose. Therapeutic moderate hypothermia (MH, Tcore33-34°C) is being studied as a promising treatment option for intracranial hemorrhage (ICH) patients. Optimal nutrition support is associated with improved outcomes, and this begins with an accurate determination of energy needs. Conventional validated equations are typically used to estimate basal metabolic rate (BMR) in stroke patients, but these may be inaccurate during MH. The primary aim of this study was to test the hypothesis that BMR as calculated by predictive equations would overestimate resting energy expenditure (REE) as measured by indirect calorimetry during MH after ICH. Methods. During a Phase-IIa/b clinical trial of temperature modulation after ICH (TTM-ICH, clinicaltrials.gov: NCT01607151), we measured BMR using the Harris-Benedict (BMR-HB) and Penn-State/modified Penn State equation (BMR-PS); and REE with the Vmax-Encore VS-29n System. We performed two measurements of BMR and REE, one at baseline (Day [D] 0) and the second between D1-3. We tested our hypothesis using a double-sided matched paired t-test for means. Results. In total, 5 ICH patients participated in this study, 2 undergoing MH and 3 normothermia (NT, Tcore 36-37°C). Median age was 62 years (IQR 26), 80% women. Overall, the mean BMR-HB at D0 was 1316±131kCal and at D1-3 was 1341±139Cal (p=0.2); mean BMR-PS at D0 was 1495±122kCal and at D1-3 was 1226±432kCal (p=0.5); and the mean REE at D0 was 1184±271kCal and at D1-3 1160±321kCal (p=0.9). In the MH group, the BMR-HB was increased from D0 to D1-3 (1331kCal vs. 1394kCal, Delta=+63kCal, p=0.01); the BMR-PS was decreased from D0 to D1-3 (1511kCal vs. 1040kCal, Delta=-471kCal, p=0.3); and REE was decreased from D0 to D1-3 (1413kCal vs. 985kCal, Delta=-428kCal, p=0.2). Within the MH group at D0, mean REE was 1413±197kCal


compared to mean BMR-PS 1511±41kCal and BMR-HB 1331kcal±165. At D1-3, mean REE was 985±4kCal compared to mean BMR-PS 1040±407kCal and BMR-HB 1394kcal±166. Conclusions. Our results suggest that in ICH patients undergoing therapeutic hypothermia, the use of predictive equations to determine BMR is suboptimal. The Harris Benedict and Penn State equations are more likely to overestimate energy requirements when compared to indirect calorimetry. MH is associated with decreased REE. Given the small sample size of this study, our results did not reach significance. We recommend larger studies to confirm these results. At this time, we support the use of indirect calorimetry to determine energy needs for patients undergoing MH. Financial Support: American Heart Association (12CRP12050342) and Academy of Nutrition and Dietetics Research Grant (2014-DNS). . S53—Prevalence of Sarcopenia and Overweight or Obesity in Cancer Patients Tania P. Tostado-Madrid, BS1; Yanin Torres-Tanus, AS1; Aurora Serralde-Zúñiga, PhD and MD2; Natalia Vazquez-Manjarrez, BS3; Ana Lara-Pulido, BS1; Angelica G. Esparza-Gómez3, AS3; Martha Guevara-Cruz, PhD and MD3 1Nutricion, Medica Sur, D.F., Mexico; 2Nutriologia Clinica, Instituto Nacional de Ciencias Medicas y Nutricion, D.F., Mexico; 3Fisiologia de la Nutricion, Instituto Nacional de Ciencias Medicas y Nutricion, D.F., Mexico Purpose: Nutritional status may be affected by the patient’s pathologies and comorbilities; cancer is one of the main diseases that cause impairment and impacts on quality of life, increase complications and toxicity of chemotherapy, and worse outcome, decreasing chances for survival. Besides loss of muscle tissue, sarcopenia includes loss of strength and functionality. On the other hand, obesity is increasing notoriously in treated cancer patients who in turn may suffer sarcopenia, resulting in an entity called sarcopenic obesity (coexistence of high fat mass with low muscle mass, carrying reduction of physical condition). Objective: Determine the prevalence of sarcopenia and obesity in cancer patients of a private Methods: Descriptive observational study carried out in the Oncology Center of a private hospital. All the ambulatory adult cancer patients were assessed during three months period. Data was gathered from their medical records and biochemical analysis, and patients were assessed with a nutritional history, a food frequency questionnaire, anthropometric measurements and an electrical bioimpedance scale (Inbody R20) to obtain fat and muscle mass, and a 6 m walk test (6MWT) and a dynamometer (Takei Grip Strength Dynamometer) to estimate running speed and muscle strength. This information was used to diagnose the patients with overweight or obesity according to the Body Mass Index (BMI), sarcopenia according to the Foundation of NIH sarcopenia Project criteria and sarcopenic obesity according to fat mass. Between the data collected form the medical records included the stage and type of cancer, differentiation, tumoral aggressiveness, type of treatment, and weight loss in the last month and in the last six months. The average of continuous variables, percent or frequency of nominal and dichotomous variables, and standard deviation were reported. The statistical analysis was made using the t-test for independent samples. Results: 120 cancer patients were studied, 68 women and 52 men aged 56.1 ± 14 years. (Figure 1). Diagnosis categorization was as follows: breast cancer thirty patients, prostate cancer 6 patients, lung cancer 6 patients, gastrointestinal tumor 21, pancreas 10, hematologic neoplasms 19, gynecological 14 and 13 other types of cancer. Anthropometric and biochemical features are shown in Table 1.Thirty percent (36) of patients assed were diagnosed with sarcopenia, sarcopenic obesity patients reached 47% (17) highlighting a 35% (6) with severe sarcopenia. (Figure 2). Comparing sarcopenic and not sarcopenic patients, the first showed a lower BMI, hip and arm circumference. (p < 0.05). Furthermore, patients with sarcopenic obesity in comparison with patients without this diagnostic, had a greater BMI, hip and arm circumference, and those with severe sarcopenia were older and had a higher hip circumference and cholesterol HDL (p < 0.05). In addition 60% of sarcopenic patients presented overweight or obesity. On the other hand, there was an important increase of BMI in the last six months of 10.9% in patients with sarcopenia, and in patients without sarcopenia of 7%. (Figure 3) Conclusions: 30% of the studied patients showed sarcopenia. From this 30% almost half had sarcopenic obesity, a third part suffered severe sarcopenia and 60% of patients had overweight and obesity. Therefore prevention and early detection of these diagnoses may play an important role in nutritional treatment and in cancer and medical treatment consequential complications. Moreover, this study exposes the importance of a detailed nutritional approach made by a professional focused on nutrition and oncology, who consider a complete diagnostic for a successful nutritional treatment.



Table 1. Anthropometric and Biochemical Characteristics of Patients.

Characteristics All Mean ± SD

Without Sarcopenia

With Sarcopenia P

Age, years 56.0 ± 14.7 54.6± 15.1 59.5±12.3 0.089 Diagnostic Time, months 30.0 ± 5.4 27.4±5.34 36.5±13.4 0.451

Weight, kg 67.6 ± 15.4 63.7 ± 14.4 76.7±13.9 0.0001 BMI, kg/m2 25.0 ± 4.86 24.4±4.75 26.5±4.87 0.031

Waist circumference, cm 91.5 ± 12.5 88.6±11.7 98.3±11.7 0.0001 Hip circumference, cm 100 ± 13.0 100±10.2 100±18.1 0.776 Arm circumference, cm 28.0 ± 6.24 26.9±3.46 30.4±9.75 0.005

% masa grasa 30.8 ± 9.10 31.8±9.07 28.8±8.94 0.099 Hemoglobin 12.8 ± 1.94 12.6±1.85 13.3±2.13 0.125

mean corpuscular volume 90.0 ± 10.8 90.2±12.0 89.4±6.84 0.752 Platelets 230027 ± 112397 236±118 213±92.1 0.365

leukocytes 7.19 ± 0.85 6.98±0.90 7.77±2.06 0.687 Neutrophils, % 60 ± 13.0 60±11.8 61.3±16.2 0.755 Glucose, mg/dl 107 ± 28.1 104±27.3 115±29.1 0.113

Blood urea nitrogen, mg/dl 14.9 ± 2.6 11.9±4.62 23.1±9.5 0.057 Creatinine, mg/dl 0.76 ± 0.22 0.73±0.24 0.82±0.16 0.126

Sodium, mg/dl 138 ± 2.58 138±2.66 139±2.19 0.068 Potassium, mg/dl 3.89 ± 0.40 3.85±0.42 3.98±0.36 0.249 Chlorine, mg/dl 106 ± 3.77 106±3.86 106±3.65 0.887 Calcium, mg/dl 9.33 ± 0.38 9.32±0.37 9.33±0.40 0.943

Phosphorus, mg/dl 3.75 ± 0.69 3.72±0.74 3.81±0.59 0.655 Triglycerides, mg/dl 164 ± 83 171±86.1 150±77.3 0.368 Total protein, g/dl 6.49± 0.69 6.40±0.74 6.70±0.50 0.100

Albumin, g/dl 3.74± 0.51 3.70±0.58 3.84±0.29 0.314 The statistical analysis was made using the t-test for independent samples.


Figure 1. Prevalence of sarcopenia and type of cancer.

Figure 2. Prevalence of patients.


Figure 3. Body mass index and presence of sarcopenia before and after six months.

S54—Association of Sub-Optimal Serum Phosphorus Levels and Anemia in Critically Ill Patients Tiffany Otero, BS1,5; Cecilia Canales, MPH1,7; D. Dante Yeh, MD2,6; Emily Johnson, RN4; Tara Harrison, RN4; Donna Belcher, RD, MS3; Sadeq A. Quraishi, MD, MHA, MMSc1,6 1Anesthesia, Critical Care and Pain Medicine, Massachusetts General Hospital, Boston, MA, USA; 2Department of Surgery, Massachusetts General Hospital, Boston, MA, USA; 3Department of Nutrition and Food Services, Massachusetts General Hospital, Boston, MA, USA; 4Department of Nursing, Massachusetts General Hospital, Boston, MA, USA; 5School of Medicine, Tufts University, Boston, MA, USA; 7School of Medicine, University of California -- Irvine, Irvine, CA, USA; 6School of Medicine, Harvard University, Boston, MA Purpose: Recent studies have suggested that deranged serum phosphate (Phos) levels are associated with anemia in hospitalized patients. It remains unclear whether such a relationship exists at the outset of critical illness. Therefore, our goal was to investigate whether abnormal Phos levels on admission to the intensive care unit (ICU) are associated with anemia in critically ill patients. Methods: We performed a retrospective analysis of data from an ongoing prospective study of nutrition in critical illness. Serum Phos and hemoglobin (Hgb) levels were obtained within one hour of ICU admission. Normal Phos level was defined as 2.5-4 mg/dL. Anemia was defined according to age, sex, and race-specific National Health and Nutrition Examination III thresholds for Hgb. To investigate the association between Phos and Hgb levels, we performed a linear regression analysis, while controlling for age, sex, race, body mass index (BMI), Nutrition Risk Screening (NRS 2002) score, Deyo-Charlson Comorbidity Index, as well as levels of creatinine, mean corpuscular volume (MCV), and albumin. Similarly, to investigate the association of Phos levels with anemia, we performed logistic regression analyses, while controlling for the same covariates. Results: 510 patients comprised the analytic cohort. Baseline characteristics are presented in Table 1. On admission, 315 (62%) patients were anemic, while 154 (30%) had Phos levels >4mg/dL and 69 (14%) had levels <2.5 mg/dL. Linear regression analysis demonstrated an inverse association of Phos and Hbg levels (β = -0.19: 95%CI -0.34 to -0.04; p = 0.016). In this model, additional variables found to be independently associated with Hgb levels were female sex (inverse relationship), MCV, and albumin levels. Logistic regression analysis demonstrated a 25% higher risk of anemia per 1 mg/dL change in Phos levels (OR = 1.25: 95%CI 1.05-1.50; p = 0.018). Moreover, patients with Phos levels >4 mg/dL had a 68% higher risk of anemia compared to patients with normal Phos levels (OR 1.68: 95%CI 1.02-2.78; p = 0.04). On the other hand, patients with Phos levels <2.5 mg/dL were not at a higher risk of anemia compared to patients with normal Phos levels (OR = 1.09: 95%CI 0.57-2.08; p = 0.79). In all logistic


regression analyses, additional variables found to be independently associated with anemia were female sex, BMI, and albumin levels. Conclusions: Our data suggest that there is an overall inverse association between Phos and Hgb levels at the outset of critical illness. ICU patients with Phos levels >4 mg/dL on admission are significantly more likely to be anemic compared to patients with normal Phos levels. These findings support the need for future studies to determine whether maintaining optimal Phos levels in ambulatory care patients reduces the risk of anemia at the outset of acute care hospitalization or critical illness. Future studies should also aim to determine whether maintaining optimal Phos levels during acute care hospitalization or critical illness influences clinical outcomes, such as the need for blood transfusions, length of stay, acute organ failure, and mortality. Financial Support received from: None


S55—Elevated Red Cell Distribution Width at Initiation of Critical Care Is Associated With Mortality in Surgical Intensive Care Unit Patients Tiffany Otero, BS1,5; Caitlin M. McCarthy, BA1,5; D. Dante Yeh, MD2,6; Cecilia Canales, MPH7; Donna Belcher, RD, MS3; Kenneth B. Christopher, MD4; Sadeq A. Quraishi, MD, MHA, MMSc1,6 1Anesthesia, Critical Care and Pain Medicine, Massachusetts General Hospital, Boston, MA, USA; 2Department of Surgery, Massachusetts General Hospital, Boston, MA, USA; 3Department of Nutrition and Food Services, Massachusetts General Hospital, Boston, MA, USA; 4Department of Medicine, Brigham and Women's Hospital, Boston, MA, USA; 5School of Medicine, Tufts University, Boston, MA, USA; 6School of Medicine, Harvard University, Boston, MA, USA; 7School of Medicine, University of California: Irvine, Irvine, CA Purpose: Elevated red cell distribution width (RDW), often thought of as a marker of systemic inflammation, is associated with mortality in various chronic diseases. Recent data suggests that in critical illness, elevated RDW is also associated with subsequent all-cause mortality. However, it is not clear whether early derangements of RDW levels in surgical intensive care unit (ICU) patients, who often have limited underlying chronic diseases compared to medical ICU patients, may have biological significance. Therefore, our goal was to investigate whether RDW levels, measured at initiation of critical care, are associated with mortality in surgical ICU patients. Methods: We performed a retrospective analysis of data from a prospective, observational cohort study of nutrition in critical illness. Only patients admitted to surgical ICUs (i.e. acute changes in health not related to the exacerbation of a chronic illness) were included in the study. RDW levels were obtained as part of a complete blood count assessment within one hour of ICU admission in all patients. Normal RDW level was defined as 11.5-14.5%. To investigate the association of RDW and 90-day mortality, we performed a logistic regression analysis, controlling for age, sex, race, body mass index (BMI), Nutrition Risk Screening (NRS 2002) score, Acute Physiology and Chronic Health Evaluation (APACHE II) score, hospital length of stay, as well as admission creatinine and albumin levels. In a subset of patients, who had c-reactive protein (CRP) levels assessed on ICU admission, we repeated the analysis with the addition of CRP as a covariate in the regression model. Results: 500 patients comprised the analytic cohort. Baseline characteristics are presented in Table 1. At the time of ICU admission, 237 (47%) patients had elevated RDW levels and subsequent 90-day mortality in the overall cohort was 28%. Logistic regression analysis demonstrated that patients with elevated RDW levels had a greater than two-fold higher risk of mortality at 90-days (OR 2.28: 95%CI 1.20-4.33; p = 0.004) compared to patients with normal RDW levels. 300 patients had admission CRP levels measured (mean 134 ± 92 mg/L). In this subset, 149 (50%) patients had elevated RDW levels and mortality was 32%. Logistic regression analysis, using the same variables as those in the overall study cohort regression model, but with the addition of CRP levels, demonstrated an almost 20-fold greater risk of mortality in patients with elevated RDW levels (OR 19.65: 95%CI 1.58-244.21; p=0.021) compared to patients with normal RDW levels. Conclusions: In our cohort of surgical ICU patients, elevated RDW levels at initiation of critical care was significantly associated with increased risk of 90-day mortality. Moreover, our data suggest that this relationship exists independent of the systemic inflammation that typically accompanies critical illness. These data support the need for future studies to determine the biological processes that lead to elevated RDW levels in early critical illness and the physiologic consequences of such derangements. Additionally, future studies are needed to identify effective interventions to normalize RDW levels and to determine whether such interventions can improve survival after critical illness. Financial Support received from: None


S56—Sliding Scale Regular Human Insulin for Critically Ill Patients Receiving Nutrition Support Sarah V. Cogle, PharmD1; Susan E. Dickey, PharmD2; George O. Maish, MD3; Gayle Minard, MD3; Martin A. Croce, MD3; Roland N. Dickerson, PharmD4 1Pharmacy Practice, Auburn University Harrison School of Pharmacy, Auburn, AL, USA; 2Pharmaceutical Services, Vanderbilt University Medical Center, Nashville, TN, USA; 3Surgery, University of Tennessee College of Medicine, Memphis, TN, USA; 4Clinical Pharmacy, University of Tennessee College of Pharmacy, Memphis, TN, USA Purpose: Blood glucose (BG) concentrations below 140 to 150 mg/dl are associated with improved clinical outcomes in critically ill trauma patients. Intermittent intravenous (IV) regular human insulin (RHI), known as sliding scale RHI therapy (SSI), is frequently employed when initiation of a continuous IV RHI infusion is not indicated. The intent of this study was to evaluate the safety and efficacy of our standardized intermittent sliding scale RHI algorithms in critically ill trauma patients receiving specialized nutrition support. Methods: Adult patients (> 17 y of age), admitted to the trauma intensive care unit between January 2014 and April 2015 and referred to the Nutrition Support Service (NSS) for enteral nutrition (EN) or parenteral nutrition (PN) were evaluated. Those who received a continuous IV RHI infusion, non-NSS directed SSI, or those who received < 3 days of SSI were excluded. Patients without a history of diabetes mellitus (DM) and whose baseline BG prior to nutrition therapy was < 150 mg/dL were given our low intensity SSI (Table 1). Patients with DM or with moderate stress-induced hyperglycemia (BG 150 to 179 mg/dL) prior to nutrition therapy were assigned our high intensity SSI


(Table 1). BG determinations were obtained every 3 h to 6 h per discretion of NSS personnel. Patients with DM or stress-induced hyperglycemia were given a low carbohydrate formula unless a different specialized EN formula was indicated. Those who failed glycemic control were evaluated by the NSS for escalation to the high intensity SSI or a continuous IV RHI infusion. SSI was discontinued if BG was maintained in the target BG range (70 mg/dL to 149 mg/dL) at goal nutrition intake with minimal RHI intake (about < 8 units/d). Patients were followed up to a maximum of 7 d while receiving concurrent SSI and nutrition therapy. Efficacy of the SSI algorithms were evaluated by the number of hours spent within the target BG range or with respect to treatment failure as evidenced by escalation to more aggressive RHI therapy. Safety endpoints were measured by the number of patients who experienced at least one episode of mild hypoglycemia (BG 40-69 mg/dL) or severe hypoglycemia (BG < 40 mg/dL). Non-nominal data were expressed as mean ± SD. Results: A total of 121 cases of patients receiving SSI were enrolled for study. Forty-two patients received the higher intensity SSI and 79 were given the lower intensity SSI therapy. A significantly greater proportion of patients who received the higher intensity SSI were infected, had DM, were older and had a higher BG at baseline prior to initiation of NS (Table 2). Both groups received similar carbohydrate intakes (Figure 1). Despite a greater daily BG and greater RHI intake (Figure 1), 37 out of 42 patients (88%) who received the higher intensity SSI therapy were considered to have effective glycemic control without need for an escalation in insulin therapy. Fifty-nine of 79 patients (75%) who received the lower intensity SSI therapy were considered to have effective glycemic control. Three patients (7%) in the higher intensity group experienced at least one episode of mild hypoglycemia and two patients (5%) experienced an episode of severe hypoglycemia (BG < 40 mg/dL). Nine patients (11%) in the lower intensity group experienced at least one episode of mild hypoglycemia and none experienced severe hypoglycemia. Other patient characteristics, clinical and glycemic outcomes for both groups are detailed in Table 2. Conclusions: Our intermittent RHI sliding scale algorithms safely and effectively controlled BG for the majority of critically ill patients with traumatic injuries receiving EN or PN when a continuous IV RHI infusion was not indicated. Financial Support received from: N/A

Table 1. Intravenous RHI Sliding Scale Algorithms* BG mg/dL Higher Intensity SSI Lower Intensity SSI

Give 25 g D50W IV; call MD Give 25 g D50W IV; call MD 40-69 Give 12.5 g D50W IV; call MD Give 12.5 g D50W IV; call MD

70-125 0 units RHI 0 units RHI 126-150 3 units RHI 2 units RHI 151-175 6 units RHI 4 units RHI 176-200 9 units RHI 6 units RHI 201-225 12 units RHI 8 units RHI 226-250 15 units RHI 10 units RHI 251-275 18 units RHI 12 units RHI 276-300 21 units RHI 14 units RHI

>300 24 units RHI; call MD 16 units RHI; call MD BG, blood glucose concentration; D50W, 50% dextrose in water; IV, intravenously; MD, physician; RHI, regular human insulin; *BG sampling frequency ranges from every 3 h to 6 h.


Table 2. Patient Characteristics and Outcomes

Variable Higher Intensity SSI (n=42) Lower Intensity SSI (n=79) P Sex, Male/Female, n/n 30/12 63/16 0.420

Race African American, n

Caucasian, n Hispanic/Other, n

10 26 6

27 39 13

0.584

Age, y 54 ± 18 45 ± 18 0.016 Injury Severity Score 26 ± 10 30 ± 14 0.231

Traumatic brain injury, n 31 (74%) 43 (54%) 0.059 Concurrent infection, n 20 (48%) 14 (18%) 0.001

BMI, kg/m2 29.1 ± 5.6 28.5 ± 6.9 0.611 Weight, kg 89 ± 19 87 ± 23 0.559

Diabetes mellitus, n 10 (24%) 5 (6%) 0.013 WBC, cells/mm3 13.5 ± 7 13.5 ± 14.5 0.998

C-reactive protein, mg/dL 25.5 ± 10 25.4 ± 11.4 0.984 Prealbumin, mg/dL 8.9 ± 3.8 8.7 ± 4.7 0.859

Albumin, g/dL 2.5 ± 0.3 2.5 ± 0.5 0.684 EN/PN/Both, n/n/n 35/1/6 68/5/6 0.346 Day NS started, d 2.6 + 1.5 2.8 + 2.4 0.560

Mortality, n 9 (21%) 12 (15%) 0.522

Hospital length of stay, d 33 ± 20 30 ± 17 0.410 ICU length of stay, d 21 ± 15 18 ± 10 0.307

Ventilator days, d 19 + 19 14 + 10 0.069

CHO intake, g/d 137 + 73 144 + 145 0.435 RHI intake, units/d 16 + 10 4 + 4 0.001

Baseline BG, mg/dL 164 + 20 123 + 33 0.001 BG during SSI, mg/dL 145 + 21 129 + 14 0.001 BG 70-149 mg/dL, h/d 16 + 5 20 + 4 0.001 BG > 150 mg/dL, h/d 8 + 5 4 + 4 0.001

BG 40 to 69 mg/dL, h/d 0.10 + 0.48 0.08 + 0.27 0.400 BG 0.04 + 0.21 0 + 0 0.109

BMI, body mass index; CHO, carbohydrate; EN, enteral nutrition; ICU, intensive care unit; NS, nutrition support; PN, parenteral nutrition, SSI, sliding scale insulin therapy; WBC, white blood cell


Figure 1. Blood glucose response to intermittent sliding scale RHI for patients who received higher intensity vs. lower intensity therapy.


S57—Enteral Nutrition Delivery Is Overestimated in Provider Documentation Jenna E. Kesey, MSN, RN, FNP-BC, CWS1,2; Amy Ashmore, MS, RDN, LD1; Breanne Durbin, MS, RDN, LD1; Emily Garcia, MSN, RN, CCRN1; Sharmila Dissanaike, MD2 1Timothy J Harnar Burn Center, UMC Health System, Lubbock, TX, USA; 2Department of Surgery, Texas Tech University Health Sciences Center, Lubbock, TX, USA Purpose: Clinicians struggle with effectively meeting the estimated energy needs of critically ill patients. Research suggests early enteral nutrition and reducing breaks in nutrition delivery is beneficial to the overall health of the patient. Literature indicates large cumulative caloric deficits have been detrimental in patient clinical outcomes; increasing ICU length of stay, ventilator days, incidence of pressure ulcers, risk of infections and overall mortality. In 1982, Bartlett and associates correlated a caloric deficit of greater than 10,000 calories with death in 76% of patients. Barriers to meeting the estimated energy needs of critically ill patients may include: late initiation of the tube feed, failure to have a registered dietitian consultation, frequent unnecessary cessation of tube feeds and a lack of understanding of nutrition in the critically ill. Many critical care units have challenges in achieving prescribed rates of enteral nutrition. Our institution has developed an aggressive feeding algorithm aiming to meet the high demand of adult critical care patients. During patient rounds, it has been observed that enteral nutrition is frequently held or stopped for various reasons. We evaluated whether the documented volume of enteral nutrition delivered was accurate when compared with actual volume delivered, as recorded by the feeding pump. Methods: All patients admitted to critical care units from June to August 2015 on enteral nutrition support were included. Data was collected for the duration that they required enteral nutrition. Dietitian recommendations, provider orders and nursing documentation were abstracted. The total volume infused, as indicated on the pump, was recorded and cleared every 24 hours; this volume was compared to the documentation from the medical record. Variables were compared using t-test in Excel. Results: Fifty four patients were included, providing a total of 169 patient-days. Patients were admitted to one of 3 adult critical care units: Medical Intensive Care Unit (MICU; n=14), Trauma/Surgical Intensive Care Unit (TSICU; n=15), and the Burn Center (BC; n=25). The average volume per day of enteral nutrition that was delivered to the patient was 1268 ml (range 44-2804 ml). The average volume documented was higher at 1384 ml (range 60-3120 ml); indicating a significant overestimation (p value =0.04). The average enteral nutrition ordered was 1558 ml (range 30-3120 ml), which was significantly higher than both amount delivered and documented (p < 0.001 and p= 0.006 respectively). Overall, 18% of ordered enteral nutrition was not given, while the actual volume delivered was 8% lower than documented in the medical record. The standard enteral nutrition formula used at this facility provides 1.5 kilocalories per milliliter; therefor this discrepancy results in an average deficit of 435 calories per day to each patient. Conclusions: There was significant discrepancy between ordered, recorded and actual delivered volume of enteral nutrition. Potential reasons for the discrepancy may be frequent interruption for nursing procedures such as repositioning, wound care, linen changes, changing bottles of tube feeding or changing tubing. Staff may underestimate the time the feeding pump is off for these small periods of time resulting in significant cessations in nutrition delivery. More research is needed regarding the disparity of actual nutrition delivered and charted volumes. Financial Support received from: none

S58—Somatostatin Analog Administration as a Possible Contraindication to Jejunal Tube Feeding After Pancreatic Resection Leah Hoffman, PhD, RD, LD, CNSC1; Allen W. Knehans, PhD1; Russell G. Postier, MD2 1Nutritional Sciences, University of Oklahoma Health Sciences Center, Oklahoma City, OK, USA; 2Surgery, University of Oklahoma Health Sciences Center, Oklahoma City, OK, USA Purpose: Within a larger prospective randomized controlled trial investigating the effects of tube feeding following pancreatic surgery, it was unexpectedly difficult obtain goal tube feeding rate due to gastrointestinal side effects. The only identified inconsistency in practice is that 1 of the 2 surgeons routinely administered the somatostatin analog octreotide; the other does not use this medication. Octreotide is known to decrease transit time in the proximal small bowel and reduce splanchnic blood flow, but its relationship to jejunal tube feeding has not been well investigated. A preliminary study was conducted to compare the provision of tube feeding at goal rate and the occurrence of gastrointestinal side effects between the participants. Methods: Participants who had been randomized to tube feeding in the larger trial were separated into 2 groups by


surgeon. These 2 groups were compared in amount of prescribed tube feeding actually received by the participant and in occurrence of gastrointestinal side effects (nausea and/or vomiting, delayed gastric emptying, ileus, diarrhea, constipation, abdominal distension, and mesenteric ischemia). Results: 11 participants had received tube feeding at this point in the trial. 5 had never received octreotide; 6 had received at least 1 dose. Participants who never received octreotide were more likely to reach their goal rate and to receive more of the prescribed dose of tube feeding. Of those who did not receive octreotide, 3 of 5 had no gastrointestinal side effects. Of those who did, all experienced at least 1 gastrointestinal side effect. Conclusions: Octreotide administration may be related to reduced tolerance of jejunal tube feeding after pancreatic surgery. Financial Support received from: None

S59—Energy and Protein Deficits Persist Throughout Hospitalization in Patients Admitted With a Traumatic Brain Injury Lee-anne S. Chapple, MND, BMSc1,2; Adam Deane, MBBS, PhD, FRACP, FCICM2,3; Lauren T. Williams, PhD, AdvAPD, PGradDipHlthSc, GradDipSocSc, GradDipNutDiet4; Kylie Lange, B.Sc. (Maths&ComputerSci) (Hons)5; Daren Heyland, MD, FRCPC, MSc6; Marianne J. Chapman, BMBS, PhD, FFARCSI, FANZCA, FJFICM2,3 1Discipline of Acute Care Medicine, University of Adelaide, Adelaide, Australia; 2National Health and Medical Research Council of Australia Centre for Clinical Research Excellence in Nutritional Physiology and Outcomes, University of Adelaide, Adelaide, Australia; 3Intensive Care Unit, Royal Adelaide Hospital, Adelaide, Australia; 4Menzies Health Institute of Queensland, Griffith University, Gold Coast, Australia; 5Discipline of Medicine, University of Adelaide, Adelaide, Australia; 6Clinical Evaluation Research Unit, Kingston General Hospital, Kingston, Canada Purpose: Following traumatic brain injury (TBI) patients experience considerable energy and protein deficits during their Intensive Care Unit (ICU) admission and these deficits are associated with adverse outcomes. While energy and protein delivery in ICU has been the subject of research, nutrition provision after ICU discharge, particularly once oral diet has commenced, has not been previously quantified. It was hypothesized that these nutritional deficits persist throughout hospitalization. Methods: All adult patients with a moderate-severe TBI (Glasgow Coma Scale score 3-12) admitted to ICU of the major neuro-trauma referral center in the state of South Australia over a 12-month period were eligible. Patients were excluded if they were expected to die imminently. Data on energy and protein delivered from all routes (including medications) were collected prospectively over the hospital admission and compared to Dietician prescriptions. To precisely quantify oral intake weighed food records were conducted on three pre-specified days per week. All food and fluids provided from breakfast to one hour post-dinner were weighed and data extrapolated to a weekly average intake. Individual meal components were weighed prior to delivery to the patient, and then after consumption to measure waste, with energy and protein intakes calculated using dietary analysis software. Data were censored 90 days after admission to hospital. Data are presented as mean (SD) unless otherwise indicated. Results: Forty-seven of 105 patients admitted with a TBI were eligible for the study. Consent was obtained for 37 patients, and these patients contributed 1512 study days. Mean age was 45.3 (15.8) years, 87% were male, 67% of TBIs were severe (GCS 3-8), and body mass index was 26.7 (6.4) kg/m2. Mechanical ventilation was required in 95% of patients. The median APACHE II and Trauma Injury Severity Scores (TRISS) were 18 and 0.63, respectively. Median length of stay was 13.4 [IQR: 6.5-17.9] days in ICU and 19.9 [IQR: 9.6-32.0] days of post-ICU ward-based care, for a total of 37.6 [IQR: 19.4-52.4] days in hospital. At day 90 all patients were alive with 11% remaining in hospital. Over the entire hospitalization patients had a mean cumulative deficit of 18,242 (16,642) kcal energy and 1,315 (1,028) g protein. Overall, mean daily intakes were 1916 (880) kcal energy and 86 (43) g protein, equivalent to 83 (36) % prescribed energy and 75 (37) % protein requirements. Mean energy and protein intakes were less in ICU than on post-ICU wards (1798 (800) vs 1980 (915) kcal energy p=0.015 and 79 (47) vs 89 (41) g protein p=0.001, respectively). When compared to Dietician prescriptions, which increased once sedation was ceased, patients met 88% energy and 65% protein requirements in the first two weeks in ICU compared to 69% energy and 62% protein requirements in the first two weeks after transfer to the acute ward. Overall, energy deficits were greater in patients ingesting nutrients orally only compared to exclusive enteral nutrition (3367 (2575) vs 1858 (2371) kcal energy, p=0.016) while protein deficit was similar (40.3 (4.9) vs 36.6 (6.2) g protein, p=0.616). Conclusions: Patients admitted to ICU with a moderate-severe TBI remain in hospital for extended periods. Not only do calorie and protein deficits occur in ICU but these persist after discharge to the ward, leading to


considerable shortfalls over the entire hospitalization. Patients consuming nutrients orally are particularly at risk of caloric deficit, which may have implications for patient outcomes. Financial Support received from: Neurosurgical Research Foundation, Australia Australasian Society of Parenteral and Enteral Nutrition

ENCORE ABSTRACT: Previously presented at the Australian and New Zealand Intensive Care Society. Not printed here at the request of the author.

S60—The Effect of Early Energy Delivery on Quality of Life and Employment Status One Year After ICU Admission: A Randomized Controlled Clinical Trial Lee-anne S. Chapple, MND, BMSc1; David Reid, BDS, MBChB2; Stephanie O'Connor, RN, Grad. Dip. Card. Nsg., MNSc1,2; Rinaldo Bellomo, MBBS MD FRACP FCICM PGDipEcho3; Heidi Buhr, RN, B.Cur (Ed&Adm): Crit Care, M.Clin.T.Prac4; Marianne J. Chapman, BMBS, PhD, FFARCSI, FANZCA, FJFICM1,2; Andrew Davies, MBBS, FRACP5; Glenn Eastwood, RN, PhD3; Suzie Ferrie, MNutDiet, BA (Philosophy) (Hons),4; Kylie Lange, BSc (Maths&ComputerSci) (Hons)9; Joanne McIntyre, RN, Crit Care6; Dale Needham, FCA, MD, PhD7; Sandra Peake, BM BS, BSc (Hons), FCICM, PhD1,6; Sumeet Rai, MBBS, DA (Hons), FCPS, DNB, FCICM, EDIC, PGCCU8; Emma Ridley, BNutrDietet MPH APD5; Helen Rodgers, B App Sci (Med Lab Sci), PhD8; Adam Deane, MBBS, PhD, FRACP, FCICM1,2 1Discipline of Acute Care Medicine, University of Adelaide, Adelaide, Australia; 2Department of Intensive Care Medicine, Royal Adelaide Hospital, Adelaide, Australia; 3Department of Intensive Care, Austin Hospital, Melbourne, Australia; 4Intensive Care Service, Royal Prince Alfred Hospital, Sydney, Australia; 5Department of Epidemiology and Preventative Medicine, Australian and New Zealand Intensive Care Research Centre, Monash University, Melbourne, Australia; 6Department of Intensive Care, The Queen Elizabeth Hospital, Adelaide, Australia; 7Outcomes After Critical Illness and Surgery Group, John Hopkins University, Baltimore, MD, USA; 8Intensive Care, The Canberra Hospital, Canberra, Australia; 9Discipline of Medicine, University of Adelaide, Adelaide, Australia Financial support: None

S61—Are Weekends Different? Comparisons of Oral Intake in Hospitalized Patients Recovering From Critical Illness Lee-Anne S. Chapple, MND, BMSc1,2; Adam Deane, MBBS, PhD, FRACP, FCICM2,3; Lauren T. Williams, PhD, AdvAPD, PGradDipHlthSc, GradDipSocSc, GradDipNutDiet4; Kylie Lange, B.Sc. (Maths&ComputerSci) (Hons)5; Daren Heyland, MD, FRCPC, MSc6; Marianne J. Chapman, BMBS, PhD, FFARCSI, FANZCA, FJFICM2,3 1Discipline of Acute Care Medicine, University of Adelaide, Adelaide, Australia; 2National Health and Medical Research Council of Australia Centre for Clinical Research Excellence in Nutritional Physiology and Outcomes, University of Adelaide, Adelaide, Australia; 3Intensive Care Unit, Royal Adelaide Hospital, Adelaide, Australia; 4Menzies Health Institute of Queensland, Griffith University, Gold Coast, Australia; 5Discipline of Medicine, University of Adelaide, Adelaide, Australia; 6Clinical Evaluation Research Unit, Kingston General Hospital, Kingston, Canada Purpose: Studies of nutrition practices in a hospitalized setting frequently disregard oral intake because measurement is considered imprecise. While time-intensive, weighed food records provide precise measures of energy and protein consumption. Studies in free-living populations recommended these be conducted on at least three days per week, comprising two week days and one weekend day, to provide a reliable representation of actual intake. However whether this methodology is appropriate in hospitalized patients has not been determined and there are considerably logistical challenges with conducting weighed food records on weekends. The objectives were to compare energy and protein intakes from investigator-led weighed food records conducted on week and weekend days, and to evaluate whether variations in intake between days occurred over time. Because of fewer fasting periods for procedures and greater periods when family were present to assist with feeding it was hypothesized that patients recovering from a traumatic brain injury (TBI) would consume more energy and protein on weekends. Methods: All adult patients admitted with a moderate-severe TBI (Glasgow Coma Scale score 3-12) to the major neuro-trauma referral center in the state of South Australia over a 12-month period were eligible. Data on energy and protein consumed orally were collected prospectively with data censored at 90 days. To precisely quantify oral intake investigator-led weighed food records were conducted on three pre-determined days per week, including two


week days and one weekend day. Individual meal components provided from breakfast to one hour post-dinner were weighed prior to delivery to the patient using Salter SJ-HS digital scales and then after consumption to measure waste. Energy and protein intakes were then calculated with Foodworks dietary analysis software. Weighed food records were excluded if they were partial days (due to hospital discharge or day leave) or if patients also received enteral or intravenous feeding. Data are presented as mean (SD). Comparisons between energy and protein intakes on week versus weekend days were determined using mixed models with a fixed effect for day of week and random effects to account for clustering of multiple days within subjects. Results: Data were collected for 32 patients who were ingesting food. On admission mean age was 43.9 (16.1) years, 88% were male, body mass index was 26.1 (6.4) kg/m2, APACHE II score was 17.1 (4.9) and 63% of patients had a severe TBI (GCS 3-8). Mean length of stay was 12.6 (8.2) days in ICU and 38.9 (30.3) days in hospital. Patients commenced oral intake on day 18 (16.4) of hospital admission and received oral intake for an average of 21.4 (23.7) days. A total of 220 days of weighed food records, 68% (n=149) on a weekday and 32% (n=71) on a weekend day. Overall, patients consumed 1367 (877) kcal/day energy and 62 (40) g/day protein orally, with similar amounts consumed on week and weekend days (1368 (894) vs 1364 (845) kcal energy, 61 (41) vs 63 (39) g protein). There were no differences in energy or protein intakes received on any individual day (p=0.937 energy, p=0.797 protein) or between combined week versus weekend days (p=0.913 energy, p=0.567 protein). There were no differences in energy or protein intakes consumed in the first (p=0.665 energy, p=0.433 protein; n=83), second (p=0.529 energy, p=0.907 protein; n=56), or third (p=0.426 energy, p=0.110 protein; n=30) week of oral intake. Conclusions: In hospitalized patients recovering from a TBI energy and protein intakes are similar during week and weekend days. Future quantification of energy and protein intakes could therefore be completed using investigator-led weighed food records during week days only. Financial Support received from: Australasian Society of Parenteral and Enteral Nutrition Neurosurgical Research Foundation.

S62—Enteral Feeding Tolerance During Induced Hypothermia Post Cardiac Arrest Mahadevi Guggari, MS RD CNSC1; Allison France, BS2; Ben Coopwood, MD3; Jacklyn Barrera, MEd, CHES2; Carlos BVR Brown, MD3; Sadia Ali, MPH4 1Clinical Nutrition, Seton Medical Center, Austin, TX, USA; 2Coordinated program in Dietetics, University of Texas, Austin, TX, USA; 3Surgery and Perioperative Care, University of Texas Dell Medical School, Austin, TX, USA; 4Trauma Services, University Medical Center at Brackenridge, Austin, TX, USA Purpose: Cardiac arrest outside of the hospital has an average survival rate at just 10%. Those who survive are at risk for a neurologic injury. Therapeutic hypothermia is the only intervention shown to improve neurologic outcome following cardiac arrest. Nutrition support is an important element in the care of a critically ill patient. In critically ill, mechanically ventilated patients, early enteral nutrition (EN) has been correlated with reductions in the intensive care unit (ICU) and hospital mortality, and infectious complications. The beneficial effect from EN is time and disease-severity dependent. However, there is limited research and a large debate on when or whether to initiate nutrition support at all during hypothermia. The purpose of the study was to evaluate enteral feeding practice and tolerance in patients undergoing therapeutic hypothermia post cardiac arrest. Methods: This is a retrospective study consisting of patients in five surgical/ medical ICUs over a fourteen-month period. All patients aged 18 years and older who received and completed therapeutic hypothermia were included in the study. Patients were excluded for the following criteria: patients under 18 years of age, patients under palliative care, patients that failed to complete the hypothermia protocol and patients that passed away during HT. Data collected included ICU admission diagnosis, age, gender, admission weight, admission height, time and date of initiation and end of HT, date and time of hospital and ICU admission and mortality at the end of hospital stay. Data on estimated energy needs and calorie and protein intake from enteral feeding during cooling (Phase I) and rewarming (Phase II) phases of hypothermia were recorded. All instances of feeding intolerances and reasons for EN interruption such as route of nutrition, amount of residuals, presence of diarrhea, ileus and abdominal distension were also recorded. Descriptive data were reported as means and standard deviations. Patients were divided into feeding during HT and no feeding during HT groups based on whether EN was initiated during HT. Outcomes between the two groups were compared using SAS version 9.3 and chi square for categorical variables and T test for continuous variable comparisons. (See Figure 1.) Results: Forty eight of the 141 reviewed patients were excluded. Among the 93 patients on hypothermia, 8 patients (9%) were fed while undergoing hypothermia. There were no recorded incidences of emesis, ileus or diarrhea in this


group. Only one (12%) of the patients experienced high residual volume. Mean peak glucose values were higher (219) during phase I of HT compared to phase II (137). (See Figure 3.) Overall mortality was 53%(49) in the group that was not fed and 47% among the group with EN. Analysis of data showed that in 47% of patients EN was started immediately after the completion of HT protocol. There were no differences in baseline demographics. Among those fed, the maximum EN rate achieved was 26 ml per hour. During phase II of hypothermia, patients received an average 345 kcal and 20 gram protein. (See Figure 2.) Conclusions: Enteral nutrition was well tolerated during rewarming phase of hypothermia when initiated. Most of the orders for initiation of EN during HT were physician-driven, but the results display a tendency among dietetic clinicians to recommend waiting until after the completion of HT to initiate EN. Given the retrospective nature of the study and lack of control group, we are unable to make causal inferences. However, as the benefit of nutrition tends to be time dependent, trickle feeds/smaller volume feeds may be considered for those at high risk for malnutrition. Randomized, prospective studies with large sample size are needed to assess the tolerance to and benefit of enteral nutrition during hypothermia. Financial Support received from: None

Figure 1. Enteral nutrition during Phase I and Phase II of hypothermia.


Figure 2. Nutrition support during hypothermia.

Figure 3. Blood glucose values in patients with enteral nutrition vs no enteral nutrition.


S63—Predictors of Skeletal Muscle Loss Are Different When Comparing the Initial Acute Phase Response versus Prolonged Mechanical Ventilation Sarah Peterson, PhD, RD1; Patricia Sheean, PhD, RD2; Urvi Taylor, MD3; Joy Sclamberg, MD3; Omar Lateef, DO4; Carol Braunschweig, PhD, RD5 1Rush University Medical Center, Chicago, IL, USA; 2Marcella Niehoff School of Nursing, Loyola University, Maywood, IL, USA; 3Department of Diagnostic Radiology and Nuclear Medicine, Rush University Medical Center, Chicago, IL, USA; 4Department of Internal Medicine, Rush University Medical Center, Chicago, IL, USA; 5Department of Kinesiology and Nutrition, University of Illinois at Chicago, Chicago, IL, USA Purpose: Exploitation of abdominal computed tomography (CT) scans completed for diagnostic purposes can be used to quantify skeletal muscle (SM) cross-sectional area. The objective of this study was to determine predictors of change in abdominal SM change in ICU patients with two or more abdominal CT scans completed during their ICU stay. Methods: Patients who met eligibility criteria for our recent randomized clinical trial (RCT), designed to assess the impact of intensive medical nutrition therapy versus standard care on clinical outcomes among patients with acute lung injury (INTACT), but not consented were included in this retrospective cohort study. Further inclusion criteria included two abdominal CT scans of the third lumbar (L3) region completed within at least 5 days and while receiving nutrition support in the ICU. Cross-sectional SM area (cm2) was computed from L3 CT scans by summing muscle tissue pixels (Hounsfield Units of -29 to +150) and multiplying by pixel surface area. Potential predictors of muscle change were collected: age, gender, BMI and severity of illness scores. Calorie delivery (including delivery of enteral nutrition, parenteral nutrition, propofol and dextrose containing intravenous fluids) and medication administration was collected between the first (CT1) and second CT scan (CT2). Percent change in cross-sectional SM area per day variable was calculated for each patient (absolute differences between cross-sectional SM area/CT1 cross-sectional SM area)/divided by number of days between scans *100). Forward, stepwise linear regression models were utilized to determine significant predictors of overall percent change per day. Additional models were created to determine if predictors of percent change of cross-sectional SM area per day were different when CT1/CT2 were completed within 8 days of requiring mechanical ventilation (n=8, early ICU scans) versus when CT1/CT2 were completed after 7 days of mechanical ventilation (n=9, late ICU scans). Results: A total of 37 patients who had two diagnostic abdominal CT scans while receiving nutrition support were included. The mean age and BMI were 55.4 ± 15.5 and 30.9 ± 8.9, respectively. Almost half (49%) of the sample was female. The average length of time between CT1 and CT2 was 16.1 ± 9.2 day. During this interval patients received 13.7 ± 6.8 Kcal/Kg and 0.7 ± 0.4 grams protein/Kg. Patients lost an average of 0.71% SM cross-sectional area per day between CT1 and CT2. Received insulin (β: 0.90, p=0.004) and BMI (β: -0.04, p=0.02) were significant predictors of overall percent change per day. Predictors of SM cross-sectional area change varied dependent upon time of ICU discharge. Significant predictors of SM cross-sectional area change per day were BMI (β:-0.07, p=0.01), APACHE II (β: -0.09, p=0.02) and insulin received (β: 0.97, p=0.03) for patients with both CT1 and 2 completed within < 7 days of intubation. When patients had both CT1 and 2 completed >8 days of intubation calorie delivery (Kcal/kg, β:0.05, p=0.005), male (β:-0.70, p=0.006), paralytic exposure (β:-0.85, p=0.003) and admit SOFA (β:-0.07, p=0.01) predicted changes in SK muscle area. Conclusions: Calorie delivery did not influence skeletal muscle during the acute interval of illness, however energy supplied days 8+ attenuated muscle loss. These findings suggest that early calorie delivery does not influence skeletal muscle changes. Increased calorie exposure provided after day 8, indicative of prolonged mechanical ventilation, prevented muscle loss. Financial Support received from: None


S64—Problems With Optimal Energy Delivery in the Pediatric Intensive Care Unit Yara M. Moreno, RD, PhD1; Daniela B. Hauschild, RD, MSc1; Eliana Barbosa, RD, MSc2; Nilzete L. Bresolin, MD, MSc3; Nilesh M. Mehta, MD4 1Graduate Program in Nutrition, Federal University of Santa Catarina, Florianópolis, Brazil; 2Nutrition, Joana de Gusmão Children's Hospital, Florianópolis, Brazil; 3Pediatric Intensive Care Unit, Joana de Gusmão Children's Hospital, Florianópolis, Brazil; 4Division of Critical Care Medicine, Department of Anesthesiology, Perioperative and Pain Medicine, Boston Children’s Hospital, Boston, MA, USA Purpose: Optimal delivery of energy is associated with improved outcomes in critically ill children and therefore is an important aspect of care. The use of standard equations to determine energy requirements may result in inaccurate prescription and other barriers impede the delivery of prescribed nutrients in the pediatric intensive care unit (PICU). The gaps between requirement, prescription and delivery of energy need to be addressed in order to optimize the nutrition therapy in this vulnerable population. We aimed to examine the agreement between the standard equations used to predicted energy expenditure (PEE) in the PICU. We also aimed to describe the prescription and actual energy delivery in the PICU. Methods: Single-center, prospective cohort study, including consecutive children (age: 1 month to 15 years) admitted to the PICU of an academic hospital in Brazil. Demographics, clinical characteristics and nutrient delivery details were recorded. The cumulative balance between the energy required and energy intake; and between the prescribed and actual energy intake (AEI) were calculated for up to 6 days in the PICU. Underfeeding was defined as energy intake < 80% of the predicted value. Categorical variables were described in absolute values and frequency. Quantitative variables as median and interquartile range [IQR]. Bland-Altman method was used to describe the agreement between PEE by Schofield, World Health Organization (WHO) and Dietary Reference Intake (DRI) equations, and the Holliday-Segar method. Results: We enrolled 130 patients, 37% female, median [IQR] age 29.43 [4.03; 100.63] months, with PICU length of stay 6 [4; 13] days. The median [IQR] values for the predicted, prescribed and actual energy intake were 47.13 [38.60; 55.38], 31.94 [13.99; 51.90] and 25.06 [10.21; 46.92] kcal/kg/day, respectively. On average, intake was 47% of the estimated requirement and, 68% of patients were underfed. Nutrient delivery (enteral or parenteral) was interrupted in 64% of patients. Mean energy intake was lower in patients with enteral nutrition interruptions in comparison to those with uninterrupted enteral nutrition [25.67 (15.02; 47.79) vs. 41.11 (13.25; 56.71) kcal/kg/day, p= 0.044]. The PEE values by Schofield and WHO equations were in agreement with a mean bias (95% Confidence Interval) of 1.06 (-29.11; 31.25)%. The DRI and Holliday-Segar values were significantly greater than those by Schofield (Figure 1). Conclusions: A majority of the children in our PICU failed to receive the prescribed energy goal in the first week, due to suboptimal prescription (gaps between the predicted requirement and the prescription of energy) and multiple feeding interruptions. These are potentially modifiable factors that might improve energy delivery in this cohort. On the other hand, the use of Holiday-Segar and DRI equations may result in unintended overfeeding and should be avoided. Financial Support received from: Post-Doctoral scholarship from the CAPES Foundation, Ministry of Education, Brazil (POS_DOC - 99999.006148/2014-00) to Moreno, YMF. Doctorate scholarship from the CAPES Foundation, Ministry of Education, Brazil to Hauschild DB.


S65—Nutritional Status on Admission Is Associated With the Duration of Mechanical Ventilation in Critically Ill Children Yara M. Moreno, RD, PhD1; Rafaela B. Grippa, RD, MSc1; Paola S. Silva2; Eliana Barbosa, RD, MSc3; Nilzete L. Bresolin, MD, MSc4; Nilesh M. Mehta, MD5 1Graduate Program in Nutrition, Federal University of Santa Catarina, Florianopolis, Brazil; 2Nutrition, Federal University of Santa Catarina, Florianópolis, Brazil; 3Nutrition, Joana de Gusmão Children's Hospital, Florianópolis, Brazil; 4Pediatric Intensive Care Unit, Joana de Gusmão Children's Hospital, Florianópolis, Brazil; 5Division of Critical Care Medicine, Department of Anesthesiology, Perioperative and Pain Medicine, Boston Children’s Hospital, Boston, MA, USA Purpose: Critically ill children admitted to the Pediatric Intensive Care Unit (PICU) are often malnourished. The prevalence of moderate and severe malnutrition varies between 20 and 53% depending on the location and the criteria used. This condition may worsen during the PICU stay and, in addition, increase the risk factor for unfavorable clinical outcomes, such as the duration of mechanical ventilation. Our aim was to examine the role of nutritional status based on anthropometric parameters on admission as a predictor of the duration of mechanical ventilation in critically ill children admitted to PICU. Methods: Single-center, prospective cohort study, including consecutive children (age: 1 month to 15 years) admitted to a tertiary center in Brazil, requiring mechanical ventilation (MV) in the PICU. Demographics, clinical characteristics and nutritional status were recorded and patients were followed up until hospital discharge. Nutritional status was evaluated by anthropometric parameters and, patients were classified as malnourished if the z-scores for the anthropometrical parameters. Malnutrition was considered when the z-score of anthropometric parameters were ≤ -2. The primary outcome was duration of mechanical ventilation and the secondary outcome was


mortality. Categorical variables were described in absolute values and frequency. Quantitative variables as median and interquartile range [IQR]. Cox regression analysis was used to determine the association between nutritional status and outcomes. Results: A total of 72 patients were included, median age (IQR) 21.1 (4.4; 82.2) months, with median duration of mechanical ventilation 5 (3; 11.5) days and median PICU length of stay 7 [4.5; 13.5] days. The prevalence of malnutrition was 41.2%, according to height-for-age z-score and 18.6% according to weight-for-height z-score. Nutrition status at admission was associated with the duration of mechanical ventilation (Figure 1). The anthropometrical parameters that predicted duration of MV were weight-for-age (Hazard Ratio [HR]: 2.73; 95% Confidence Interval [CI]: 1.44; 5.18); height-for-age (HR:2.49; 95% CI: 1.44; 4.28); and mid-upper-arm muscle area-for-age (HR: 5.22; 95%CI: 1.19; 22.76). Nutrition status at PICU admission was not associated with hospital mortality. Conclusions: Malnutrition, based on a variety of anthropometric variables, was significantly associated with duration of MV in this cohort of critically ill children. Assessment of nutritional status by anthropometry should be performed on admission to the PICU so as to allow targeted nutritional rehabilitation for the subset of children with existing malnutrition. Financial Support received from: Post-Doctoral scholarship from the CAPES Foundation, Ministry of Education, Brazil (POS_DOC - 99999.006148/2014-00) to Moreno, YMF. Master scholarship from the CAPES Foundation, Ministry of Education, Brazil to Grippa, RB.


S66—Enteral Nutrition Administration Record Prescribing Process Using Computerized Order Entry: A New Paradigm and Opportunities to Improve Outcome in the Critically Ill Amir Y. Kamel, PharmD, BCNSP1,7; Stephen J. Lemon, PharmD, BCPS1,7; Sandra W. Citty, PhD, ARNP-BC3; Fatima E. Guadamuz, PharmD Candidate7; Lee Marlowe, RDN, LDN6; Lynn Westhoff, MSN, MHA, CCRN, RN5; Shawn Sanders, BA4; Frederick A. Moore, MD, FACS, MCCM2 1Pharmacy, University of Florida, Gainesville, FL; 7College of Pharmacy, University of Florida, Gainesville, FL, USA; 3College of Nursing, University of Florida, Gainesville, FL, USA; 6Nutrition, UF Health, Gainesville, FL; 5Nursing, UF Health, Gainesville, FL, USA; 4Pharmacy Informatics and Automation, UF Health, Gainesville, FL, USA; 2Surgery, University of Florida, Gainesville, FL, USA Purpose: Malnutrition occurs in up to 60% patients of hospitalized patient and is associated with complications such infection, poor wound healing, increased morbidity and mortality. Nutrition support is a vital component in improving clinical outcomes in the critically ill and has been shown to reduce avoidable re-hospitalization, and decrease incidence of pressure ulcers. At UF Health, we have established Enteral Nutrition (EN) protocols for the critical ill to allow for a standardized practice throughout different intensive care units. Guideline driven protocols have been shown to reduce infections, overall complication, and hospital length of stay. Hospitals that use feeding protocols are more effective at starting nutritional support an average of 3.1 days earlier. Protocols can also improve the proportion of patients meeting at least 80% of their nutritional goals. Compared to hospitals without a feeding protocol there is a 10% decrease in mortality and shorter length of stay with protocols. While the current Computerized Physician Order Entry (CPOE) allows for documenting the initiation and rate changes of EN, there remain significant limitations. There is an inherent redundancy that can create confusion when viewing and/or documenting EN orders. Nurses have to document the initiation and rate changes of EN in multiple places. Relevant information has to be searched through multiple areas within the system yet one cannot determine full caloric prescription from a given formula and additional supplements at a given time. Incorporating nutritional information into one location can reduce medical errors through better access to patient data and enhanced communication with the health care team. Additionally, frequent interruptions due to procedures, imaging, or surgery lead to feeding interruptions where EN could be stopped for many hours limiting the amount of nutrition the patient receives down to 76% compared to 87% with no interruptions, creating a caloric deficit according to one study. Creating a new Electronic Nutrition Administration Record (ENAR) with a linked nutrition tab will enable clear documentation and modification to EN located in a single place. It will give an opportunity for nurses to document reasons why they are holding, stopping or interrupting EN. It will help users find information more efficiently and enable relevant data to be easily accessible. This also may result in better adherence to established protocols and early caloric goals achievements that ultimately may result in better patient outcome. Methods: A multidisciplinary nutrition improvement team was established and met weekly to discuss strategies, barriers, and evaluate current process and to establish workflow from point of order entry, delivery, administration and electronic documentation of enteral orders nutritional supplements. ENAR was built to allow for easy documentations of EN formula, rate, rate changes, interruptions or stoppage, additional protein or supplements and flushes. Separate orders panels were built for continuous, cyclic and flush orders with option to add supplemental protein and additives as needed. Feedback and modification of initial built were thought to allow for variant local ICU practices. Nursing and healthcare practitioners’ education materials were developed and training sessions were implemented. Results: The new ENAR went alive with minimal disruption and we have received mostly positive feedback from ordering prescribing, nursing and various ICU medical teams. Conclusions: Utilizing technology to build a full electronic nutrition administration record would result in optimizing patient care. Financial Support received from: W. Martin Smith Interdisciplinary Patient Quality and Safety Awards Program (IPQSA)


S67—Albumin and Prealbumin Do Not Correlate With Nutrient Delivery or Outcomes in Surgical ICU Patients Daniel D. Yeh, MD1; Emily Johnson, RN2; Tara Harrison, RN2; Luis Alfonso Ortiz-Reyes, BS1; Haytham Kaafarani, MD, MPH1; Jarone Lee, MD, MPH1; Peter Fagenholz, MD1; Kathryn Butler, MD1; George Velmahos, MD, PhD1 1Surgery, Massachusetts General Hospital, Boston, MA, USA; 2Massachusetts General Hospital, Boston, MA, USA Purpose: Albumin and prealbumin are commonly believed to be markers of nutritional status and are often serially measured to assess the adequacy of nutritional delivery. As negative acute phase reactants, however, serum levels of these proteins will be decreased during times of stress regardless of nutritional status, thus decreasing their clinical usefulness. The predictive value of these biomarkers has not been extensively described in critically ill surgical patients. Therefore, we aimed to investigate the association of initial and serial serum albumin and prealbumin with nutritional delivery and clinically important outcomes in surgical ICU patients. Methods: We performed a retrospective analysis of an ongoing prospective registry of adult surgical ICU patients. Data collected included demographic data such as age, body mass index (BMI), cumulative ICU calorie/protein deficits (actual delivered calories/protein subtracted from prescribed calories/protein), serial albumin, prealbumin, and C-reactive peptide (CRP) values, and clinical outcomes such as mortality, length of stay (LOS) and 28 day ventilator free-days (28-VFD). Pearson two-tailed correlations were performed to assess the relationship between initial and serial values of albumin, prealbumin, and CRP with nutritional adequacy and clinical outcomes. A p value of <0.05 was considered significant. Results: A total of 213 subjects were included. Initial albumin and prealbumin levels were correlated (r=0.341, p=0.008), as were changes in albumin and prealbumin during hospital stay (r=0.357, p=0.074), though this was not statistically significant. There was no significant correlation between initial albumin or prealbumin with CRP nor were changes in albumin or prealbumin correlated with changes in CRP. Changes in albumin and prealbumin were not correlated with total caloric or protein deficit nor with clinically meaningful outcomes such as ICU LOS, hospital LOS, 28-VFD, or discharge disposition. Initial albumin, prealbumin, and CRP levels were not predictive of clinical outcomes. Conclusions: Traditional biomarkers of nutritional status and adequacy of nutritional delivery are inaccurate in critically ill surgical patients and should not be used to prognosticate outcome nor assess adequacy of nutrition delivery. Financial support: None.


Table 1. Demographics, Nutrition Delivery, and Outcomes.

APACHE = Acute Physiology and Chronic Health Evaluation; BMI = body mass index; CCI = Charlson Comorbidity Index; CRP = C-reactive peptide; EN = enteral nutrition; ISS = Injury Severity Score; LOS = length of stay; VFD = ventilator free-days


S68—Association of Psoas Muscle Characteristics and Clinical Outcomes in Critically Ill Surgical Patients Daniel D. Yeh, MD1; Luis A. Ortiz, BS1; Sadeq A. Quraishi, MD, MHA, MMSc2; Nalin Chokengarmwong, MD1; Laura Avery, MD3; Haytham Kaafarani, MD, MPH1; Jarone Lee, MD, MPH1; Peter Fagenholz, MD1; Kathryn Butler, MD1; George Velmahos, MD, PhD1 1Surgery, Massachusetts General Hospital, Boston, MA, USA; 2Anesthesia, Massachusetts General Hospital, Boston, MA, USA; 3Radiology, Massachusetts General Hospital, Boston, MA, USA Purpose: Cross-sectional area (CSA) of the psoas muscle is associated with sarcopenia and frailty in hospitalized patients. Abdominal/pelvic computed tomography (CT abd/pel) is frequently obtained in critically ill patients, yet the relationship between psoas muscle quality and important clinical outcomes is unclear. We sought to investigate whether psoas muscle CSA (pCSA) and density (in Hounsfield Units, HU) are associated with intensive care unit (ICU) length of stay (LOS), hospital LOS, and discharge destination in surgical ICU patients. Methods: In a retrospective analysis of a registry of nutritional assessments in surgical ICU patients, adults with at least one CT abd/pel within 72 hours of ICU admission were included in the study. pCSA and HU were assessed bilaterally at the L4-L5 intervertebral disk space level and averaged to derive a mean value for each patient. Norms for pCSA and HU have not been established, so we categorized these variables into their respective quartiles. Pearson’s two-tailed correlations were performed. A p value <0.05 was considered statistically significant. To investigate the association of pCSA/HU with ICU and hospital LOS, we performed Poisson regression analyses, controlling for age, sex, body mass index (BMI), Acute Physiology and Chronic Health Evaluation (APACHE) II score, total protein deficit during ICU care, and total complications during hospitalization. In patients with follow-up CT imaging during hospitalization, we performed logistic regression analyses to test the association of percent change in pCSA/HU with discharge destination (home vs. non-home), while controlling for similar confounders. Results: 97 patients met inclusion criteria (Table 1). Pearson’s test demonstrated correlations between pCSA with age (r=-0.461, p<0.001), wt (r=0.575, p<0.001), BMI (r=0.219, p=0.031) and Charlson Comorbidity Index (CCI) (r=-0.452, p<0.001). Additional correlations were demonstrated between HU and age (r=-0.496, p<0.001), wt (r=0.207, p=0.042), and CCI (r=-0.436, p<0.001). Poisson regression analysis demonstrated that compared to patients in the 4th(highest) CSA quartile, those in the 1st (lowest) quartile were likely to have a 29% higher ICU LOS (IRR 1.29, 95%CI 1.01-1.65). Compared to patients in the 4th HU quartile, those in the 1st quartile were likely to have an 11% higher ICU LOS (IRR 1.11, 95%CI 1.01-1.21). In addition, compared to patients in the 4th CSA quartile, those in the 3rd, 2nd, and 1st quartile were likely to have a 19% (IRR 1.19, 95%CI 1.01-1.49), 21% (IRR 1.21, 95% CI 1.02-1.41), and 23% (1.23, 95%CI 1.03-1.37) higher hospital LOS, respectively. Similarly, compared to patients in the 4th HU quartile, those in the 3rd, 2nd, and 1st quartile were likely to have a 17% (IRR 1.17, 95%CI 1.02-1.34), 26% (IRR 1.26, 95% CI 1.08-1.47), and 31% (IRR 1.31, 95%CI 1.09-1.58) higher hospital LOS, respectively. Logistic regression analyses in patients with follow-up CT scans (n=35) demonstrated that each percentage loss in pCSA was associated with a 2% (OR 1.02, 95% 1.01-1.05) increased risk of non-home discharge, and each percentage loss in HU was associated with a 3% (OR 1.03, 95%CI 1.01-1.07) increased risk of non-home discharge. Conclusions: In surgical ICU patients, initial psoas muscle size and quality are associated with age, weight, pre-existing co-morbid illness (reflective of overall frailty) and ICU and hospital LOS. Moreover, declining psoas muscle size and density was associated with a higher likelihood of non-home discharge after acute care hospitalization. Larger, prospective studies are needed to verify and build upon our findings. Future studies are needed to determine whether aggressive nutritional strategies can influence psoas muscle quality and clinical outcomes. Financial support: None.


Table 1. Demographics, Psoas Muscle Characteristics, and Outcomes.

APACHE = Acute Physiology and Chronic Health Evaluation; BMI = body mass index; CCI = Charlson Comorbidity Index; CSA = cross sectional area; Day3CalDef = cumulative caloric deficit on ICU day 3; EN = enteral nutrition; HU = Hounsfeld units; ISS = Injury Severity Score; LOS = length of stay; VFD = ventilator free-days



S69—An Evaluation of Multifrequency Bioimpedance Phase Angle and Impedance Ratio as Markers of CT-Measured Low Muscularity and Live ICU Discharge in Critically Ill Patients: Preliminary Results From the VALIDation of Bedside Ultrasound of Muscle Layer Thickness of the Quadriceps in the Critically Ill Patient (VALIDUM Study)

Adam Kuchnia, MS, RD1; Carrie Earthman, PhD, RD1; Abigail Cole, PhD1; Gregory Beilman, MD2; Andrew Day, MSc3; Roger Leung, MSc3; Willem Looijaard, MD4; Peter Weijs, PhD4; Heleen Oudemans-van Straaten, PhD4; Marina Mourtzakis, PhD5; Michael Paris5; Charlene Compher, PhD, RD6; Rupinder Dhaliwal, RD3; Sarah Peterson, RD7; Hannah Roosevelt, RD7; Daren Heyland, MD3 1Food Science and Nutrition, University of Minnesota–Twin Cities, Minneapolis–St Paul, MN, USA; 2Surgery, Division of Critical Care/Acute Care, University of Minnesota Medical Center, Minneapolis, MN, USA; 3Clinical Evaluation Research Unit, Kingston General Hospital, Kingston, Ontario, Canada; 4Intensive Care Medicine, VU University Medical Center Amsterdam, Amsterdam, Netherlands; 5Kinesiology, University of Waterloo, Waterloo, Ontario, Canada; 6School of Nursing, University of Pennsylvania, Philadelphia, PA, USA; 7Food and Nutrition, Rush University Medical Center, Chicago, IL, USA

Purpose: Lean tissue depletion has been associated with increased morbidity and mortality, infections, and hospital length of stay. Moreover, loss of lean tissue in clinical populations is often masked by increased adiposity, lending to the exacerbation of this issue. Estimation of lean tissue is not commonly performed in hospital settings in part due to the limited availability of valid bedside methods. Inaccuracies in whole-body lean tissue estimates have led to the investigation of raw bioelectrical impedance analysis (BIA) parameters (eg, 50-kHz phase angle [PA] and 200/5-kHz impedance ratio [IR]) as potential markers of nutrition status, prognosis, and clinical outcomes. Our objective was to test whether PA and IR could be similarly used to assess low muscle index measured by computed tomography (CT) and evaluate whether these BIA parameters are independent predictors of live intensive care unit (ICU) and hospital discharge. Furthermore, we assessed whether PA and IR cut points, created on the basis of low fat free mass index as measured by dual-energy X-ray absorptiometry in a healthy population, could be used to further predict live ICU and hospital discharge.

Methods: This was a multicenter prospective observational study, which included all patients greater than 18 years of age admitted to the ICU. Further inclusion criteria consisted of having an abdominal CT scan performed for clinical reasons less than 24 hours prior to admission and up to 72 hours after admission. Moribund patients not expected to survive were excluded. CT scans were landmarked at the third lumbar vertebra and were analyzed for skeletal muscle cross-sectional area. Multifrequency BIA was performed according to standard protocol within 72 hours following the abdominal CT scan.

Results: Of our 71 patients, 62% were male. Overall, patients were 57 ± 16 years old and had a BMI of 29 ± 8 kg/m2 (mean ± SD). Mean APACHE II score was 16 ± 7. According to BMI, 39% were classified as normal weight; 31% were overweight; and 30% were class I obese. CT scans revealed that 57% of patients had lower-than-normal muscularity, defined as CT-derived muscle index <55.4 cm2/m2 for males and <38.9 cm2/m2 for females. Based on linear regression, PA alone was able to predict 28% of the variance in CT muscle index and 58% of the variance when covariates were added to the model (age, sex, BMI, Charlson comorbidity index, and admission type; Table 1). Similarly, IR alone was also able to predict 26% of the variance in CT muscle index and 57% of the variance when covariates were added. The area under the receiver operator curve (c index) to predict CT-defined low muscle mass index was 0.73 for both PA and IR. With covariates added to logistic regression models including PA and IR, the c indexes were 0.87 and 0.86, respectively (Figures 1 and 2). PA and IR were both able to predict live ICU discharge. When low fat free mass index cut points for PA and IR were used, they were able to predict live ICU discharge with greater certainty (c index = 0.77, 0.79 and P = .04, .01, respectively; Table 2).

Conclusions: Despite the fact that 61% of the patients were overweight or obese, PA and IR were useful tools in predicting low muscularity defined by abdominal CT scans. When information such as BMI, age, sex, and comorbidity index were known, PA and IR were strong predictors of muscularity and live ICU discharge. Our preliminary results illustrate the potential utility of PA and IR as markers to identify patients with low muscularity who could benefit from early and rigorous intervention.



Table 1. Use of PA in Linear Regression Models to Predict Muscle Index and Area.

Outcome Predictors R2 Adjusted R2

RMSE P Value Model

P Value PA

CT muscle index (cm2/m2) PA 0.28 0.27 9.8 <.0001 <.0001

CT muscle index (cm2/m2) Covariatesa 0.50 0.46 8.4 <.0001 NA

CT muscle index (cm2/m2) Covariates+PA 0.58 0.54 7.74 <.0001 .0008

CT muscle area (cm2) PA 0.21 0.20 37.7 <.0001 <.0001

CT muscle area (cm2) Covariatesa 0.60 0.57 27.5 <.0001 NA

CT muscle area (cm2) Covariates+PA 0.65 0.61 26.2 <.0001 .008

CT, computed tomography; PA, phase angle. aCovariates are age (linear), sex (binary), BMI (linear), Charlson comorbidity index (linear), and admission type (binary).

Table 2. Ability of PA and IR to Predict Relative Time to Live Discharge by Cox Regression.a

Predictors Outcome c Index of Model

P Value of PA or IR

PA (continuous) Time to live ICU discharge 0.599 .028

PA (continuous) Time to live hospital discharge 0.528 .49

PA (continuous)+covariatesb Time to live ICU discharge 0.611 .008

PA (continuous)+covariatesb Time to live hospital discharge 0.599 .43

PA<5% cut point Time to live ICU discharge 0.774 .04

PA<5% cut point Time to live hospital discharge 0.613 .70

IR (continuous) Time to live ICU discharge 0.584 .032

IR (continuous) Time to live hospital discharge 0.513 .46

IR (continuous)+covariatesb Time to live ICU discharge 0.608 .0095

IR (continuous)+covariatesb Time to live hospital discharge 0.587 .39

IR<5% cut point Time to live ICU discharge 0.786 .01

IR<5% cut point Time to live hospital discharge 0.563 .75

ICU, intensive care unit; IR, impedance ratio; PA, phase angle. aPA cut point is 6.75 for men and 5.85 for women. IR cut point is 0.78 for men and 0.81 for women. Cut points are based on bioimpedance spectroscopy data from NHANES IV and V, with low muscle mass defined by fat-free mass index at or below the 5th percentile. bCovariates in model include sex, age (linear), and BMI (linear).


Figure 1. ROC curve for phase angle with covariates to predict low muscle index.

Area Under the Curve = 0.87

1 - Specificity

Sen

sitiv

ity

Points labeled by predicted probability


Figure 2. ROC curve for impedance ratio with covariates to predict low muscle index.

S70—The Patient- and Nutrition-Derived Outcome Risk Assessment (PANDORA) score as a predictor of mortality in critically ill patients Anna Nakayama, RD, LDN1; Cecilia Canales, MPH2; Daniel D. Yeh, MD4; Donna Belcher, MS, RD, CNSC, LDN1; Caitlin M. McCarthy, BA6; Sadeq A. Quraishi, MD, MHA, MMSc3,5 1Nutrition and Food Services, Massachusetts General Hospital, Boston, MA, USA; 2University of California School of Medicine, University of California, Irvine, Irvine, CA, USA; 3Department of Anesthesia, Critical Care and Pain Medicine, Massachusetts General Hospital, Boston, MA, USA; 4Department of Surgery, Massachusetts General Hospital, Boston, MA, USA; 5Harvard Medical School, Harvard University, Boston, MA, USA; 6Tufts University School of Medicine, Tufts University, Boston, MA, USA Purpose: The ability to accurately predict mortality, early in the course of critical illness, may influence goals of care in the intensive care unit (ICU). Despite growing evidence to support its association with mortality after critical illness, nutritional status remains largely underappreciated. Indeed, even the most commonly used mortality prediction tool worldwide, the Acute Physiology and Chronic Health Evaluation (APACHE II) score, does not specifically account for nutritional status. Recently, the Patient- and Nutrition-Derived Outcome Risk Assessment (PANDORA) score was developed and validated as a tool for predicting 30-day mortality in hospitalized patients; however, its utility in the ICU setting is not known. Therefore, our goal was to investigate whether PANDORA scores are associated with mortality in critically ill patients and to compare PANDORA scores to APACHE II scores as predictors of mortality after initiation of ICU care. Methods: We performed a retrospective analysis of data from an ongoing study of nutritional status and outcomes in critically ill patients. Daily protein and caloric deficit (estimated nutrient requirement based on American Society

Area Under the Curve = 0.86

Points labeled by predicted probability

1 - Specificity

Sen

sitiv

ity


for Parenteral and Enteral Nutrition guidelines–actual nutrient delivery) was summed over the duration of ICU stay for each patient. Mortality rate was assessed at 30, 90, and 180 days after hospitalization. To investigate the association of PANDORA scores with mortality, we performed logistic regression analyses, while controlling for age, sex, race, body mass index, as well as protein deficit, caloric deficit, and length of stay in the ICU. We also investigated the association of APACHE II scores with mortality in our cohort of patients, while controlling for the same covariates. We then compared PANDORA to APACHE II as a predictor of mortality after critical illness by constructing receiver operating characteristic curves and using the “roccomp” function in STATA v13 (StataCorp: College Station, TX) to discriminate between the areas under the curve (AUC) for each time-point. Results: 312 patients comprised the analytic cohort (Table 1). Mortality rate was 21%, 28%, and 31%, at 30, 90, and 180 days, respectively. Each unit increment in PANDORA and APACHE II scores was associated with an 8% (OR 1.08: 95%CI 1.04-1.13; p<0.001) and 12% (OR 1.12: 95%CI 1.08-1.16; p<0.001) increased risk of 30-day mortality, respectively. PANDORA scores were equivalent to APACHE II scores as a predictor of mortality at 30-days (AUC 0.69: 95%CI 0.62-76 vs. 0.73: 95%CI 0.68-0.81, p=0.29 - Figure 1). At 90 days, each unit increment in PANDORA and APACHE II scores was associated with a 9% (OR 1.09: 95%CI 1.03-1.12; p<0.001) and 12% (OR 1.12: 95%CI 1.06-1.13; p<0.001) increased risk of mortality, respectively. PANDORA scores were equivalent to APACHE II scores (AUC 0.71: 95%CI 0.63-0.77 vs. 0.74: 95%CI 0.67-0.80, p=0.52) as a predictor of mortality at 90-days. And finally, at 180 days, each unit increment in PANDORA and APACHE II scores was associated with a 10% (OR 1.10: 95%CI 1.06-1.15; p<0.001) and 13% (OR 1.13: 95%CI 1.08-1.17; p<0.001) increased risk of mortality, respectively. PANDORA scores were equivalent to APACHE II scores (AUC 0.73: 95%CI 0.67-0.79 vs. 0.75: 95%CI 0.69-0.81, p=0.66; Figure 1) as a predictor of mortality at 180-days. Conclusions: In our cohort of critically ill patients, PANDORA scores were independently associated with mortality at 30, 90, and 180 days after initiation of critical care. PANDORA scores were also found to be equivalent to APACHE II scores for mortality prediction at 30, 90, and 180 days after initiation of critical care. Future studies are needed to assess whether prospective assessments of PANDORA scores can help to predict outcomes and influence care in ICU patients. Financial Support received from: None


Figure 1. Receiver operating characteristic curves for the Acute Physiology and Chronic Health Evaluation II and the Patient- and Nutrition-Derived Outcome Risk Assessment scores for 30- and 180-day mortality in critically ill patients (n = 312). APACHE, Acute Physiology and Chronic Health Evaluation; AUC, area under the curve; PANDORA, Patient- and Nutrition-Derived Outcome Risk Assessment. A: AUC for 30-day mortality (p = 0.29). B: AUC for 180-day mortality (p = 0.66).


S71—Loss of Lean Tissue and Fat Gain in Exercising Breast Cancer Patients May Suggest Need for Promoting Greater Protein and Reduced Fat Intake: Preliminary Results Lisa Bos, BASc1; Amanda Milburn, MSc1; Schuyler Schmidt, BSc1; Caryl Russell, MSc1; Marina Mourtzakis, PhD1 1Kinesiology, University of Waterloo, Waterloo, Canada Purpose: Loss of lean and gain of fat mass are hallmark changes in breast cancer (BC) patients and survivors. Despite the high 5-year survival rates (88%) in BC, these deleterious body composition changes are associated with increased risk of diabetes, cardiovascular disease, and cancer recurrence. Exercise is increasingly used to manage these body composition changes; however, dietary interventions and assessments of dietary changes during active treatment are lacking. Combining nutrition and exercise interventions can optimally improve metabolic and health benefits. To better understand the appropriate dietary intervention for BC patients who exercise during treatment, we sought to better understand dietary intake in these BC patients. Thus, our objectives were: 1. To quantify total caloric, fat, and protein intake of BC patients who are on active treatment. 2. To describe longitudinal change in total caloric, fat, and protein intake, relative to changes in body composition, during the exercise program in a subset of BC patients. Methods: BC patients, at any stage of disease, and any regimen or phase of treatment, were recruited. Participants exercised 2x/week and each session consisted of 1 hour of supervised light-moderate intensity cardiovascular and strength exercise. Dietary intake was assessed using 3-day food records at 2 different time-points: baseline (onset of the exercise program; n=36) and following the exercise program (~12-24 weeks; n=23). Food diaries were analyzed using the ESHA Food Processor and compared to the Harris Benedict Equation (HBE) as well as current dietary recommendations for the BC population. Dual energy x-ray absorptiometry (DXA) as well as body mass index (BMI) were used to assess body composition. Results: At baseline, patients were (mean+SD) 53±10 years old, had 39.7±6.1% body fat, BMI of 27.1±5.4 kg/m2, resulting in 56% who were overweight/obese. Total caloric intake was 1813±332 kcal/day, protein intake was 78±18 g/day and total fat intake was 69±19 g/day and average macronutrient distribution was 50±6% carbohydrates, 34±4% fat, and 17±3% protein. Our preliminary results indicate that 18/36 (50%) of the participants exceeded HBE recommendations (calculated using an activity factor of 1.375) for total caloric intake. Also, 94% of participants had fat intakes that were >20% of total calories and 67% of patients had fat intakes >30% of total calories. About 89% of participants met their protein requirements based on 0.8g/kg/day, while only 47% of participants met their requirements if based on 1.2 g/kg/day. For the subset of patients who completed the exercise intervention, despite the lack of changes in total caloric, protein and fat intakes during the study trajectory, % fat mass tended to increase (Pre vs post exercise: 39.95±6.67 vs 40.02±6.41%, p=0.0546, n=22) and total lean mass may be decreasing (pre vs post exercise: 37.76±5.66 vs 35.79±9.02 kg, p= 0.206; n=22). Conclusions: With 67% of BC patients consuming >30% fat and with 47% consuming <1.2 g/kg/day protein from their diet, these features of dietary intake may contribute to fat gains and lean tissue losses reported in the literature in BC survivors. In the present study, the tendency for increased fat and reduced lean tissue in exercising BC patients may also be attributed to these features in dietary intake, despite the lack of changes over the study trajectory. It is possible that without exercise, these fat gains and lean tissue losses may have been worse. Our preliminary findings suggest that future work may focus on the potential need for promoting additional protein and reducing fat intakes in BC patients who are exercising during cancer treatment. Financial Support received from: Canadian Institute for Health Research (CIHR)


S72—Severe Vitamin D Deficiency in Critically Ill Patients With Traumatic Injuries Roland N. Dickerson, PharmD1; Jonathan R. Van Cleve, Pharm.D1; Joseph M. Swanson, PharmD1; George O. Maish, MD2; Gayle Minard, MD2; Rex O. Brown, PharmD1 1Clinical Pharmacy, University of Tennessee, Memphis, TN, USA; 2Surgery, University of Tennessee, Memphis, TN, USA Purpose: Vitamin D has been shown to be an important modulator in the immunologic response to infection. Recent studies indicate that critically ill patients are susceptible to vitamin D insufficiency or deficiency. However, studies examining vitamin D deficiency in trauma patients are lacking. The purpose of this study was to evaluate the prevalence and risk factors for severe vitamin D depletion in critically ill patients with traumatic injuries who require nutrition support therapy. Methods: Critically ill adult patients (18 years of age or older) admitted to the trauma intensive care unit (TICU) who were referred to the Nutrition Support Service (NSS) between June 2013 and June 2014 were retrospectively evaluated. Patients were routinely screened for vitamin D depletion within several days following admission to the ICU and post-fluid resuscitation to avoid erroneously depressed serum 25-hydroxy vitamin D concentrations (25-OH vit D). Serum 25-OH vit D was assayed by the hospital laboratory via an automated chemiluminescent microparticle technique (minimum detection limit of 8 ng/mL, intra and inter-assay coefficient of variations were between 2.4% and 4.6%). Normal 25-OH vit D was established as within 30 to 80 ng/mL. Patients with a 25-OH vit D of 20 - 29.9 ng/mL were considered vitamin D insufficient whereas those with a 25-OH vit D of 13.1 to 19.9 ng/mL were categorized as deficient and < 13 ng/mL as severely deficient. Patients known to receive vitamin D beyond a daily multivitamin supplement prior to hospital admission or those without a 25-OH vit D determination within 2 weeks of ICU admission were excluded. Patients who were deficient or severely deficient were supplemented with 50,000 IU of ergocalciferol once to 3 times weekly while in the ICU when enteral access was available. This study was approved by the University IRB and hospital office for medical research. Continuous data were given as mean ± SD. Differences among groups were evaluated by one way ANOVA. Pearson correlation analysis was used to assess relationships. Results: A total of 158 consecutive patients were evaluated. One hundred and twenty-one patients were either deficient or severely deficient and 31 were vitamin D insufficient (77% and 20% of the total population, respectively). Forty-six patients (38% of the total population) exhibited severe deficiency (Table 1). African-Americans had a lower 25-OH vit D (15.4 ± 4.3 vs. 19.4 ± 7.3 ng/mL, P = 0.001) and a greater proportion of patients with 25-OH vit D deficiency (91% vs. 64%; Table 1, P = 0.004) than Caucasians. There was no significant difference in baseline serum 25-OH vit D for those who were admitted in the winter (November through March; n = 66) vs. the other months (n = 92): 17.4 ± 7.0 vs. 17.4 ± 5.5 ng/mL, respectively, P = 0.642). Serum c-reactive protein concentration or injury severity score did not correlate with 25-OH vit D (Figures 1 and 2). Nutritional intake and clinical outcomes including survival, hospital/ICU length of stay, and the prevalence of nosocomial infections were not significantly different among groups (Table 2). However, these clinical data may have been skewed by administration of ergocalciferol for the majority of deficient patients (Table 2). Conclusions: The vast majority of critically ill patients with traumatic injuries admitted to the ICU experienced vitamin D insufficiency, deficiency, or severe deficiency. African-American patients had lower 25-OH vit D concentrations than Caucasians and were at risk for deficiency. Extent of stress or inflammation or season did not influence the prevalence of vitamin D deficiency. Further study to ascertain whether vitamin D depletion and vitamin D supplementation influences clinical outcomes for critically ill trauma patients is warranted. Financial Support received from: None


Table 1. Patient characteristics.

Variable Normal Insufficient Deficient Severely Deficient P ≤

25-OH vitamin D conc range, ng/mL 30 to 80 20 to 29.9 13.1 to 19.9 ≤ 13

N 6 31 75 46 Male/ Female, n/n 5/1 27/4 61/14 37/9 0.885

Race Caucasian, n(%) African-American, n (%) Hispanic/Other, n (%)

5 (83%) 1 (17%) 0 (0%)

22 (71%) 5 (17%) 4 (12%)

34 (45%) 32 (43%) 9 (13%)

13 (28%) 27 (59%) 6 (13%)

0.004

Etiology for admit MVA, n (%) GSW/KSW,n(%) Fall/Assault, n (%)

Other, n (%)

1 (17%) 0 (0%)

5 (83%) 0 (0%)

21 (68%) 1 (3%)

6 (19%) 3 (10%)

48 (64%) 14 (19%) 9 (12%) 4 (5%)

18 (39%) 19 (41%) 6 (13%) 3 (7%)

0.001

Age, yr 60 ± 30 46 ± 22 45 ± 18 43 ± 17 0.243 Weight, kg 70 ± 20 85 ± 17 88 ± 23 87 ± 19 0.206

BMI, kg/m2 22 ± 5 27 ± 4 28 ± 7 29 ± 5 0.033 Injury Severity Score 25 ± 6 30 ± 14 27 ± 14 29 ± 13 0.399

C-reactive protein, mg/dL 20.4 ± 12.0 18.3 ± 12.0 22.1 ±

12.5 23.7 ± 11.6 0.330

Prealbumin, mg/dL 8.9 ± 4.3 9.5 ± 5.8 8.8 ± 4.3 8.7 ± 3.8 0.969 Alcohol abuse, n (%) 1 (17%) 10 (32%) 17 (23%) 12 (26%) 0.720

Admitted Nov-Mar, n (%) 3 (50%) 10 (32%) 29 (39%) 24 (52%) 0.300

Serum ionized calcium, mmol/L 1.19 ± 0.06 1.17 ± 0.06 1.18 ±

0.09 1.16 ± 0.07 0.407

Serum phosphorus, mg/dL 2.9 ± 0.3 3.7 ± 0.8 3.6 ± 1.0 3.5 ± 1.0 0.053 25-OH vitamin D, ng/mL 39.2 ± 7.3 23.4 ± 2.5 16.0 ± 2.2 12.8 ± 0.5 0.001

Hospital day of vitamin D determination, d 6.2 ± 2.9 9.6 ± 5.8 7.4 ± 5.3 6.6 ± 4.3 0.070 BMI, body mass index; GSW, gun shot wound; KSW, knife stab wound; Mar, March; MVA, motor vehicle accident; N, number; Nov, November


Table 2. Clinical Outcomes, Vitamin D Intake, and Nutrition Therapy.

Variable Normal Insufficient Deficient Severely Deficient P ≤ 25-OH vitamin D conc range, ng/mL 30 to 80 20 to 29.9 13.1 to 19.9 ≤ 13

N 6 31 75 46 Survived, n (%) 5 (83%) 27 (87%) 64 (85%) 26 (80%) 0.835

ICU LOS, d 14 ± 7 23 ± 9 23 ± 18 21 ± 16 0.053 Hospital LOS, d 22 ± 13 38 ± 20 35 ± 24 38 ± 22 0.132 Infection, n (%) 6 (100%) 28 (90%) 54 (72%) 35 (76%) 0.108

Hospital day of infection, d 6 ± 2 9 ± 7 6 ± 5 8 ± 6 0.215 Receivedsuppl ergocalciferol 0 (0%) 5 (16%) 57 (76%) 35 (76%) 0.001 Day ergocalciferol started, d - 16 ± 4 10 ± 7 11 ± 8 0.286

Ergocalciferol duration, d - 27 ± 24 15 ± 13 20 ± 17 0.095 Caloric intake, kcals/d 1060 ± 524 1397 ± 473 1386 ± 562 1420 ± 412 0.527

Protein intake, g/d 109 ± 41 114 ± 41 106 ± 41 108 ± 30 0.797 EN/PN/both, n/n/n 6/0/0 26/1/4 55/6/14 33/6/7 0.526

Vitamin D from EN/PN, IU/d 818 ± 279 953 ± 326 769 ± 316 800 ± 280 0.084 EN, enteral nutrition; LOS, length of stay; N, number; PN, parenteral nutrition; ICU, intensive care unit

Figure 1. Relationship between 25-OH vitamin D concentration and severity of injury.


Figure 2. Relationship between 25-OH vitamin D concentration and extent of inflammation as assessed by CRP concentration.

S73—Interleukin-18 Mediates Left Ventricular Dysfunction Induced by Western Diet in the Mouse Salvatore Carbone, MS1,2; Paul J. Lee, BS1,3; Adolfo G. Mauro, MS1,3; Jessica A. Regan, BS1,3; Donatas Kraskauskas, MS1,3; Raffaella Buzzetti, MD2; Benjamin W. Van Tassell, PharmD1,4; Antonio Abbate, MD, PHD1,3; Stefano Toldo, PhD1,3 1Internal Medicine/Cardiology, Virginia Commonwealth University, Richmond, VA,USA; 2Experimental Medicine, Sapienza University of Rome, Rome, Italy; 3Victoria Johnson Research Laboratories, Virginia Commonwealth University, Richmond, VA, USA; 4School of Pharmacy, Virginia Commonwealth University, Richmond, VA, USA Purpose: Western diet (WD), rich in saturated fat and sugars, is a risk factor for obesity, cardiovascular disease and heart failure (HF). WD induces left ventricular (LV) systolic and diastolic dysfunction through unknown mechanisms. The proinflammatory cytokine interleukin (IL)-18 impairs myocardial contraction and relaxation and is increased in HF. We hypothesize that IL-18 mediates WD-induced LV. Methods: Adult C57BL/6J (WT) and IL-18 knock out (KO) (N=8/group) mice were randomly assigned to either standard mouse diet (SD, saturated fat 0.8%, sucrose 0%) or WD (saturated fat 12.8%, sucrose 30%) for 8 weeks. ELISA and real time PCR were used to measure IL-18 plasma levels and the heart IL-18 mRNA. Transthoracic echocardiography was performed at baseline and at 8 weeks to measure LV fractional shortening (LVFS), isovolumetric relaxation time (IRT) and myocardial performance index (MPI). A conduction catheter was used to measure LV end diastolic pressure (EDP). LV fibrosis was measured with Masson’s Trichrome staining. Results: WT mice fed with WD had increased myocardial IL-18 mRNA and plasma levels, compared with SD-fed mice. Baseline echocardiographic characteristics were not statistically different between WT and IL-18KO. After 8 weeks, WD-fed WT mice had significantly lower LVFS and higher IRT and MPI than WD-fed IL-18KO mice (see Figure 1). Interstitial myocardial fibrosis was increased in the WT WD group (+88%; p=0.01 vs WT SD) and was lower in the IL-18KO mice (0.73±0.11 vs 1.18±0.12 of the WD-fed WT mice; - p=0.035). Conclusions: WD leads to an increase in IL-18 expression in the heart, and genetic deletion of IL-18 significant attenuates WD-induced LV dysfunction. This suggests that inhibiting IL-18 may prevent the adverse cardiac effects of WD. Financial support: None.


Figure 1. IL-18 induced by Western diet mediates cardiac dysfunction.

S74—Western Diet Leads to the Formation of the Inflammasome in Cardiac Endothelial Cells in the Mouse Salvatore Carbone, MS1,2; Adolfo G. Mauro. MS1,4; Eleonora Mezzaroma,PhD 1,3 Jessica A. Regan, BS1,3; Donatas Kraskauskas, MS1,4; Raffaella Buzzetti, MD2; Benjamin W. Van Tassell, PharmD1,3; Antonio Abbate, MD, PHD1,4; Stefano Toldo, PhD1,4 1Internal Medicine/Cardiology, Virginia Commonwealth University, Richmond, VA; 2Experimental Medicine, Sapienza University of Rome, Rome, Italy; 4Victoria Johnson Research Laboratories, Virginia Commonwealth University, Richmond, VA; 3School of Pharmacy, Virginia Commonwealth University, Richmond, VA Purpose: Western diet (WD) is a risk factor for cardiovascular disease (CVD). The mechanisms by which WD leads to CVD are not completely understood. Dietary fat and sugars exert pro-inflammatory effects in endothelial cells in vitro by inducing the formation of the inflammasome. The aim of the study was to determine whether WD induced the formation of the inflammasome in the heart. Methods: Outbreed-ICR CD-1 8 weeks old male mice were fed for 8 weeks with standard diet (SD) (n=6)(saturated fat 0.8%, sucrose 0%), or WD (n=8)(saturated fat 12.8%, sucrose 30%). Mice underwent transthoracic echocardiography to measure left ventricular (LV) function, followed by pathology to measure the formation of the inflammasome, staining for ASC (apoptosis speck-like protein containing a caspase recruitment domain). Results: WD, and not SD, induced a significant impairment in cardiac function (measured as a decrease in LV fractional shortening (-14%, P=0.001) and increase in isovolumetric relaxation time (+47%, P<0.001), and this was associated with the presence of inflammasome specks in endothelial cells in WD-fed mice, but not in SD (Figure 1). Conclusions: WD induces the formation of the inflammasome in cardiac endothelial cells of mice. The pro-inflammatory effects on the cardiac endothelium in vivo may explain the cardiac dysfunction and increased risk of CVD associated with WD. Financial Support received from: None


Figure 1. Western diet induces inflammasome-specks in cardiac endothelial cells.

S75—Parenteral Selenium as Monotherapy or in Combination With Other Antioxidant Micronutrients in the Critically Ill: An Updated Systematic Review and Meta-Analysis William Manzanares, MD, PhD1; Margot Lemieux, RD2; Gunnar Elke, MD3; Pascal L. Langlois, MD4; Frank Bloos, PhD5; Konrad Reinhart, MD5; Daren Heyland, MD, FRCPC, MsC2,6 1Department of Critical Care, Universidad de la República, Montevideo, Uruguay; 2Clinical Evaluation Research Unit, Queens University, Kingston, Canada; 3Department of Anaesthesiology and Intensive Care Medicine, University Medical Center Schleswig-Holstein, Campus Kiel, Kiel, Germany; 4Anesthesia and Reanimation, Université de Sherbrooke, Sherbrooke, Canada; 5Klinik f. Anästhesiologie und Intensivmedizin, Universitätsklinikum Jena, Jena, Germany; 6Department of Medicine, Queens University, Kingston, Canada Purpose: Selenium is an essential trace element with antioxidant, anti-inflammatory and immunomodulatory effects. So far, several randomized controlled trials (RCT), systematic reviews and meta-analyses have demonstrated that parenteral selenium may be able to improve relevant clinical outcomes in intensive care unit (ICU) patients, particularly in those with sepsis and severe sepsis. Nonetheless, the largest trial on selenium therapy in sepsis has recently finished and its results were discouraging. With the aim to update our previous data, we conducted a systematic review and meta-analysis on parenteral selenium alone or in combination with other antioxidant micronutrients in the critically ill. Methods: We searched computerized databases from 1980 to 2015 and included trials that evaluated clinical outcomes associated with parenteral selenium as a single or combined strategy in parenterally or enterally fed patients. Several subgroup analyses were done to elucidate the effects of selenium on mortality and infections (selenium monotherapy vs. combined therapy, selenium loading dose vs. no loading dose, and selenium high dose vs. low dose). Results: 21 RCTs involving 4129 patients met our inclusion criteria. When the data from these 21 trials were aggregated, selenium supplementation had no effect on mortality (risk ratio, RR 0.99, 95 % confidence interval, CI 0.90, 1.08, P= 0.79, heterogeneity I2=0%). When the results of 10 trials that reported on infections were statistically aggregated, parenteral selenium was associated with a trend towards a reduction in infections (RR 0.95, 95 % CI 0.88, 1.02, P=0.15, I2=0%) and when the results of 4 studies that reported on ventilator associated pneumonia (VAP) were aggregated, selenium was associated with a significant reduction in the incidence of VAP (RR 0.62, 95 % CI 0.58, 0.89, P=0.002, I2=0%). With the 12 studies that reported ICU length of stay (LOS), we did not find any


effect (weighted mean difference, WMD 0.32. 95% CI -0.80, 1.43, P=0.43, I2=4%). As well, no effect was seen when analyzing the 8 RCT that reported hospital LOS (WMD -0.83, 95 % CI -3.71, 2.04, P=0.57, I2=0%). The subgroup analysis results are shown in Table 1. Conclusions: Parenteral selenium as monotherapy or in combination with other antioxidant micronutrients has no effect on mortality, ICU LOS, hospital LOS, and ventilator days in critically ill patients. According to our findings, the role of selenium therapy in the critically ill is still uncertain and selenium substitution cannot be recommended to ICU patients. Financial Support received from: None



S76—Clinical Value of Probiotic Treatment for Burn Patients

Yuli Jiang, BS1; Kwaku Opoku1; Eneko Larumbe-Zabala, PhD1; Jenna E. Kesey, APRN, FNP1; John A. Griswold, MD, FACS1; Sharmila Dissanaike, MD, FACS1 1Texas Tech University Health Sciences Center, Lubbock, TX, USA

Purpose: Probiotics are often prescribed to help restore alterations in gut microflora from antibiotic usage in critical illnesses such as burns. While many studies exist on the efficacy of probiotics in gastrointestinal diseases such as Clostridium difficile infections and antibiotic-associated diarrhea, there are very few studies on the efficacy of probiotics in burn patients. One study analyzed the effects of probiotics on 10 pediatric burn cases. They noted a trend toward reduced incidences of fungal infection, improved GI outcomes, and reduced wound healing times, but unfortunately, they lacked sufficient sample size for a definitive conclusion. This study aims to assess the clinical usefulness of probiotics in burn patients.

Methods: A retrospective cohort study in adult burn patients given >1 million CFU/d of Lactobacillus acidophilus and Lactobacillus rhamnosus as routine supplementation. Subjects were divided into 3 groups: patients receiving ≥5 days of probiotics starting in the first week (n = 35), those receiving ≥ 5 days of probiotics starting later than the first week (n = 16), and a control group receiving <5 days of probiotics during their entire stay (n = 57). Data were collected on outcome variables: diarrhea, number of daily bowel movements, emesis, gastric residuals, C difficile infection, sepsis, serum CRP and prealbumin values, length of stay, and mortality. Confounding variables were laxative use and antimicrobial use. Multivariate analysis was performed with adjustment for TBSA and age using STATA.

Results: Demographics and injury characteristics were similar among groups, with mean body surface area of burn, 23.2% (0.2%–69.0%), and age, 46.7 years (18–82). There was no significant difference in clinical outcomes of C difficile infection, sepsis, emesis, gastric residuals, length of stay, or mortality. However, diarrhea occurred more frequently in subjects who received early probiotics (P = .001) and who had an average of 9.5 daily bowel movements vs 4.7 in controls (P = .001) in the first week, which increased to 17.4 vs 11.3 (P = .002) in the second week. Serum CRP was found to be higher for those who received probiotics, 12.45 vs 9.42 pg/mL (P = .015), while prealbumin was lower, 13.52 vs 17.59 pg/mL (P = .015).

Conclusions: Probiotic administration was associated with increased diarrhea and worse serum protein markers in adult burn patients, without perceptible clinical benefit. Alterations in gut permeability and integrity with severe burn may represent a potential cause for this finding. Probiotics should be used with caution in this population, and not in conjunction with other laxatives.

Financial support: None. S77—Impact of BMI and Hyperlipidemia on Life Expectancy of Patients With Amyotrophic Lateral Sclerosis Doris A. Delgado1,2; Elizabeth Bailey 1,2; Anna Plisova1,2; Dacyl Spillane, RDN, LD1,2; Samer Abdul-Hay, PhD, PharmD3; Karen Overstreet, MS, CCRC2; Melissa Stewart, RDN/LD1 1Department of Nutritional Services, Mayo Clinic, Jacksonville, FL, USA; 2Department of Neurology, Mayo Clinic Florida, Jacksonville, FL, USA; 3Department of Pharmacy, Mayo Clinic, Jacksonville, FL, USA Purpose: Length of survival for amyotrophic lateral sclerosis (ALS) patients averages two to five years from time of diagnosis. However, duration of survival in this disease is variable and patients have lived for five years or greater. Recent literature suggests a specific body mass index (BMI) range and the presence of hyperlipidemia as factors affecting survival. The purpose of this study is to determine if these factors contribute to extended duration of survival in ALS patients. Methods: This is a retrospective study analyzing 181 patients diagnosed with ALS. Data analyzed included patient BMI and whether there was a history of hyperlipidemia. Raw data was standardized and outliers excluded. A general linear regression (normal distribution) was used to analyze the effects of BMI and hyperlipidemia (p<0.05) on duration of survival with gender and age of onset as covariates. Results: General linear regression model analysis produced an unexpected result, revealing increased BMI in women improved duration of survival (p<0.0442, Table 1). An inverse effect was displayed in men (p>0.2462) with results indicating decreased length of survival with a higher BMI (Figure 1). Age at onset also showed an inverse


impact on survival for men and women (0.0023, 0.6539, respectively). History of hyperlipidemia showed no significant improvement on duration of survival (Table 1). Conclusions: The findings support the hypothesis that BMI may contribute to prolonged survival in ALS patients. Specifically, these findings suggest that an increased BMI in women and throughout disease progression appears to improve survival outcome with an inverse effect observed in men. Additionally, findings did not find a history of hyperlipidemia to pose a significant impact on duration of survival. Given the ease of providing early nutritional intervention through a high calorie, high fat diet to impact BMI or metabolic lipid chemistry, future studies may benefit from using lipid panels and quantifying power needed to elucidate whether hyperlipidemia impacts survivability in ALS patients. Financial Support received from: None

Table 1. Linear Regression Results

Linear regression and parametric survival fit test for BMI, hyperlipidemia, gender and age at onset; (n=181).


Figure 1. BMI and gender linear graph. Graphical linear regression representation of BMI and survival in days in both male and female subjects; p<0.001 (n=181). S78—Effects of Protein Supplementation on Muscle Functionality, Inflammatory Response, and Ex Vivo Cell Viability in Patients Waiting for Liver Transplantation—A Pilot Study Yani G. Mizubuti, RD1,4; Daniel C. Lima, PhD1; Erica L. Vieira, PhD3; Simone V. Generoso, PhD2; Agnaldo S. Lima, PhD3; Thales A. Silva4; Isabel Correia3 1Ciência de Alimentos, Universidade Federal de Minas Gerais, Belo Horizonte, Brazil; 2Nutrição, Universidade Federal de Minas Gerais, Belo Horizonte, Brazil; 3Medicina, Universidade Federal de Minas Gerais, Belo Horizonte, Brazil; 4Nutrição, Pontifícia Universidade Católica de Minas Gerais, Belo Horizonte, Brazil Purpose: Inadequate ingestion of proteins modifies biological functions in chronic liver disease patients. Whey protein, due to its amino acid profile, digestibility and bioactive compounds may be beneficial. Methods: A randomized, double-blind, intervention study was performed at Hospital das Clínicas, Belo Horizonte. Patients were randomly assigned to receive packages of 20g of whey protein (WP) or casein (CA) to take twice a day (20g in the morning and 20g at night), as a supplement, during 30 days. They were monitored weekly by phone calls. Regular usual diet was maintained. Patients underwent muscle functionality assessment by handgrip dynamometry and 6-min. walking test and, the inflammatory response was assessed by plasmatic cytokines. All tests were performed at the beginning and the end of the study. The ex vivo viability of resting and PHA stimulated peripheral blood mononuclear cells (PBMC) was assessed by flow cytometry. Data were analyzed by bootstrapped two-sample t test for unequal population variances and small sample sizes, and the One-way ANOVA Tukey Post Hoc Significance Test for the ex-vivo experiment. The protocol was approved by the Ethics Committee. Results: Twelve patients were evaluated, 53 ± 11 yrs. old, 9 men and 3 women. There were no differences between treatments regarding the effects on muscle functionality and inflammatory response (Table 1). However, the WP reduced the cell dead of resting and PHA cells in a dose-dependent manner, effects not observed with CA (Figure 1). Conclusions: Despite the immunomodulatory effect associated with WP supplementation, there were no differences in the other parameters when comparing WP and CA. This study will continue including more patients to improve reliability and, better test our hypothesis.

Financial Support received from: Nutricum FAPEMIG


Figure 1. Effects of different concentration of WP and CA on the cell viability of resting or PHA stimulated PBMC. *ƚ

*Values are means. ƚ Data are expressed as the percentage of dead cells determined by propidium iodide staining and, were analyzed by one-way ANOVA and Tukey HSD test.

0

10

20

30

40

50

60

70

80

%Cell Dead PI+ % Lymphocytes PI-A+

%Monocytes PI+

Medium

Casein 40ug

Whey 40ug

Medium+PHA

Casein 40ug+PHA

Whey 40ug+PHA

b b b

a a a

a a a

ac bc b

ab a ab

bd cd cd

0

10

20

30

40

50

60

70

80

%Cell Dead PI+ % LymphocytesPI-A+

%Monocytes PI+

Medium

Casein 400ug

Whey 400ug

Medum+PHA

Casein 400ug+PHA

Whey 400ug+PHAa ac ad

a a a

b b b

ac bc bcd a ac a

ac bc bc

0

10

20

30

40

50

60

70

80

%Cell Dead PI+ % Lymphocytes PI-A+

%Monocytes PI+

Medium

Casein 4mg

Whey 4mg

Medium+PHA

Casein 4mg+PHA

Whey 4mg+PHA

a a a

b b b

a a a

a b a ac ac ac

acd bd bc


Table 1. Effects of protein supplementation on muscle functionality and inflammatory response*

Variable Before supplementation After supplementation T test between

WP and CA after supplementation

WP CA WP CA

Handgrip dynamometry (kg)

22.78 (n=7)

28.52 (n=5)

25.86 (n=5)

37.03 (n=3) p=0.22

6 min. walking test (m)

382.64 (n=7)

351.18 (n=3)

446.6 (n=4)

446.05 (n=4) p=0.97

IL-6 pg/mL 6.42 (n=5) 19.00 (n=4)

1.03 (n=4) 0.00 (n=3)

p=0.35

IL-10 pg/mL 1.14 (n=5) 0.34 (n=4) 1.80(n=4) 1.37 (n=3)

p=0.77

*Values are means S79—Early Aggressive Nutrition Intervention Improves Nutrition Outcomes in Patients on Extracorporeal Membrane Oxygenation Therapy in the Pediatric Intensive Care Unit Jayne M. McMurdy, RD, CNSC, CDN1; Jennifer Zarrilli, RD, CNSC, CDN1 1Clinical Nutrition, Connecticut Children's Medical Center, Hartford, CT, USA Purpose: Extracorporeal Membrane Oxygenation (ECMO) is used in the Pediatric Intensive Care Unit (PICU) for children with severe respiratory failure refractory to conventional mechanical ventilation. Nutrition deficits may begin prior to initiation of ECMO due to critical illness, and exacerbated during therapy due to delayed or inadequate nutrition support in the setting of catabolic stress. Our objective was to improve nutrition delivery to critically ill children before, during, and immediately after ECMO therapy to prevent weight loss and preserve nutritional status. Methods: The ECMO physician coordinator and PICU registered dietitian (RD) identified the need for a process change aimed at improving nutrition outcomes in ECMO patients. Barriers to adequate nutrition were identified for a recent ECMO patient "A": Delay in start of nutrition support, misperception about lipid effects on oxygenator precluding lipid administration; hyperglycemia limiting dextrose calories; delayed advancement and interruptions to enteral feeding, optimism about the ability to discontinue nutrition support and feed orally, and first days of oral feeding limited to fluids. Our interventions for a subsequent ECMO patient "B" included: Immediate initiation of nutrition support, inclusion of lipids with parenteral nutrition, RD presence at daily interdisciplinary patient care rounds to assess intake, adequate overlap of parenteral and enteral nutrition, and continued enteral feeds post-ECMO until adequate oral intake demonstrated. Data collected retrospectively for patients A & B included timeliness of nutrition ordered, amount of nutrition ordered, amount of ordered nutrition delivered, and percentage of nutritional needs met for each day of PICU admission before, during, and after ECMO. Weight measurements at PICU admission and discharge were collected. Results: For patient A, nutrition support was not initiated until day 4 of admission and nutritional needs were not met until day 9. Fifty-six percent of nutritional needs were met during ECMO therapy and forty-nine percent of needs were met throughout PICU stay. Twelve percent weight loss was seen over 14 days, and two criteria for severe malnutrition were met (percent intake and degree of weight loss). For patient B, nutrition support was initiated and nutrition needs were met on day 1. Ninety percent of nutrition needs were met while on ECMO and seventy percent of needs were met throughout PICU stay. Minimal weight loss was seen and nutritional status was maintained. Conclusions: By demonstrating the need for earlier and more aggressive nutrition intervention for ECMO patients and implementing a process change addressing this need, we have been able to preserve nutritional status and maintain weight in a critically ill child. With increased awareness of the essential role of nutrition therapy and increased visibility of registered dietitians as an integral part of the health care team, this same model for process improvement can be applied for other diagnoses both in and out of the PICU to improve nutrition outcomes for our patients. Financial Support received from: None


S80—Early Nutrition Therapy During Pediatric Extracorporeal Membrane Oxygenation (ECMO) Katelyn Ariagno, RD, LDN1; Craig Smallwood, BS, RRT2; Mary Arbuthnot-Thorpe, DO3; Nilesh M. Mehta, MD4,1 1GI/Nutrition, Boston Childrens Hospital, Boston, MA, USA; 2Department of Anesthesia, Division of Critical Care Medicine, Respiratory Care, Boston Childrens Hospital, Boston, MA, USA; 3Department of General Surgery, Boston Childrens Hospital, Boston, MA, USA; 4Division of Critical Care Medicine, Department of Anesthesiology, Perioperative and Pain Medicine, Boston Childrens Hospital, Boston, MA, USA Purpose: Extracorporeal Membrane Oxygenation (ECMO) therapy is employed to support children with cardiorespiratory failure secondary to a variety of illnesses. In 2010 the A.S.P.E.N. guidelines for this population recommended early nutrition support with optimal protein and energy delivery, and an emphasis on the enteral route when feasible. However, a variety of myths related to the safety of the enteral strategy have impeded nutrient delivery in children on ECMO therapy. Protein catabolism during ECMO may result in significant negative nitrogen balance in the absence of adequate nutrient intake. In our current study, we aimed to describe the route and adequacy of energy and protein intake in children during the first 2 weeks on ECMO therapy. In particular, we examined the feasibility of enteral nutrition (EN), and variables associated with failure to deliver EN in this cohort. Methods: In a retrospective cohort study, we reviewed data from children (newborn to 18 years) who were admitted to the pediatric intensive care unit (PICU) and required cannulation for either veno-arterial (VA) or veno-venous (VV) ECMO therapy over a 2 year period (2012-2014). Patients who survived for at least 24 hours after cannulation were included. Daily energy and protein intake via parenteral nutrition (PN) and EN routes was recorded for 7 consecutive days and also on day 14 after cannulation; and described as a percentage (adequacy) of the goal prescribed by the dietitian. We identified variables associated with failure to administer EN in this cohort. Results: Of the 36 patients that received ECMO therapy in the PICU, 28 patients with median (IQR) ECMO duration of 9 (6-24) days met eligibility criteria and were included in the analyses. Table 1 describes the clinical and demographic characteristics of this cohort. Within the first 7 days on ECMO therapy, 93% of the cohort received PN, with the median (IQR) duration of 20 days (8-30). Median (IQR) time to initiation of EN in this cohort was 6 days (2-16). Figure 1 demonstrates energy and protein intake via both the EN and PN routes during the first 2 weeks after ECMO initiation. By day 7, 16/28 (57%) patients were on EN, of which 2/28 (7%) advanced to full EN, being able to wean off PN. A majority of this subset (69%) was fed via the gastric route and the remainder received post pyloric feeds. By day 14, 9/14 (64%) patients were advanced to full EN support. At least 66% of the prescribed energy and protein goal (via EN + PN) was delivered by day 4, which is reflected in Figure 2. Table 2 shows the differences in variables between patients that received any EN and those that did not by day 14. Compared to patients on some EN, the patients that did not receive any EN were more likely to be on vasoactive medications (36% vs. 0%, p = 0.495), had a higher incidence of indwelling umbilical arterial catheter (UAC, 50% vs. 21%, p = 0.115), and had a higher 90-day mortality (57% vs. 14%, p = 0.018). The remaining variables were not different between the 2 subgroups. Conclusions: Over 2/3 of the prescribed energy and protein goals were delivered by day 4 in this cohort of pediatric ECMO. PN was the predominant mode of nutrition support during the first 2 weeks following ECMO cannulation. EN was initiated early but mainly as trophic feeding in the first week. By day 14, 1/3 of macronutrient requirements were delivered via EN in 50% of the cohort. The use of UAC and vasoactive agents was higher in patients that did not receive EN. Current practices surrounding the use of EN in this cohort should be further reviewed given the known benefits of early EN in the critically ill population. Financial Support received from: None


Table 1. Clinical and Demographic Characteristics.


Table 2. Differences Between Subjects That Received EN vs. No EN by Day 14 After ECMO Initiation.

Figure 1. Energy (A) and Protein (B) intake via enteral and parenteral routes during the first 2 weeks after extracorporeal membrane oxygenation initiation in children.


Figure 2. Adequacy (intake as a percentage of prescribed goal) of energy and protein intake during pediatric extracorporeal membrane oxygenation.

S81—Nutritional and Clinical Parameters Are Associated With the Duration of Length on Mechanical Ventilation in Critically Ill Children With Acute Respiratory Failure Larissa T. Feiber, RD1; Daniela B. Hauschild, RD, MSc1; Julia C. Ventura, RD1; Eliana Barbosa, RD, MSc2; Nilzete L. Bresolin, MD, MSc3; Yara M. Moreno, RD, PhD1 1Graduate Program in Nutrition, Federal University of Santa Catarina, Florianopolis, Brazil; 2Nutrition, Joana de Gusmão Children's Hospital, Florianópolis, Brazil; 3Pediatric Intensive Care Unit, Joana de Gusmão Children's Hospital, Florianópolis, Brazil Purpose: Respiratory failure is one of the most common causes of admission in the Pediatric Intensive Care Unit (PICU). Studies suggest that nutrition status and nutrition therapy might influence clinical outcomes, such as duration of mechanical ventilation (MV), PICU length of stay and mortality of critically ill children. The study aimed to determine the association between nutritional status, nutrition therapy and clinical parameters with the duration of MV in critically ill children admitted to a PICU due to acute respiratory failure. Methods: A prospective cohort study, including critically ill children with respiratory failure, aged 1 month to 15 years, requiring MV. Demographics, clinical characteristics, nutrition status and nutrient delivery details were recorded. Patients were followed up until the 21st day after admission or PICU discharge Clinical characteristics were evaluated by Prognostic Index of Mortality 2 (PIM2), PICU length of stay and presence of Systemic Inflammatory Response Syndrome (SIRS), acute respiratory distress syndrome (ARDS), nosocomial infections and, fluid overload. Nutritional status was evaluated by anthropometry and classified based on the z-scores of anthropometrical parameters. The predicted energy expenditure (PEE) was calculated by Schofield equation. Undernutrition was considered when the body mass index-for-age or weight-for-height were ≤-1 z-score and overweight/obese ≥ 2z-score. The cumulative balance between the required and actual intake of energy (AEI) and protein were calculated for up to 7 days in the PICU. Overfeeding was defined as AEI >110% of the PEE. The outcome was duration of MV. Categorical variables were described in absolute values and frequency. Quantitative variables as median and interquartile range [IQR]. Poisson regression, crude and adjusted (for gender, age and PIM 2) were performed, considering P-value <0.05 as significant. Results: A total of 23 patients were included, median age of 7.2 [IQR: 3.0; 26.7] months, median PIM 2 of 1.7 [IQR:1.1; 11.7]%. Pneumonia was the main reason for PICU admission and, 65.2% of patients had ARDS and 74% SIRS. Median duration of MV was 11 [IQR:6; 21] days, median PICU length of stay was 13 [IQR:7; 33] days. The mortality incidence was 17.4%. Regarding the nutritional status on admission, 43.5% of patients were undernourished and, 13% were overweight/obese. Underfeeding was observed in 21.7%, overfeeding in 43.5% and,


fluid overload in 78.3% of patients. The adequate nutritional status at PICU admission was associated with reduction of the duration of mechanical ventilation (incidence rate ratio (IRR): 0.66; 95% Confidence Interval [CI]: 0.50; 0.88) (p=0.005). Overfeeding was associated with the duration of MV (IRR: 1.77; 95%CI: 1.45; 2.16) (p<0.001) in crude analysis, but not after adjusted analysis (IRR: 1.14; 95%CI: 0.88; 1.49) (p=0.315). Fluid overload (IRR 2.36; 95%CI: 1.65; 3.37) (p<0,001), SIRS (IRR: 1.34; 95%CI: 1.03; 1.75, (p=0,029) and nosocomial infection (IRR: 1.97; 95%CI: 1.56; 2.50 (p<0,001) were associated with the duration of MV (Table 1). Conclusions: There was observed a high prevalence of malnutrition at PICU admission. The adequate nutritional status at admission was associate with the reduction of the duration of MV. Overfeeding, as well, fluid overload, SIRS and nosocomial infection might increase the duration of MV. Thus, the clinical care of critically ill children with respiratory failure should prioritize the adequacy of nutritional status and nutrition therapy; the management of fluid balance as well, the prevention of nosocomial infection. Financial Support received from: Master of Science scholarship from the CAPES Foundation, Ministry of Education, Brazil to Feiber, LT and Carvalho, JC. Doctorate scholarship from the CAPES Foundation, Ministry of Education, Brazil to Hauschild DB.


S82—One Third of PICU Patients Do Not Meet Technical Criteria for Indirect Calorimetry Megan Beggs, MSC, RD1; Gonzalo Garcia Guerra, MSc, MD 2; Bodil Larsen, PhD, RD1 1Nutrition Services, Stollery Children's Hospital, Edmonton, Canada; 2Pediatric Critical Care, Stollery Children's Hospital, Edmonton, Canada Purpose: Indirect calorimetry (IC) is considered gold standard for assessing energy needs of critically ill children as predictive equations and clinical status indicators are unreliable. Accurate assessment of energy requirements in this vulnerable population is essential given the high risk of over or underfeeding and the consequences thereof. It is unknown how many patients during critical illness meet technical criteria required to perform IC measurement. In the current study, we aimed to determine the daily proportion of pediatric intensive care unit (PICU) patients who met technical criteria for IC. Methods: Prospective, observational, single-center study conducted in a cardiac and medical/surgical PICU (1000 admissions/year). All patients admitted for more than 3 days were included in analysis. 288 patients were included for a total of 3590 patient days. Variables collected for each patient included age at admission, admission diagnosis, and if technical criteria for IC was met. These technical criteria included provision of supplemental oxygen, ventilator settings, endotracheal tube leak, diagnosis of physiologic air leak, provision of external gas support (i.e. nitric oxide), and provision of extracorporeal membrane oxygenation (ECMO). Descriptive statistics were used to analyze variables. The Mann Whitney test or Kruskal-Wallis test were used to compare continuous variables and Chi-square test was used for categorical variables. P<0.05 was considered statistically significant. Analysis was performed using Prism version 5.0 (GraphPad Software, Inc.). Results: The median patient age was 0.7years (interquartile range 0.3-4.6 years). The median length of PICU stay was 7 days (interquartile range 5-14 days). Only 34% of patient days met technical criteria for IC (Table 1). Moreover, 27% of patients did not meet technical criteria for IC on any day of their PICU stay. Reasons for why IC could not be performed are illustrated in Figure 1. Conclusions: In the current study, technical criteria to perform IC were not met for 27% of patients and were not met on 63% of patient days. This data represents a major gap in the feasibility of current recommendations for assessing energy requirements of this population. Future studies are needed to improve methods of predicting and measuring energy requirements in critically ill children who do not meet current criteria for indirect calorimetry. Financial Support received from: Not applicable.


Pediatric and Neonatal Posters S83–S118

S83—Variability in Pediatric Parenteral Nutrition Prescribing While Utilizing CPOE and Standard Dose Prompts Collin Anderson, PharmD, PhD, BCPS1; Mark MacKay, RPh1; Jared Olson, PharmD, BCPS1; Jake Luke, PharmD1; Susan Sorenson, PharmD1; Sabrina Boehme, PharmD, BCPS1 1Pharmacy, Primary Children's Hospital, Salt Lake City, UT Purpose: Pediatric parenteral nutrition (PN) requirements vary widely based on age and disease state. Published recommendations regarding macronutrient, micronutrient, and electrolyte requirements are available. A computerized physician order entry system (CPOE), with dose prompts, is utilized at our free standing pediatric hospital. The purpose of this quality improvement study is to review all parenteral nutrition orders from January 2007-August 2015 and assess the variability and compliance to recommended standard ranges for electrolyte and macronutrient components. Methods: All parenteral nutrition orders from January 2007-August 2015 were obtained from the institution’s electronic data warehouse. Standard electrolyte ranges for our institution were defined as follows: sodium 2-4 mEq/kg, potassium 2-3 mEq/kg, calcium 0.3-3 mEq/kg, phosphorus 0.3-2 mmol/kg, magnesium 0.3-0.5 mEq/kg. The data were retrospectively analyzed to determine the percentage of PN orders with electrolyte components outside of each standard dose range. Median osmolarity of PN for peripheral and central lines were calculated. Average dose of each macronutrient was calculated. Medication error events for PN were evaluated over the specified time frame. Results: A total of 108,587 PN orders were evaluated. Of the electrolytes, sodium and potassium were within the institutional recommended standard dose range for approximately 60% of the orders. Magnesium and calcium had approximately 80% of the orders within the standard range, while phosphorus was ordered within range for nearly 90% of the orders (Figure 1). Median osmolarity for PN ordered for central administration was 1355 mOsm/L (Q1 1103 mOsm/L, Q3 1661 mOsm/L). Peripherally administered PN had a median of 937 mOsm/L (Q1 863 mOsm/L, Q3 980 mOsm/L). Extravasation was the most reported event causing temporary harm, with nine events reported. Average dextrose dose was 14.6 ± 3.7 g/kg for patients ≤2.5kg, 14.2 ± 4.0 g/kg for patients >2.5kg-10kg, and 8.8 ±4.1 g/kg for patients >10kg. Amino acids were dosed at an average of 3.2 ± 0.5 g/kg in patients ≤2.5kg, 2.8 ± 0.7 g/kg in patients >2.5kg-10kg, and 1.6 ± 0.6 g/kg in patients >10kg. Average lipid dose was 2.3 ± 0.7 g/kg for patients ≤2.5kg, 2.2 ± 0.8 g/kg in patients >2.5kg-10kg, and 1.3 ± 0.6 g/kg in patients >10kg. Conclusions: A pediatric PN CPOE must take into account the differing nutritional needs of patients based on age. Dosing prompts related to macronutrients are used in the CPOE system to guide practitioners. The retrospective quality analysis demonstrates that dextrose, amino acids, and lipids ordered have on average been within the recommended range, while still allowing the prescriber variability to adjust according to disease state and specific patient need. The sodium and potassium components of PN orders varied widely. There were a greater percentage of sodium orders above the institutional standard range than those below the standard range. The CPOE system provides prompts when a patient is dosed outside a standard range. Given the high percentage of sodium orders above the standard range, and the desire to limit unnecessary warnings which lead to alert fatigue, consideration is being given to expanding the institutional standard range for sodium within the CPOE system. Although potassium had a high percentage of orders outside of standard, the institutional range will not be changed due to the higher risk associated with potassium administration. All other electrolytes were primarily in range. Financial Support received from: None


S84—Pediatric Malnutrition: Multidisciplinary Approach for Screening, Assessment and Documentation at a Large Academic Pediatric Center Dina Al-Zubeidi, MD1; Douglas Robertson, RDN, LD2; Jean Ryan, RDN, CSP, LD, CNSC2; Mary Kay Brooks, RN, MSN, CPHQ3; Bridget Drapeaux, MA, RDN, LD2; Kilee Dunham, RDN, CSP, LD, CNSC2 1Pediatrics, Gastroenterology, University of Iowa Hospitals and Clinics, Iowa City, IA, USA; 2Food and Nutrition Services, University of Iowa Hospitals and Clinics, Iowa City, IA, USA; 3Clinical Quality, Safety, and Performance Improvement Office, University of Iowa Hospitals and Clinics, Iowa City, IA, USA Purpose: The risk of malnutrition in hospitalized pediatric patients is a real risk. Pediatric literature suggests a prevalence of up to 35%. Upon review of our documented diagnoses during hospitalization at our center, the documented prevalence of malnutrition was unexpectedly low at < 1.1% for the last 2 years. Our pilot study aimed to develop screening and assessment tools to more accurately identify malnourished patients by utilizing the 2015 Consensus Statement of the Academy of Nutrition and Dietetics/American Society for Parenteral and Enteral Nutrition. Furthermore, we implemented an educational program with materials designed to enhance early identification, improved documentation of diagnosis and increase provider awareness. Methods: Patient population: Five pediatric services were selected for the pilot study. The services selected were pediatric intensive care, pediatric cardiology, pediatric gastroenterology, pediatric hematology oncology and pediatric surgery. Education: Educational material for nurses and medical providers were designed and distributed before the pilot kickoff with follow up formal educational sessions scheduled 6 weeks after the pilot started. Screening tools: Nurses and registered dietitians identified and evaluated patients on admission and during hospitalization utilizing the 2015 Consensus Statement of the Academy of Nutrition and Dietetics/American Society for Parenteral and Enteral Nutrition. Electronic medical records (EMR): Malnutrition committee worked with EMR specialist to develop a dietitian malnutrition assessment tool and an electronic smart phrase to be inserted in all inpatient progress notes that automatically pulls in dietitian’s assessment under the objective section for providers. Documentation of diagnosis and incorporating intervention in the plan section remained the provider’s


responsibility. Documentation optimization team followed up with providers for more accurate documentation. Results: We systematically identified 172 patients to be at nutritional risk during the pilot period (June- August 2015). All 172 patients underwent nutrition assessment by a registered dietitian. Further assessment revealed no nutritional concerns in 42 patients (24%). Of the 172 patients assessed 59 patients (34%) where identified to be at risk for malnutrition but were not malnourished. Twenty nine Patients (17%) with mild (non-severe) malnutrition, 21 patients (12%) with moderate (non-severe) malnutrition and 21 patients (12%) with severe protein calorie malnutrition. A preliminary review of 70 pilot patients diagnosed with malnutrition showed potential increases in expected length of stay (LOS) days of 23.4 and inpatient reimbursement of $70,982. Patients with a malnutrition diagnosis that discharged in the first month of the pilot had a 23.6% reduced LOS index compared to patients with no malnutrition diagnosis. Conclusions: Malnutrition is under- recognized in hospitalized pediatric patients. In our pilot, 42% of patients assessed by a dietitian were malnourished. Nursing screening tool developed for nutrition risk and dietitian assessment tool for malnutrition are useful for early patient identification and prevention of diagnosis being delayed. Additional follow up screening with dietitians using more advanced tools e.g. weight trends and diet histories along with direct consultation from physicians ensures no patients are overlooked. Provider education further facilitates accurate documentation of diagnosis and intervention needed. Improving documentation by specifying type of malnutrition and treatment results in more accurate reportable hospital outcome measures (Severity of Illness, Risk of Mortality, and expected LOS) and hospital reimbursement. Financial Support received from: None

S85—Implementation of Nutrition Assessments for Pediatric Neuromuscular Scoliosis Patients Prior to Spine Surgery: A Quality Improvement Project Elizabeth Spoede, MS, RD, CSP, LD1; Lucille Mullins, MA, RD, CSP, LD1; Nicki Canada, MS, RD, LD, CNSC1; Starla Garcia, MEd2 1Texas Children's Hospital, Houston, TX, USA; 2University of Houston, Houston, TX, USA Purpose: Pediatric patients with neuromuscular scoliosis are at high risk for post-operative complications following spinal surgery. Due to the complex nature of this patient population and surgery, Texas Children's Hospital developed a multi-disciplinary neuromuscular spine surgery review committee to develop a spine surgery care plan. The team included, but is not limited to, orthopedic surgery, neurology, anesthesiology, safety experts, nursing, neurology, critical care, pulmonary, physical medicine, nutrition, blood bank, quality and outcomes measures, and spine surgery nurse coordinators. The team identified a comprehensive list of issues leading to sub-optimal outcomes which was analyzed by the Quality and Safety Department. Malnutrition was identified as one of the issues resulting in sub-optimal outcomes following neuromuscular scoliosis spine surgery. Methods: A quality improvement (QI) project was initiated in 2014 for neuromuscular patients being considered for spine surgery which consists of several evaluations including a nutrition assessment by a Registered Dietitian. A retrospective analysis of all patients with neuromuscular scoliosis undergoing general anesthesia for spine surgery from January 2012 through June 2015 was completed to determine if patients are receiving a nutrition assessment within 6 months prior to the procedure after implementation of the QI project. The analysis also included all nutrition assessments, even if outside of the 6 month preoperative period. Results: Eighty eight surgeries were included in the analysis. Nutrition assessments were completed 6 months prior to the surgery in 38% of all surgeries in the entire 2012-2015 time frame. Prior to initiation of the QI project, nutrition assessments were completed at a rate of 11% in 2012 and 9% in 2013. After implementation of the QI project, nutrition assessments were completed at a rate of 53% in 2014 and 100% in 2015. Additionally, prior to implementation of the QI project, 78% of the patients in this analysis had never had a nutrition assessment documented in the electronic medical record. After implementation this decreased to 37% in 2014 and 0% in 2015. Conclusions: A systematic, multidisciplinary team approach to implement a surgery care plan was implemented for pediatric patients being evaluated for neuromuscular spine surgery. Since implementation of the QI project in 2014, the percent of patients obtaining a nutrition assessments within 6 months of the procedure has improved. The most recent data indicates that 100% of the targeted population are obtaining a nutrition assessment indicating success implementation of the QI project. Next steps for the QI project include evaluating the population’s nutritional status and identifying nutritional markers that may identify patients at high risk for post-operative complications to further improve patient outcomes.



S86—Cycling Total Parenteral Nutrition in a Neonatal Surgical Patient: An Argument for Increased Utilization Katherine W. Gonzalez, MD1; Katrina L. Weaver, MD1; Deborah J. Biondo, MS, RD, LD, CNSC2; Joel Lim, MD3; Richard J. Hendrickson, MD1 1Department of Pediatric Surgery, Children's Mercy Hospital, Kansas City, MO, USA; 2Department of Nutrition, Children's Mercy Hospital, Kansas City, MO, USA; 3Department of Gastroenterology, Children's Mercy Hospital, Kansas City, MO, USA Purpose: Cycling of total parenteral nutrition (TPN) has been increasingly utilized in adult and pediatric patients with intestinal failure to limit hepatic dysfunction and improve quality of life. However, cycled TPN is infrequently utilized in the neonate population due to immature glucose homeostasis and concern for hypoglycemia. Here we present a successful case of cycled TPN in a neonate with complicated gastroschisis. Description: Patient is a Caucasian male born at 36 weeks gestation to a G2P1 mother with prenatally diagnosed gastroschisis. Following delivery he was taken to the operating room for surgical closure and found to have multiple atretic intestinal segments. He was transferred to a tertiary children’s hospital for further care. The patient required 3 additional surgeries: resection of intestinal atresias with ileostomy, stricture resection and ileostomy reversal, and anastomotic revision. Prior to the second procedure, the patient developed increasing liver dysfunction from TPN cholestasis, with a high direct bilirubin of 3.4, and was transitioned to cycled TPN. Results: The patient received TPN for 22 hours with ½ NS or NS fluids to keep the central line patent at 2 ml/hr for the remaining 2 hours. The TPN was tapered to facilitate a gradual change in the patient’s glycemic status. Each day at 4 pm the TPN was initiated at ¼ goal rate, at 5 pm this was increased to ½ goal rate. From 6 pm to 12 pm the following day, the TPN was run at goal. From 12 to 1 pm it was then decreased to ½ goal rate again. From 1-2 pm, ¼ goal rate; and from 2-4 pm the TPN was discontinued (Figure 1). Rates were calculated based on the total volume of the daily TPN. Upon admission the pt was initially started on fat emulsion 20% at 1 g/kg/d. This was incrementally advanced up to 3 g/kg/d. However, when the patient developed liver dysfunction, his lipid regimen was again decreased to 1 g/kg/d where it remained for the duration of his therapy. The glucose infusion rate ranged from 11.1 mg/kg/min to 16.3 mg/kg/min and was adjusted in order to maintain adequate caloric intake. The amino acid concentration (Premasol) was kept at 4 g/kg/d to facilitate postoperative healing. Due to the high gastric output during periods of intestinal obstruction, the patient required elevated chloride concentrations (up to 11 mEq/kg/d) with minimal acetate. Similarly the sodium concentration was increased up to 10 mEq/kg/d. With the return of bowel function, decrease in gastric output, and initiation of enteral feeds, these concentrations were decreased to expected ranges. The patient demonstrated expected fluctuations in his weight associated with each surgical procedure. However, overall he maintained constant growth at just under the 3rd percentile (Figure 2).In regards to hypoglycemia, the patient had only two incidents where his glucose level was <50 mg/dL. Each of these occurred in a time of acute illness due to a proximal obstruction. Upon initiating cycling, the patient’s liver dysfunction improved and ultimately returned to normal limits. He did not require therapy with Phenobarbital or Actigall. Conclusion: Tapered cycling TPN can be safe and effective in the neonatal population at providing nutrition while limiting hepatic complications. Financial Support received from: None


Figure 1. Cycled TPN regimen.

Figure 2. Growth chart.


S87—Infant Formula: Minimum Levels of Docosahexaenoic acid (DHA) and Arachidonic acid (AA) in Infant Formula Required to Make a Claim for “Normal Physical Brain and Eye Development” Kevin C. Wood, MSc1; Mona Eskander, MD1; Daniel Ko, MD1 1Bureau of Nutritional Sciences, Health Canada, Ottawa, Canada Purpose: Docosahexaenoic acid (DHA) and arachidonic acid (AA) are the major n-3 and n-6 long chain polyunsaturated fatty acids (LCPUFAs) in the membrane structural lipid of the brain and retina. DHA and AA can be endogenously synthesized from their essential fatty acid precursors alpha-linolenic acid (ALA) and linoleic acid (LA) respectively (see Figure 1), but the conversions are limited in infants. DHA accumulates predominantly in the last trimester of pregnancy and during the first 2 years of life. After birth, human milk (HM), or infant formula (IF) containing DHA and AA, are usually the only exogenous sources of DHA and AA for the infant. Historically, IF contained only the precursor fatty acids, LA and ALA. The Codex Alimentarius’ Infant Formula Standard (Codex Stan 72-1981, Revised 2007) allows for the optional addition of DHA to IF, at a guidance upper level (GUL) of 0.5 % of total fatty acids (% FA), and states that when DHA is added, AA should reach at least the same concentration. However, due to the optional use of LCPUFAs in IF, manufacturers have added DHA and AA to some of their IFs and are interested in making claims related to their role in brain and eye development. The Canadian Food and Drug Regulations allow for the addition of any nutritive substance to IF when that substance is added at an equivalent level to that found in HM. Furthermore, as per Canada’s Food and Drugs Act, information on product labels must be truthful and not misleading. As HM is considered optimal for feeding infants, the LCPUFA needs of infants can be estimated from the average percentage of DHA and AA in HM. This data was used to estimate the minimum levels of DHA and AA supplementation in IF to support the claim: “DHA, an omega-3 fatty acid and AA, an omega-6 fatty acid, support the normal physical development of the brain and eyes”. Methods: Health Canada evaluated current literature to determine the minimum levels of DHA and AA for addition to IFs to support a claim for physical brain and eye development. Recommendations related to DHA and AA requirements for IF from international organizations and some other governments were tabulated (see Table 1). Results: Based on the studies and data evaluated by Health Canada, DHA can be considered a conditionally essential fatty acid for infants (EFSA 2014 and FAO 2010). The worldwide mean (± SD) concentration of DHA in HM (expressed as weight percentages of total fatty acids) is 0.32% ± 0.22% and that of AA is 0.47% ± 0.13% (Brenna et al., 2007). An FAO expert panel has recommended adequate intakes of LCPUFAs to be 0.2-0.36% of total fatty acids (% FA) for DHA, and 0.4-0.6% FA for AA (FAO 2010). The most recent data concerning Canadian HM content of LCPUFAs (Ratnayake et al., 2014) found that the mean % FA was 0.3% (± 0.2 SD) for DHA and 0.4% (± 0.1 SD) for AA in HM samples collected in 2011. Conclusions: Therefore, Canadian HM samples contain amounts of DHA and AA which are similar to worldwide levels and similar to the recommendations of international organizations and some other governments. This information can be used to guide appropriate supplementation of IF. Financial Support received from: None


Table 1. Summary of recommended levels of DHA and AA to add to infant formula from international organizations. DHA (% FA*) AA (% FA*)

Organization Minimum Maximum Minimum Maximum

EFSA 20141 0.36 – 0.49%a 0.9 – 1.23%b NS NS

FSANZ Standard 2.9.1 (2014)2 NS 1% (maximum 1% for all n-3 FAs)

NS 1% (maximum 2% for all n-6 FAs)

FAO 20103 0.2% 0.36% 0.4% 0.6%

USFDA 20154 NS NS NS NS World Association of Perinatal Medicine, Early Nutrition Academy, and Child Health Foundation 2008

0.2% 0.5% 0.2% 0.5%

Codex Stan 72-1981 Infant Formula (Revised 2007)5

NS 0.5% GUL NS NS

European Commission Directive 2006/141/EC 2006

NS 1% of total fat (maximum 1% of total

fat for all n-3 FAs)c

NS 1% of total fat (maximum 2% of total

fat for all n-6 FAs)

ESPGHAN 20056 NS 0.5% of total fat NS NSd

NS = Not specified, GUL = Guidance Upper Limit *% FA = % of total fatty acids unless otherwise indicated. In all cases, EPA content must not exceed DHA content. DHA = Docasahexaenoic Acid AA = Arachidonic Acid 1 European Food Safety Authority (EFSA) 2014. Scientific Opinion on the essential composition of infant and follow-on formulae. EFSA Panel on Dietetic Products, Nutrition and Allergies (NDA). 2 Food Standards Australia New Zealand (FSANZ) Standard 2.9.1 – Infant Formula Products. Revised 2014. 3 Food and Agriculture Organization of the United Nations. Fats and fatty acids in human nutrition. 2010. 4 United States Food and Drug Administration. Section 350a Infant Formulas. Revised 2010. 5 Codex Alimentarius Standard for Infant Formula and Formulas for Special Medical Purposes intended for infants. 6 Global Standard for the Composition of Infant Formula: Recommendations of a European Society for Paediatric Gastroenetrology, Hepatology, and Nutrition (ESPGHAN) Coordinated International Expert Group. 2005. a Based on EFSA NDA Panel 2013 Scientific Opinion on nutrient requirements and dietary intakes of infants and young children in the European Union. b Based on Brenna et al. 2009 for International Society for the Study of Fatty Acids and Lipids (ISSFAL). Alpha-Linolenic acid supplementation and conversion to n-3 long-chain polyunsaturated fatty acids in humans. c DHA content may only be declared if DHA content is >0.2% FA. d AA content should reach at least the same concentration as DHA if DHA is added to the infant formula.


S88—Factors Influencing Adequacy and Achievement of Parenteral Nutrition Prescription on Growth in Neonatal Intensive Care Unit Infants at Rush University Medical Center Kirstie Ducharme-Smith, MS, RD1; Celina M. Scala, MS, RD, LDN, CNSC1; Yimin Chen, MS, RD, LDN, CNSC1 1Clinical Nutrition, Rush University Medical Center, Chicago, IL, USA Purpose: The importance of administering adequate parenteral nutrition (PN) prescription to achieve goal growth in infants in the neonatal intensive care unit (NICU) is well recognized. Factors impeding the administration of goal PN need to be established to improve PN prescription and protocol. The objective of this quality improvement study is to review the adequacy of PN prescription and its impact on growth in NICU infants. Methods: NICU infants who received PN from the years 2011 through 2014 at Rush University Medical Center (RUMC) were included. Electronic medical records were used to collect PN prescription and laboratory values. Achievement of goal PN prescription was operationalized as achievement of goal kilocalories (90 kcal/kg; yes/no), protein (3.5 gm/kg; yes/no), and lipids (3 gm/kg; yes /no). Growth was calculated as grams/day (weight) and centimeters (cm)/week (head circumference and length). Laboratory values were classified as hyperglycemia (>180 mg/dL; yes/no) and direct hyperbilirubinemia (>2mg/dL; yes/no). Current RUMC NICU protocol involves daily trace element provision in PN to meet chromium needs (when available) and administration of Ursodiol (hydrophilic bile acid) /Lovaza (enteral omega-3 lipid supplement) post development of direct hyperbilirubinemia (when receiving minimum of 50 ml/kg of enteral feeding). Implementation of RUMC protocol was determined by presence of trace element shortage (yes/no) and administration of Urodiol/Lovaza (yes/no) when appropriate. Independent t tests were performed to determine a difference in growth parameters between infants who achieved goal PN prescription and those who did not. Descriptive statistics were used to describe laboratory values and achievement of RUMC protocol of/in infants who did not achieve goal PN prescription. Results: A total of 666 NICU infants with a mean gestational age of 31 weeks and birth weight of 1550 grams were included. The average percentage of infants who achieved goal kilocalories, protein, and lipids from the years 2011-2014 was 64%, 75%, and 77%, respectively. Infants who achieved goal macronutrients showed significantly higher growth parameters for weight, head circumference, and length (Table 1). Of the infants who did not achieve goal kilocalories, 21% were hyperglycemic. Of the infants who did not achieve goal lipids, 17% had direct hyperbilirubinemia (Table 2). During periods of trace element shortages 23% of infants were hyperglycemic, compared with 0% during periods of no shortage (Table 3). Administration of Ursodiol/Lovaza occurred 46% of the time in infants with direct hyperbilirubinemia. Conclusions: Achievement of goal PN prescription (kilocalories, protein, lipids) resulted in statistically higher growth parameters (weight, HC, length) in this sample of NICU infants. Factors that prohibited the achievement of goal PN prescription included presence of hyperglycemia and direct hyperbilirubinemia, which subsequently resulted in reduced growth. During trace element shortages, a higher percentage of infants were hyperglycemic. National shortages affect RUMC inhibiting achievement of goal PN prescription. Therefore, advocacy to prevent further shortages as well as conservative management of effected components are crucial. Although administration of Ursodiol/Lovaza was not routinely delivered post development of direct hyperbilirubinemia per protocol, more research is needed to determine the effects of Ursodiol/Lovaza administration on improvement in cholestasis. Financial Support received from: None


S89—Lower Vitamin D Status Is Associated With Hypertension and Abdominal Obesity but Not Mental Health Comorbidities in Pediatric Obesity Krista M. MacDonald, BSc1; Andrea M. Haqq, MD2; Kristine Godziuk, MA3; Rena LaFrance, MD3; Mohammad Ansarian, MD3; Jason Yap, MD2; Diana Mager, PhD, MSc, RD,1,2 1Agricultural, Food and Nutritional Science, University of Alberta, Edmonton, Canada; 2Department of Pediatrics, University of Alberta, Edmonton, Canada; 3Pediatric Centre for Weight & Health (PCWH), Misericordia Hospital, Edmonton, Canada Purpose: Metabolic dysregulation, suboptimal vitamin D status and mental health co-morbidities are commonly observed in childhood obesity. The study purpose was to describe associations between serum 25-hydroxy vitamin D (25(OH)D) levels and anthropometric, cardiometabolic, liver and mental health parameters in obese children. Methods: A retrospective chart review was conducted in (n=100) children aged 2-18 years referred to the Pediatric Centre for Weight and Health (PCWH) at the Misericordia Hospital in Edmonton, Alberta. The following variables were assessed: anthropometric (weight, height, body mass index (BMI), BMI standard deviation score (SDS), waist circumference (WC)), cardiometabolic (systolic blood pressure (SBP), diastolic blood pressure (DBP), fasting blood glucose (FBG), insulin, homeostasis model assessment for insulin resistance (HOMA-IR), triglyceride, high density lipoprotein (HDL), low density lipoprotein (LDL), total cholesterol, liver function (alanine aminotransferase (ALT), gamma-glutamyl transferase (GGT)), dietary intake and mental health diagnosis (number, diagnosis) using validated methodologies. Vitamin D status was defined as follows: insufficient 25(OH)D <50nmol/L, suboptimal 25(OH)D 50-75 nmol/L and sufficient 25(OH)D >75 nmol/L. ALT levels >20 U/L were considered abnormal. Results: Children (12.0 ± 3.1 yrs; 55M/45F) had a mean BMI-SDS of 2.2 ± 0.5 and WC of 99.1 ± 15.7 cm (48% >100cm). The prevalence of vitamin D insufficiency (<50 nmol/L) was 23% (mean: 63 ± 20 nmol/L). Patients had a mean GGT of 18 ± 11 U/L and a mean ALT of 30 ± 20 U/L (64% ALT >20U/L). For SBP, n=12 (12.4%) patients had pre-hypertension (HTN), n=19 (19.6%) had stage 1 HTN and n=7 (7.2%) had stage 2 HTN. For DBP, n=4 (4.1%) patients had pre-HTN and n=4 (4.1%) had stage 1 HTN. The prevalence of anxiety, attention deficit hyperactivity disorder (ADHD) and mood disorders was 20% (n=20), 10% (n=10) and 10% (n=10) respectively. Age (p=0.03), height (p<0.01), weight (p<0.01), BMI (p<0.01), WC (p=0.02), SBP (p<0.01), insulin (p=0.01) and HOMA-IR (p=0.02) showed weak but negative correlations with 25(OH)D. WC >100cm (approximate mean) was associated with lower 25(OH)D levels (58 ± 19 nmol/L vs. 68 ± 17 nmol/L, p=0.02). Higher vitamin D status (25(OH)D>50nmol/L) was associated with lower SBP (115 ± 13 mmHg vs 124 ± 10 mmHg, p<0.01) and lower DBP (64 ± 8 mmHg vs. 68 ± 9 mmHg, p=0.04). History of weight gain (≥5%; mean 156 ± 84 days) did not influence vitamin D status. No significant associations between vitamin D and mental diagnoses were observed. Conclusions: Overweight and obese individuals attending a specialized pediatric weight management clinic had reduced vitamin D levels, cardiometabolic and liver dysfunction. Lower vitamin D levels were observed in those with greater abdominal obesity and were associated with higher levels of SBP and DBP. Financial Support received from: G. Woodrow Wirtanen Studentship Alberta Diabetes Institute


S90—Medication Reconciliation of Compounded Home Parenteral Nutrition Solutions—A Quality Improvement Initiative at a Pediatric Teaching Hospital Margaret K. Murphy, PharmD, BCNSP1,2; Kathleen M. Gura, PharmD, BCNSP, FASHP,FPPAG,FASPEN1,2; Glendalis Grullon, RN, MSN, CPNP2; Bram Raphael, MD2,3 1Pharmacy, Boston Children's Hospital, Boston, MA; 2Home Parenteral Nutrition Program, Boston Children's Hospital, Boston, MA; 3Division of GI, Hepatology and Nutrition, Boston Children's Hospital, Boston, MA Purpose: Parenteral nutrition (PN) is a complex intravenous admixture that may contain over 25 different components. This high-risk therapy is often prescribed for use in the home setting. The number of pediatric patients utilizing PN at home is growing as new treatments have improved survival rates for children with serious digestive disorders. The use of medication reconciliation at all points of care has been shown to be an effective measure for preventing adverse drug events (ADEs). The Home Parenteral Nutrition (HPN) program at Boston Children’s Hospital (BCH) is a multi-disciplinary program involving clinical pharmacists specialized in nutrition support that follows more than 100 patients receiving this life-sustaining therapy. During outpatient encounters a clinical pharmacist routinely reconciles compounding records of HPN solutions to identify potential discrepancies. Methods: Between January 2013 and July 2015, as part of a quality improvement initiative, BCH HPN pharmacists reviewed individual patient’s commercial HPN compounding records prior to each of their clinic visits. Each component of the compounded HPN was calculated by the pharmacist and reconciled with the BCH prescription in order to assess the accuracy of the commercially compounded solution. Any discrepancies were reported to the team. A discrepancy was defined as any difference in formulation of the compounded solution from the prescription. Appropriate interventions were determined by the team based on the nature of the discrepancy and the patient’s clinical status. The commercial compounding facility was also contacted to discuss and resolve the discrepancy. The total number of prescriptions reviewed as well as the number of HPN solutions found to have one or more discrepancies was calculated each month (Figure 1). Results: A total of 1116 commercial HPN compounding records were reconciled with the BCH HPN prescription by a pharmacist in conjunction with a CAIR or HPN clinic visit. Of these, 294 records (26%) were found to have at least one discrepancy. Types of discrepancies found were: total volume of HPN, dextrose amount, electrolyte amount, anion balance, trace element amount and amino acid brand substitution. This process has resulted in a trend toward decreased discrepancies. Conclusions: Commercially compounded HPN prescriptions do not always contain the exact ingredient amounts prescribed. Reconciliation of compounded HPN solutions at outpatient encounters may be critical to detect discrepancies. The ability to interpret complicated HPN solutions and to communicate with commercial HPN vendors places pharmacists in a unique role among the clinical team. While HPN vendors provide an indispensable function, errors in compounding this high-risk therapy can and do occur. Further study would be useful to determine clinical implications of HPN compounding errors in the pediatric population. Financial Support received from: None


S91—Case Report: Aluminum Toxicity in a Pediatric Patient With Intestinal Failure, on Long-Term Home Parenteral Nutrition Stephen C. Adams, MS, RPh, BCNSP1 1VITALine Infusion Pharmacy Services, Geisinger Health System, Danville, PA, USA Purpose: The toxicity of parenteral Aluminum (Al) was first noted in the mid to late 70’s in dialysis patients where dialysis encephalopathy and osteomalacic dialysis osteodystrophy were associated with Al in the dialysate solutions. (1) (2) Subsequently, patients who received parenteral nutrition with casein hydrolysate were found to have substantial elevation of serum Al, increased urinary excretion of Al and very high content of Al in trabecular bone. (3) In 2004, the FDA amended labeling requirements for aluminum content in large and small volume parenterals and pharmacy bulk packages used to compound parenteral nutrition solutions. Large-volume parenterals are required to contain less than 25 mcg/L of aluminum. Small-volume parenterals are required to be labeled with the maximum aluminum content at expiration. (4) In response, the A.S.P.E.N. Aluminum Task Force released a statement on Aluminum in Parenteral Nutrition Solutions. (4) While noting that it may be impossible to provide < 4-5 mcg/kg aluminum per day, as recommended in the FDA ruling, the statement recommends that all healthcare providers involved in the provision of parenteral nutrient should ensure that aluminum exposure is limited when possible in at risk populations. Poole, et al revisited the FDA ruling in 2008, calculating aluminum exposure in pediatric patients receiving parenteral nutrition. (5) Patients weighing less than 3 kg received the largest calculated amounts of aluminum, ranging from 30-60 mcg/kg/day. The authors conclude that with the aluminum content of currently available products, it is impossible to meet the FDA rule of less than 5 mcg/kg/d of aluminum exposure in patients weighing less than 50 kg. This case report describes an incident of aluminum toxicity in a pediatric patient


with intestinal failure on long-term home parenteral nutrition. The patient is a 5 year old male with a variant of the filamin-A gene and neuropathic chronic intestinal and colonic pseudo-obstruction. Home parenteral nutrition (HPN) was initiated in July of 2010 after the patient failed to tolerate enteral feedings. The patient presented to clinic in September of 2014 with lower extremity pain with a differential diagnosis of TPN-associated aluminum toxicity versus musculoskeletal issues than can accrue in individuals with variants of the filamin-A gene. In addition to X rays of bilateral lower extremities and hips, aluminum levels were drawn and were reported as 61 mcg/L (normal range < 7 mcg/L, for dialysis patients < 40 mcg/L). A negative deferoxamine challenge was completed in October 2014. Aluminum content for the current HPN solution was calculated (Table 1) and a sample of the HPN solution was submitted for analysis of aluminum content (Table 2). HPN components were adjusted to minimize the total aluminum exposure (Table 3). Aluminum levels were repeated at monthly clinic visits. By substituting components with lower labeled Al content, total daily Al was decrease from 33.28 mcg/kg/day to 3.93 mcg/kg/day. Methods: N/A Results: N/A Conclusions: N/A Financial Support received from: none

Table 1. Calculated Aluminum Content Al mc/gL Volume (ml)/day Al (mcg)/day Current weight (kg)

Premasol 25 340 8.5 19.64 Intralipid 20% 25 130 3.25

Na Acetate 400 3 1.2 NaCl 200 4.25 0.85

K Phos 31000 4.25 131.75 Ca gluconate 9400 53.78 505.532

MgSO4 300 0.84 0.252 Peditrace ? Zinc Cl 150 1 0.15

Selenium 250 0.85 2.125 Total Al (mcg)/day 653.1

Al mcg/kg/day 33.28 Substitute for lower Al content

Al mcg/L Vol (ml)/day Al (mcg)/day Na Phos 16300 4.25 69.275

KCl 100 3.35 0.335 N Acetate 200 6 1.2

NaCl 200 3 0.6 Al (mcg)/day 77.187 Al mcg/kg/day 3.93


S92—Linear and Head Circumference Growth in Premature Neonates Receiving Fortified Human Milk With Supplemental Protein Xing Pan, MS, RD1; Sharon Foley, PhD, RD, LDN1; Aloka Patel, MD1; Yimin Chen, MS, RD, CNSC, LDN1; Celina M. Scala, MS, RD, CNSC, LDN1 1Rush University Medical Center, Chicago, IL, USA Purpose: Preterm infants are at high risk for post-natal growth failure. Providing preterm infants with insufficient protein may lead to suboptimal weight, length and head circumference (HC) growth, particularly length and head circumference. Human milk (HM) as a sole source is insufficient in meeting the increased protein needs of the preterm infants. A liquid protein supplement can be added to HM solely or to HM fortified with human milk fortifier (HMF) to facilitate meeting the increased protein requirements of infants without relying on the use a preterm formula (PF). The purpose of this study was to determine the effect of a liquid protein (LP) product added to fortified HM on weight, linear and HC growth in preterm infants. Methods: Data from a convenience sample of infants who were patients in the neonatal intensive care unit (NICU) of a large urban medical center were used after obtaining approval from the Institutional Review Board. The electronic medical records of the infants were retrospectively reviewed. Differences in linear and HC growth between preterm infants who received fortified human milk (HM) with additional LP (LP, n = 46) were matched and compared to a control group (n = 46) who received fortified HM only over a 21-day period. Historical control subjects were matched with LP subjects based on birth gestational age (GA), gender, birth weight, whether infants were fed fortified HM at the same gestational age at which the LP group started to receive LP, and the type and percentage of formula the infants received. Pair-t tests were used to analyze the differences in weight, length and HC between matched subjects. Results: There was statistically significant mean growth within each group for length (LP group: 3.7 ± 1.5 cm; Control group: 3.3 ± 1.3 cm, p < 0.001 for both) and HC (LP group: 2.9 ± 0.9 cm; Control group: 2.8 ± 1.0 cm, p < 0.001 for both). However, there was no significant mean difference in length or HC between the feeding groups. Significant differences were found in mean weight gain within each feeding group (LP group: 466.7 ± 127.5 g; Control group: 405.0 ± 127.6 g; both p < 0.001), as well as between groups (mean difference: 61.7 ± 145.9 g, p = 0.007). Conclusions: Additional protein supplementation may improve weight gain in preterm infants but this effect may not be detected with length and HC growth. The long-term clinical significance of liquid protein on preterm infant growth remains unclear. A strong prospective study design would be beneficial to help control for confounders. Future studies should be directed toward the long-term effects of protein supplementation on growth in preterm infants. Financial Support received from: None

S93—Implementation of a TPN Taper Algorithm on a Pediatric Oncology Service Jennifer Townsell, PharmD, BCOP1,3; Elizabeth Welin, MS, RD, LD2,3 1Pharmacy, Cincinnati Children's Hospital Medical Center, Cincinnati, OH, USA; 2Nutrition Therapy, Cincinnati Children's Hospital Medical Center, Cincinnati, OH, USA; 3Cancer and Blood Diseases Institute, Cincinnati Children's Hospital Medical Center, Cincinnati, OH, USA Purpose: At Cincinnati Children’s Hospital Medical Center the nursing staff on the oncology service self-identified inconsistencies within their practice of administering total parenteral nutrition (TPN). When holding TPN for administration of incompatible medications or procedures, there was no consensus on whether nursing should taper the rate of the TPN down or abruptly stop the TPN. This variable practice with a high-risk medication raised concerns for potential safety events, caused frequent wasting of a high-cost medication, and caused some patients to receive less nutrition daily than prescribed. To create a standardized approach for administration of TPN, a multidisciplinary group was formed to review the literature and create an evidence-based guideline to improve practice. Methods: A multidisciplinary group was formed to create a guideline for TPN administration on the oncology service. A dietitian, a pharmacist and nurses evaluated the literature for data regarding tapering TPN in pediatric patients. Six criteria were identified to be useful in decision making – patient age, glucose infusion rate, enteral nutrition status, insulin, anticipated time off TPN and diluent of incompatible medication to be given. These criteria were organized into a decision algorithm for nursing (Figure 1). Following implementation, data was collected to


evaluate the safety of the process change. Nursing staff was also surveyed for feedback and to evaluate impact on workflow. Results: Six patients received cyclic TPN during the study time period for a total of 44 days on TPN. One patient on cyclic TPN did not meet criteria for no tapering because of concurrent insulin therapy. Across the 44 TPN days, there were zero episodes of hypoglycemia defined as blood glucose level < 70 mg/dL. Nursing surveys indicated the guideline had a positive impact on nursing workflow and revealed opportunities for improving the electronic orders which were successfully implemented. Conclusions: The TPN taper algorithm was successfully implemented in a step-wise process on the pediatric oncology service. Targeted education was provided to staff prior to implementation of the pilot, and surveys of bedside nurses followed to evaluate for ease, effectiveness and to identify issues. Consistent communication to providers at various levels allowed this quality improvement project to be successful. Standardization of this process for administration of a high-risk medication will decrease the risk for potential safety events. Following the success of this initiative on the oncology service, this TPN taper algorithm may be used on additional units at Cincinnati Children’s Hospital Medical Center. Financial Support received from: None


S94—Incidence of Cholestasis After the Implementation of Lipid Dosing Guidelines at a Children’s Hospital Kylee Kastelic, PharmD.1,2; Beth Deen, PharmD., BCNSP2; Mary F. Lynch, MD.2; Scott Raynaud, PhD.2 1Pharmacy, Oregon Health and Science University, Portland, OR, USA; 2Pharmacy, Cook Children's Medical Center, Fort Worth, TX, USA Purpose: Current evidence suggests that length and amount of lipid exposure are risk factors for total parenteral nutrition (TPN)-associated cholestasis, which can eventually lead to parenteral nutrition associated liver disease (PNALD). Mirtallo et al. and Diamond et al. in 2010 suggested that decreasing exposure to these lipid preparations might help prevent TPN cholestasis or limit its progression. In August 2012, lipid minimization guidelines were implemented in the neonatal intensive care unit (NICU) at Cook Children’s Medical Center (CCMC), reducing the maximum daily dose from 3 g/kg to 2.5 g/kg. The objective of this study is to determine whether implementation of lipid minimizing dosing guidelines will decrease the incidence of cholestasis and direct bilirubin levels without negatively impacting growth. Methods: This institutional review board approved retrospective cohort study compared data before and after dosing guideline implementation. Patients admitted to the NICU who received TPN for 14 or more days during hospitalization stay were included. Patients were excluded if they had a preexisting liver disease as documented in the medical record by abnormal liver function tests prior to TPN exposure or had congenital cardiac anomalies requiring cardiac surgery. Patients admitted between July 1st 2010 and June 30th 2012 were screened to be in the pre-protocol group, and patients admitted between January 1st 2013 through December 30th 2014 were screened for the post-protocol group. Chi-square and Fischer’s Exact tests were performed on dichotomous data such as baseline characteristics, percent of non-protein calories, number of patients with sepsis, number of patients receiving lipids >2.5 g/kg/day, and the primary objective a direct bilirubin greater than 2 mg/dL. Wilcoxon tests were performed to determine whether there were significant changes in weight and total number of days on TPN. A linear regression was done to determine whether sepsis was a confounder for cholestasis. Results: This study included 114 patients total, with 47 in the pre-protocol group and 67 in the post-protocol group. There was no difference between groups in regard to gestational age at the start of TPN, gender, weight at the start of TPN, or the number of days on TPN. This study found 17/47 (36.2%) patients in the pre-protocol group had a direct bilirubin >2 mg/dL compared to 14/67 (20.9%) patients in the post-protocol group (P=0.0712) [figure 1]. No differences in weight were found between groups (P = 0.6378). The rates of sepsis were 11/47 (23%) vs. 8/67 (12%) in the pre and post-protocol groups, respectively (P = 0.1059), with a linear regression showing it to be a confounder (P=0.0002) [figure 2]. Conclusions: There was a reduction in the amount of cholestasis with the implementation of a lipid dosing protocol, however, it was not statistically significant and sepsis was found to be a confounder in this study. There was not a negative impact on growth with the lipid dosing reduction showing that the lipid reduction protocol is safe to use in this population. Financial Support received from: None


S95—Incidence of Low Copper and Zinc Levels in High-Risk Infants With Intestinal Failure Laura A. Romig, MBA, RD, LD1; Leonor A. Massieu, RD, CNSC, LD1; Andrea O'Donnell, RD, CNSC2; Amy Hair, MD2 1Clinical Nutrition Services, Texas Children's Hospital, Houston, TX, USA; 2Section of Neonatology, Department of Pediatrics, Baylor College of Medicine and Texas Children’s Hospital, Houston, TX, USA Purpose: High-risk infants, especially those with intestinal failure including anatomic or functional short bowel syndrome, have increased requirements for micronutrients. Current guidelines continue to recommend withholding manganese from PN in the presence of cholestasis. Previous literature recommended reduction or omission of copper in the presence of cholestasis due to concern for toxicity as intravenous copper is primarily excreted in bile. Recent data suggests that standard dosing (20 mcg/kg/day) of intravenous copper in infants with cholestasis may not significantly increase the incidence of copper toxicity or contribute to worsening liver function. Most current recommendations support the provision of copper at standard dosing with regular monitoring of serum copper levels in cholestatic infants. As a result, the provision of parenteral multi-trace element (TE) products is often reduced for patients with cholestasis receiving parenteral nutrition (PN). Generally, this is achieved by providing multi-TE products two to three times weekly rather than daily in PN. Single-entity TE products can be utilized to meet individual patient needs but this may be challenging given ongoing shortages of minerals for intravenous use. High-risk infants with intestinal failure may require additional zinc and copper above maintenance requirements due to increased gastrointestinal losses. Typically, multi-TE products available in the United States are inadequate to meet maintenance zinc requirements of high-risk infants and individual parenteral zinc products must be added to the PN formula. Methods: A retrospective analysis was completed on infants followed by the Intestinal Rehabilitation Team during admission to the Neonatal ICU from August 2013 – July 2015. Nutrition status was assessed by serum copper levels, serum zinc levels, and conjugated bilirubin levels to determine the presence of cholestasis. Copper deficiency was defined as serum copper level < 75 mcg/dL. Zinc deficiency was defined as serum zinc level < 60 mcg/dL. Cholestasis was defined as a conjugated bilirubin > 2 mg/dL during admission. Results: A total of 164 infants were included: gestational age at birth 32 2/7 ± 5 3/7 weeks, birth weight 1896 ± 1036 grams. In infants with serum copper levels (n=60) the prevalence of low serum copper levels was 76.7% (46.6 ± 16.4 mcg/dL). The prevalence of cholestasis in these patients was 84.8% with a peak conjugated bilirubin level during admission of 7.7 ± 6.5 mg/dL. In infants with serum zinc levels (n=65) the prevalence of low serum zinc levels was 15.4% (49 ± 14.2 mcg/dL). The prevalence of cholestasis in these patients was 80% with a peak conjugated bilirubin level during admission of 6.8 ± 3.2 mg/dL. Conclusions: High-risk infants with intestinal failure and a history of cholestasis are at an increased risk for copper and zinc deficiency when multi-TE products are limited. The use of individual parenteral zinc products to meet maintenance requirements when multi-TE products are limited has led to a lower incidence of zinc deficiency in infants with intestinal failure and a history of cholestasis. Copper as well as zinc may need to be proactively added back as a single-entity TE when trace minerals are reduced. A standardized approach should be developed for the routine monitoring of copper and zinc levels in infants requiring prolonged PN, especially for infants with cholestasis. Further research is needed to determine an evidenced-based approach to the provision of copper and zinc in cholestatic infants requiring PN. Financial Support received from: None

S96—Evaluation of the Incidence and Resolution of Cholestasis in Infants on Prolonged Parenteral Nutrition Lindsay M. Hubbard, PharmD1,2; Michael L. Christensen, PharmD1,2; Oscar R. Herrera, PharmD1,2; Catherine Crill, PharmD1,2 1Clinical Pharmacy, University of Tennessee Health Science Center, Memphis, TN, USA; 2Pharmacy, Le Bonheur Children's Hospital, Memphis, TN, USA Purpose: Parenteral nutrition (PN) has a significant impact on ensuring adequate growth and development among infants unable to receive enteral nutrition. However, parenteral nutrition does place the patient at risk for complications, especially with prolonged therapy. Long-term PN administration has been associated with the risk of developing cholestasis. Cholestasis can lead to serious complications including parenteral nutrition associated liver disease (PNALD), hepatic fibrosis, fulminant liver failure and even death. Treatment strategies for patients who develop cholestasis at our institution often include any combination of medications, lipid restriction, PN cycling, and


enteral feedings as tolerated. The objective of this study was to determine the incidence of cholestasis at our institution and identify patient characteristics associated with development and resolution of cholestasis in infants on prolonged PN. Methods: This was a retrospective chart review conducted on all infants receiving parenteral nutrition at Le Bonheur Children’s Hospital between January 1, 2012-December 31, 2014. To be included, patients had to be less than 12 months of age and have received parenteral nutrition for at least 30 days. Patients aged greater than 12 months or who received PN for less than 30 days were excluded. The primary endpoint was the incidence of cholestasis, defined as direct bilirubin greater than or equal to 2 mg/dL, in infants on prolonged PN therapy. Clinical factors leading to onset and resolution of cholestasis were assessed as secondary endpoints. Other data extracted included demographics, indication for PN, length of PN therapy and enteral feeding initiation. Results: During the study period, 81 patients met inclusion criteria. Fifty-one percent of patients developed cholestasis compared to 48.2% who did not develop cholestasis. Patients with lower gestational age, lower birth weight and sepsis diagnosis were at increased risk for developing cholestasis (p <0.05). Infants that developed cholestasis had significantly longer duration of PN and had PN discontinued at an older age compared to infants that did not develop cholestasis (p<0.05). Other clinical factors including length of stay, type of PN (individualized vs. standardized), age PN initiated, NPO duration, age enteral feeding initiated, PN cycling and medications were not significantly different between the two groups. For patients that achieved resolution of their cholestasis, no significant differences in treatment strategies and clinical characteristics were observed. Conclusions: For infants on parenteral nutrition for at least 30 consecutive days, more than half developed cholestasis at our institution. Clinical factors associated with an increased risk of cholestasis include lower gestational age, lower birth weight, sepsis and longer duration of PN. As observed during the study period, considerable variation exists among treatment strategies that could influence resolution of cholestasis. Financial Support received from: None

S97—Aluminum Content of Neonatal Parenteral Nutrition Solutions Containing Calcium Gluconate in Plastic Vials and Glycerol Phosphate With and Without Added Cysteine Robert K. Huston, MD1; Carl F. Heisel, RPh, BCPS2; Benjamin R. Vermillion, BSc2 1Neonatology, Northwest Newborn Specialists PC and Pediatrix Medical Group, Portland, OR, USA; 2Neonatal Pharmacy, Randall Children's Hospital at Legacy Emanuel, Portland, OR, USA Purpose: Calcium chloride (CaCl2) has been the only calcium additive available in North America that, when added to neonatal parenteral nutrition (PN), limits aluminum (Al) exposure to near recommended levels. Calcium gluconate in glass vials (CaGlu-Gl) has been shown to be a primary source of Al contamination in neonatal PN. Solutions made with calcium gluconate in plastic vials (CaGlu-Pl) and sodium glycerol phosphate (NaGP) have been shown to be low in Al content but these additives have not previously been available in North America (Frey OM and Maier L, Ann Pharmacotherapy, 2000). The purpose of this study was to measure Al concentrations in representative neonatal parenteral nutrition (PN) solutions containing CaGlu-Pl and NaGP with and without added cysteine (Cys). Methods: Solutions of Trophamine® (Braun Medical Inc, Irvine, CA) containing 3% amino acids (AA) were compounded with 2 mmol/dL (80 mg/dL) of elemental calcium as CaGlu-Pl and 2 mmol/dL of NaGP. TPN solutions contained standard neonatal dosages of dextrose, electrolytes, magnesium, copper, zinc, selenium, multivitamins, heparin, and levocarnitine. Seven solutions containing CaGlu-Pl and NaGP were compounded with or without cysteine, 90 mg/dL. Samples were sent to NMS Labs (Willow Grove, PA) for analysis of Al content using inductively coupled plasma/mass spectrometry. Five samples of the cysteine additive were also sent for analysis. The predicted Al content was calculated using the manufacturer’s maximum labeled Al concentration of additives. Means of the Al concentration of each study group were compared using Student’s t tests after entering data into an Excel spreadsheet (Microsoft Corp, Redmond WA). Results: Results (N=7 for each group) are shown in the Table. There is no significant difference in the measured Al content between solutions containing cysteine versus those without cysteine. The measured Al content for both groups was 1.5-1.6% of predicted. The measured Al content of the cysteine additive was 126±5 mcg/L (N=5) compared to the predicted content of 5000 mcg/L. The manufacturer’s listed Al content of CaGlu-Pl was the same as that for CaGlu-Gl (9400 mcg/L). Conclusions: This study is consistent with the results of a previous study that showed that the measured amount of Al contamination in solutions related to cysteine does not appear to be significant. It is also consistent with previous


studies that have shown that solutions containing calcium gluconate in plastic vials or calcium chloride and glycerol phosphate are low in Al content, leading to the possibility for limiting aluminum intakes from neonatal PN to levels within the FDA guideline of <=5 mcg/kg/day. The high predicted amount of Al in the cysteine additive and the CaGlu-Pl compared to actual Al content likely explain a large portion of the difference in measured versus predicted Al content of the PN solutions. Financial Support received from: A research grant from Northwest Newborn Specialists, PC

Aluminum Content of Neonatal Parenteral Nutrition Solutions Containing Calcium Gluconate in Plastic Vials and Glycerol Phosphate with and without Added Cysteine (90 mg/dL)

Group Predicted No Cys Measured No Cys Predicted Cys Measured Cys Al(mcg/dL) 84.8 1.3±0.2 93.8 1.5±0.2

S98—An Intestinal Rehabilitation Team in the Neonatal Intensive Care Unit: A Multidisciplinary Approach Andrea O'Donnell, RD1,2; Laura A. Romig, MBA, RD, LD3; Amy Hair, MD1,2 1Pediatrics, Baylor College of Medicine, Houston, TX, USA; 2Neonatology, Texas Children's Hospital, Houston, TX, USA; 3Clinical Nutrition Services, Texas Children's Hospital, Houston, TX, USA Purpose: Caring for infants with complex intestinal diseases presents many nutritional challenges. At Texas Children’s Hospital, we have established a multidisciplinary model for managing infants with feeding difficulties. This team is comprised of a neonatologist who is also the Medical Director of Nutrition, multiple neonatal dietitians, pediatric surgeons, physicians from the Gastrointestinal (GI) Department and dietitians specializing in GI diseases. As a team, we round weekly at each patient’s bedside and discuss the patient’s current nutritional status and plan of care. Recommendations communicated to the patient’s primary medical team include the following: type of feeding, route of feeding, advancement rate and schedule, need for alternative lipid emulsions, plan for upcoming surgeries and discharge needs. This team approach began in January 2012. Throughout the years, we have continued to define the role and management of our team. Methods: A retrospective chart review was conducted for patients followed by our Intestinal Rehabilitation (IR) Team. For the purposes of our review, we assigned patients to one of two groups: Group One consisted of patients admitted to our Neonatal Intensive Care Unit (NICU) and followed by our IR Team during the 18 month period from January 2012 through July 2013. Group Two consisted of patients who were admitted to our NICU and followed by our IR Team during the 18 month period from August 2013 through December 2014. This delineation was made for the purpose of determining if a structured, multidisciplinary team has improved outcomes with its evolution and improvements over time. We collected baseline data and growth specific data including z-scores for admission and discharge weight, length and head circumference and length of stay. Z-scores were obtained using the Fenton gender specific growth curve until 50 weeks post-menstrual age (PMA) and the World Health Organization (WHO) growth curves were utilized beyond 50 weeks PMA (corrected for prematurity when appropriate). Results: 162 patients met criteria for screening. 79 patients met criteria for inclusion in Group One; however, 12 patients were excluded for deaths leaving 67 patients for analysis. 83 patients met criteria for inclusion in Group Two; however, 5 patients were excluded for deaths leaving 78 patients for analysis. The average gestational age and birth weight for patients in Group One was 31 4/7 +/- 7 3/7 weeks and 1718 +/- 982 grams respectively and for patients in Group Two was 32 4/7 weeks +/- 5 weeks and 1868 +/- 944 grams. Primary diagnoses were similar between the two groups (Table 1). The average length of stay in Group One was 127 +/- 81 days and in Group Two was 98 +/- 73 days (p=0.028). There were no significant differences in growth velocity or z-scores between the groups. Conclusions: A multidisciplinary model for patients with complex intestinal diseases in the NICU is a novel approach to nutritional management. It provides a framework for the coordination of care both for inpatient management and in preparation for discharge. In our analysis, a decrease in mortality was observed as well as a significantly decreased length of stay. It was not within the scope of this research to evaluate the benefits of an earlier discharge; however, they are many including both financial and family-centered considerations. Financial Support received from: None


Table 1. Primary Diagnosis of Intestinal Rehabilitation Team Patients Group One Group Two

Gastroschisis 16 (24%) 24 (31%) Omphalocele 1 (1%) 6 (8%)

NEC 12 (18%) 11 (14%) Perforation 8 (12%) 16 (21%)

PDA 5 (7%) 1 (1%) Obstruction/Atresia/Volvulus 8 (12%) 9 (12%)

Hirschprungs 5 (7%) 0 (0%) Feeding Intolerance Otherwise Not Specified 12 (18%) 11 (14%)

S99—Reduced Use of Total Parenteral Nutrition in a Paediatric Hematopoietic Stem Cell Transplant Population Rivanna Stuhler, RD1 1Clinical Dietetics, Sickkids, Toronto, Canada Purpose: Total Parenteral Nutrition (TPN) is a commonly used source of nutrition in the paediatric hematopoietic stem cell transplant (HSCT) population. This is as a result of the severe side effects of myoablative conditioning regimens as well as post-transplant complications that impact the ability to meet nutritional needs orally. Early nasogastric (NG) feeding tube insertion and initiation of enteral nutrition (EN) has been successful in some paediatric HSCT patients, often reducing or alleviating the need for TPN. Provision of nutrition via a nasogastric (NG) feeding tube is safer, more physiological and less expensive than PN. In our institution, PN has been the traditional first choice for nutrition support on the HSCT unit. However, with an organizational push to decrease length of stay (LOS) coupled with perceived overuse of TPN in our patient population, EN has been promoted as a superior option for first-line nutrition, with an intent to reduce the use of TPN. Methods: As TPN use was thought to be a barrier to reducing LOS, the HSCT team began to wean TPN early, either before engraftment in those children showing interest in oral intake, or at engraftment, with the hope of stimulating oral intake or making clear the need for EN support. With coaching from the HSCT Clinical Dietitian, staff physicians began to discuss the benefits of early EN in HSCT with families in the pre-admission consult, which takes place approximately one week prior to admission. The dietitian met with patients and families on admission to discuss early initiation of EN. Data from TPN Pharmacy, and the Sickkids HSCT and hospital databases were collected and compared. As complete data was only available for 18 months, and this initiative has only been in existence for 1 year, statistical significance was not expected, and was not explored. Downward linear trends were expected. Results: Complete TPN cost data was available from January 2013-June 2015. Total TPN cost for the HSCT program was compared for the first half of the calendar year for the years 2013, 2014, and 2015 (Figure 1). There was an 11% cost reduction between 2013 and 2014, and a 25% cost reduction between 2014 and 2015. The overall cost reduction between 2013 and 2015 was 33%. In order to assess change over a continuous length of time versus three discrete time periods, TPN cost from January 2013 to June 2015 was looked at in 6 month blocks. Overall, the total TPN cost decreased between 2013 and 2015 (Figure 2). Looking at cost alone is not sufficient to declare overall decreased TPN use, as the amount of money spent can be affected by the number of HSCTs as well as fluctuations in patient days. Number of TPN days per HSCT was compared (Figure 3), showing a downward trend from 2013 to 2015, with a reduction from 21 TPN days/HSCT in the first half of 2013 to 18 days in the first half of 2015. A reduction of 3 days may impact overall length of stay, and therefore be fiscally significant. TPN cost/HSCT was compared, and again showed a downward trend, with a reduction of $258.49 per HSCT between 2013 and 2105. Conclusions: This informal Quality Improvement study shows that even small changes in the approach to nutrition during hematopoietic stem cell transplant can have a positive fiscal impact. Presenting enteral nutrition as a superior first-line nutrition option to patients and families over parenteral nutrition, and ensuring that feeding tubes are inserted early allows for a reduction in TPN use and cost, and may decrease overall LOS. This benefits not only the patients, but the hospital as well. Ongoing education is crucial to maintaining a positive attitude towards EN, thus


minimizing the use of unnecessary TPN. Data collection should continue, with the hope of showing statistical significance over time. Financial Support received from: None

Cost reduction 2013-2014 = 11% ($8,527.89). Cost reduction 2014-2015 = 25% ($17,835.09). Overall cost reduction 2013-2015 = 33% ($26,362.98)


S100—Implementation of a Standardized Feeding Guideline in the NICU Setting Shannon M. Cunningham, RD, CNSC, CLEC; Vicki Liu, MS, RD, CSP; Jennifer Kim, MS, RD; Marina Vigil, RD; Arwin Valencia, MD Clinical Nutrition, Miller Children's and Women's Hospital Long Beach, Long Beach, CA Purpose: Recent evidence has shown that implementation of a standardized feeding protocol in the NICU setting can improve outcomes by 1) Reduction in time to achieve full enteral feedings (150 mL/kg/day); 2) Fewer parenteral nutrition days; 3) Reduced risk for central line associated blood stream infections; 4) Reduced risk of late-onset sepsis; 5) Reduced risk for NEC (necrotizing enterocolitis) and 6) Reduction of EUGR (extra uterine growth restriction) at discharge. Utilizing a standardized feeding guideline schedule promotes consistency with initiation and advancement rates using birth weight categories and reduces variability among practitioners. Mother's own expressed breastmilk is the first choice for feedings in most cases, followed by pasteurized donor milk (if infant meets eligibility requirements) or preterm infant formula starting within 48 hours after birth in stable infants. Methods: Feeding guidelines were developed through the Memorial Care Neonatal Best Practice Team and launched January 2013 for infants < 1000 g birth weight, then expanded to infants < 2500 g birth weight beginning January 2014 (Figure 1). An algorithm for management of gastric residuals was also adopted to be utilized in conjunction with the guidelines. Laminated copies of both documents were placed at every infant's bedside for easy access and reference. Education to the NICU physicians, nurses and staff was provided by the nutrition team with ongoing education. The PDSA (Plan/Do/Study/Act) process improvement model for tracking and monitoring progress with compliance of the guidelines was utilized. Results: The primary aim was to achieve 80% compliance in utilizing the guidelines for VLBW (very low birth weight; < 1500 g) infants by December 31, 2014. Secondary outcome metrics included: 1. Day of life for first feeding; 2. Days to reach full feedings (Figure 2); and 3. Reasons for NPO status. Compliance was compared to a


sample of historic data. Pre-project compliance to the guidelines was found to be 73.1% and increased to 86.5% by the end of the 4th quarter. For all infants, time to achieve goal feedings was decreased from 18.6 days to 16.4 days and for ELBW (< 1000 g) infants, time to achieve goal feedings was reduced from 24.8 days to 21.5 days. NEC rates also declined from 5.9% in 2012 to 3.8% by the end of 2014. A cost savings analysis was completed which revealed a savings of 1.33M dollars in year 2014 by a reduction in number of TPN days, reduction of number of central line associated blood stream infections and less NEC cases when compared to historic (pre guideline) data. Conclusions: Utilization of a standardized feeding protocol in the NICU setting can serve as an effective tool in improving patient safety and nutrition outcomes, leading to cost savings over the course of admission for VLBW infants. By defining goals for initiation and advancement of enteral feedings, there is less variability in feeding practices, thereby reducing TPN and central line days. Limitations with tracking data included difficulty in measuring advancement compliance, infants with complicated medical courses necessitating intermittent NPO status and lack of feeding advancement over weekends despite access to guidelines. Further plans include overcoming barriers to earlier initiation of feedings and evaluation of reasons for NPO status. Financial Support received from: None

Figure 1. MCHLB NICU feeding guidelines.

Figure 2. Number of days to reach full feedings.

18.6 18.9 17.5 16.4

24.828.7

24.821.5

14.4 12.910.2

13.9

0

5

10

15

20

25

30

35

1st Quarter 2nd Quarter 3rd Quarter 4th Quarter

Days to Reach Full Feeds

All

<1000 g

1000-1500


Encore Abstract: Previously presented at Oct. 2015 NASPGHAN Conference

S101—Laparoscopic Gastric Fundoplication in Children With Spinal Muscular Atrophy Type 1 Andrew P. Rogers, MD1; Luther Sigurdsson, MD2; Mary Schroth, MD2; Jocelyn Burke, MD1; Peter Nichol, MD, PhD1 1Surgery, University of Wisconsin, Madison, WI, USA; 2Pediatrics, University of Wisconsin, Madison, WI, USA Purpose: Spinal Muscular Atrophy type-1 (SMA-1) patients are at high-risk for lethal aspiration pneumonia due to combination of profound respiratory muscle weakness, dysfunctional swallowing, and an inability to protect their airway from gastroesophageal reflux (GER). Prior to the laparoscopic era, reflux control surgery in this population was relatively contraindicated as a result of an acute decrease in respiratory effort caused by post-operative pain and narcotic pain medications. Currently, SMA-1 patients at select institutions undergo early laparoscopic fundoplication (LF) and gastrostomy tube (g-tube) placement to control GER and reduce the risk of a lethal aspiration event. This approach, however, has not been widely adopted and reported outcomes are limited. We hypothesize that due to the small size of the incisions, patients with SMA-1 undergoing LF would have minimal impairment in respiratory effort as measured by length of time to post-operative extubation. To test this, we retrospectively compared outcomes in our SMA-1 population to pediatric patients with other neurological diseases (OND) group undergoing LF. Methods: Medical records of children with SMA type 1 and OND who underwent LF/g-tube placement between 2001 and 2014 were reviewed. Time to extubation, length of stay (LOS), time to initiation of gastric feeds and time to full feeds were determined. Long-term data was obtained on LF failure, the need for gastric prokinetic medications, and conversion to jejunal feeds when prokinetics failed. Results: 38 children with SMA-1 and 23 with OND underwent LF. Six of the SMA-1 patients had a pre-existing g-tube; in all others a g-tube was placed at surgery. All SMA-1 and OND patients were extubated the day of surgery. Follow-up was available in 36 SMA-1 and 23 OND patients (6 months to 12 years). There were 12 deaths in the SMA-1(32%) group and none in the OND group. All deaths in the SMA-1 population occurred months to years after the operation and were the result of long term sequelae of respiratory failure from SMA-1. None of these deaths were attributable to a GER aspiration event. There was one failed fundoplication in the SMA-1 group and none in the OND group. SMA-1 patients underwent surgery earlier (p<0.043), had a longer LOS (P<0.023). However, there was no difference in their time to initiation of feeds or time to full feeds post-operatively. Upon long term follow-up SMA-1 patients trended towards a higher rate of jejunal feeds (16% vs. 4%, P=0.238)) and prokinetic use (48% vs. 26%, p=.179) compared to the OND group however this did not reach statistical significance. Conclusions: These results support our hypothesis that LF surgery can be performed safely and has minimal untoward effect on post-operative respiratory effort in SMA-1 patients. SMA-1 patients underwent LF significantly earlier than the OND group, likely due to our approach that once a diagnosis of SMA-1 is made, surgery is undertaken prior to the loss of swallowing function. The greater use of prokinetics and jejunal feedings in SMA-1 patient suggest the presence of intestinal motor dysfunction in this population. Studies with a larger patient cohorts are needed to determine if these trends are significant. Financial Support received from: None

SMA I (n = 38) OND controls (n = 23) p- value

Median age at time of operation (months) 7 (3-18) 10 (1-96) 0.043 Time to initiation of feeds (hours) 20 (4-144) 10 (2-70) 0.814

Time to full feeds (hours) 40 (20-144) 33 (12-220) 0.497 Length of stay (days) 4 (3-8) 3 (2-7) 0.023 Need for jejunal feeds 6 (16%) 1 (4%) 0.238

Need for prokinetic agents 17 (48%) 6 (26%) 0.179


S102—Italian Paediatric Nutrition Day Antonella Lezo, MD1; Antonella Diamanti, MD2; Teresa Capriati, MD2; Paolo Gandullia, MD3; Paolo Fiore, MD3; Laura Lacitignola, MD4; Simona Gatti, MD5; Maria Immacolata Spagnuolo, MD, PhD6; Nicola Cecchi, MD7; Giovanna Verlato, MD8; Sara Borodani, MS-IV9; Maria Luisa Forchielli, MD10; Roberto Panceri, MD11; Elena Brunori, MD12; Maria Pastore, MD13; Sergio Amarri, MD14 1Dietetics and Clinical Nutrition, OIRM-S.Anna Hospital, Città della Salute e della Scienza, Turin, Italy, Turin, Italy; 2Gastroenterology and Clinical Nutrition, Pediatric Hospital Bambino Gesù, Rome, Italy; 3Gastroentrology and Nutrition, Istituto G Gaslini, Genoa, Italy; 4Gastroenterology and Nutrition, Meyer Pediatric Hospital, Florence, Italy; 5Pediatrics, Salesi Hospital, Ancona, Italy; 6Pediatrics, University Federico II, Naples, Italy; 7Pediatrics, Santobono-Paousilipon Hospital, Naples, Italy; 8Dietetics and Clinical Nutrition, Padova University Hospital, Padua, Italy; 9School of Medicine, University of Turin, Turin, Italy; 10Pediatrics, Bologna University, Bologna, Italy; 11Pediatrics, San Gerardo Hospital, Monza, Italy; 12Pediatric Neurology, Stella Maris Foundation, Pisa, Italy; 13Pediatrics, San Giovanni Rotondo Hospital, Foggia, Italy; 14Pediatrics, S. Maria Nuova Hospital, Reggio Emilia, Italy Purpose: Introduction: the prevalence of malnutrition and its impact on outcomes in children is under recognized by clinicians in Italy as well as worldwide. A novel definition of paediatric malnutrition has been recently proposed by the A.S.P.E.N. working group based on the correlation with illness and the use of z-scores of anthropometric parameters. A uniform and simple definition should help diagnosing and treating malnutrition in hospitalized paediatric patients. The aim of our study was to investigate the prevalence of malnutrition and related nutritional support among hospitalized children in Italy, in a nation-wide survey performed in a single day observation (16/4/2015). Methods: An open access website (http://nday.biomedia.net) collected data from 73 hospitals and 101 wards from 14 Italian regions (1.994 patients). Anonymous information were gathered on hospitals’ characteristics, patient’s anthropometric data, admission diagnosis (16 choices), presence of chronic diseases (none or one of 17 choices) and use of nutritional support: oral nutritional supplements (ONS), enteral nutrition (EN) or parenteral nutrition (PN). Z-scores of anthropometric parameters, calculated with Epi Info 7.1.5, defined nutritional status: wasting was identified by BMI and Weight-for-Length z-score (<-1 mild, <-2 moderate, <-3 severe), stunting by Height-for-Age z-score (<-2 moderate, <-3 severe). In absence of height or length, Weight-for-Age z-score was utilized to define wasting. WHO and CDC 2000 growth charts were used respectively for children younger and older than 2 years old. Results: 1790 valid records were obtained for hospitalized patients aged 0-20 years old (53.3% males). For other 141 patients, only weight was available. More than 50% of children (52.9%) were aged 0-6 years. 58.8% of children suffered from chronic diseases. Wasting was detected in 28.7% being more represented in the age range 0-6 and 14-20 years while 17.3% of patients showed stunting; surprisingly almost 27% of them were aged 0-2 yrs (table 1). The admission diagnosis correlated with higher malnutrition rate were psychiatric (eating disorders), infective, pulmonary, gastrointestinal and cardiological pathologies, with >30% of malnourished patients. The prevalence of wasting was higher amongst children with chronic diseases (34.1% vs. 27.1%); stunting prevalence tripled in chronic disease patients (24.5% vs. 8.3%); see figure 2 for the ranking of chronic diseases by malnutrition rate. Nutritional support (any kind) was given only to 23.5% of malnourished children (17%, 25.6% and 36.7%, respectively mild, moderate and severe malnutrition), see figure 1. Furthermore 12.1% of nonmalnourished and 14% of obese patients (BMI or Weight-for-Length z-score>2) received some nutritional support. Nutritional support: 11,7% of malnourished children received ONS (modular or complete), 11.5% enteral nutrition (6.4% via NG tube, 5.1% via gastrostomy) and 6.8 % received PN; in some patients a combination of two. Nutritional support is better represented among stunting patients with 39.5% of treated children. In the category of only weight available patients, 77% of those having Weight-for-Age z-score <-2 receive nutritional support: 32.3% ONS, 58% EN and 6.5% PN (one or a combination of two). 23.4% of the patients in this category have neurological diseases; clinical reasons impeded to detect height or length in the rest of patients. Conclusions: Malnutrition of any grade was observed in nearly 1/3 and stunting in 17% of the reported hospitalized children; it is likely to be under-recognised as the nutritional support reached only a small part of the malnourished children. Financial Support received from: We receive the organizative support by the Italian Society of Gastroenterology and Hepatology and Nutrition.


Table 1. Nutritional status of hospitalized pediatric patients.


S103—Pediatric Renal Revascularization and Nutrition Outcomes Katherine M. Ludwig, MPH, RD1; Jennifer Beehler, MPH, RD1; Kathleen Byrne, RD, CSP1 1Patient Food and Nutrition Services, University of Michigan Mott Children's Hospital, Ann Arbor, MI, USA Purpose: In 2013 the University of Michigan Mott Children’s Hospital developed MTool™, a formalized method of diagnosing pediatric malnutrition based on the recently published, etiology based definition endorsed by A.S.P.E.N, the Academy of Nutrition and Dietetics and the American Academy of Pediatrics. As pediatric renal revascularization surgical patients are a unique subset of the hospital’s population, the MTool™ was applied to a retrospective study cohort to help identify population specific trends on the prevalence, severity, and distribution of illness related pediatric malnutrition and the possible relationship to clinical outcomes. Methods: A retrospective chart review was conducted on 73 patients admitted to the PICU for surgical intervention for renal artery revascularization from 2010-2014. Seven patients were excluded secondary to incomplete medical documentation and/or for receiving nutrition support prior to surgical intervention. Each case was analyzed to determine if the patient was malnourished on admission and again at discharge using a retrospective application of our MTool™ to determine the severity of malnutrition. Data was then evaluated to determine possible associations between malnutrition and clinical outcomes. Results: After exclusions, 66 patients were included for review. Our results found that on admission prior to the implementation of MTool™ only 1.5% (N =1) of the population was diagnosed with malnutrition. When the MTool™ was retrospectively applied, the number of patients identified as malnourished on admission increased to 15.1% (N = 10). Of those identified as malnourished, 80% (N=8) were identified as at risk/mild and 20% (N=2) were identified as severe. At discharge using MTool™ criteria the number of malnourished patients increased to 27.3% (N = 18) with 44% (N=8) identified as at risk/mild, 39% (N=7) as moderate and 16.7% (N= 3) as severe. On average, patients were NPO for 5 days prior to nutrition support start and/or diet advancement, weight loss was present in 80% (N=53) of the population with an average loss of 6%, 27% (N=18) developed pancreatitis, and 27% (N=5) of those with pancreatitis had their lipids held during PN therapy. Pearson’s Correlation (2 tailed) was utilized in SPSS for statistical analysis. Results revealed a moderate positive correlation between malnutrition at discharge with malnutrition on admission (r=0.500, N=66, P=0.01). There were no statistically significant correlations with malnutrition at discharge associated with length of stay, pancreatitis, holding of IV lipids or weight loss. However, correlations existed between length of stay and pancreatitis (moderate, r=0.438, N=66, P=0.01), holding of IV lipids during PN use (weak, r=0.316, N=66, P=0.01) and PN days (very strong, r=0.927, N=66, P=0.01 respectively). A moderate positive correlation existed between pancreatitis and holding of IV lipids (r=0.468, N=66, P=0.01) as well as with number of PN days and pancreatitis (r=0.575, N=66, P=0.01). Conclusions: Due to the highly specialized practice of surgical renal revascularization patients, minimal data exists regarding feeding practices in this unique population. Our data revealed a trend toward weight loss and malnutrition at discharge although not statistically significant. Establishment and implementation of nutrition support protocols in this population is warranted to help prevent weight loss and/or worsening of malnutrition throughout admission. Financial Support received from: None

S104—Inadequate Access to Community Registered Dietitian Services for a High-Risk, Vulnerable Patient Population—A Pediatric Situational Assessment Susan M. Bird, BSc, MScFN (c) 1; Janet Madill, PhD, RD1; Colleen O'Connor, PhD, RD1 1Food & Nutrition, Brescia University College, London, Canada Purpose: Adequate nutritional intake during the critical years of zero to five can prevent a multitude of developmental and psychological deficits. To avert these issues, children require access to an interdisciplinary team which includes the specialized services of registered dietitians (RDs). Observations made by experienced health care professionals and agencies in London, Ontario indicate that community RD accessibility for the pediatric community is limited with referrals being commonly rejected. The situational assessment will 1) quantify the gap between patients able to access registered dietitians and those who are not; 2) identify the nutrition-related problems of most frequent concern and 3) provide qualitative information regarding the negative effects of inaccessibility to RD services and patient impact statements. Methods: Using a grounded theory approach, a list of agencies and pediatricians who frequently require access to RD services, was generated. Eligibility criteria were being a health care provider (HCP) who interacted with children aged 0-5 and having no access to dietetic services on staff. HCPs tracked patient referrals that were


requested, recommended and accepted using a tally sheet for a period of eight weeks. A HCP questionnaire was also completed. Parents of young children were recruited to participate in focus groups to generate impact statements regarding their experiences with community dietetic services. Inclusion criteria included being a caregiver in London, Ontario and having experience trying to access community RD support for children (aged 0-5). All data collection tools were tested for clarity and understanding prior to data collection. Results: Six agencies and 17 pediatricians were contacted to participate, of which two agencies and seven pediatricians completed data collection. Preliminary results revealed 53% (8/15) requested and 93% (88/95) recommended and requested referrals were denied, leaving children without access to community dietetic support, despite their identified need. The most frequently cited reasons for RD referrals were feeding issues (29%; 44/154), followed by failure to thrive (FTT) (24%; 37/154), and general infant nutrition (14%; 22/154). Results from the HCP survey (n=17) show that the majority of HCPs (70%; 12/17) find community RD services difficult to obtain for their pediatric patients reporting that they are “almost impossible to obtain” and “there seems to be few options that are publicly funded”. Almost one quarter of HCPs (23%; 4/17) indicated only referrals for the most complex patients were accepted, and only 1 HCP reported a positive experience regarding community RD accessibility. HCPs also conveyed that inaccessibility is challenging for them with 87% (14/16) stating it affects their practice and/or their patients. Noteworthy is that HCPs recognize that RDs provide unique, specialized and valuable information with 100% of respondents in agreement. Five global themes emerged from the focus group discussions: 1) Frustration with the health care system, 2) Fear, 3) Need for service in the community, 4) Unique skills and role of the RD, and 5) Value of communication among the interdisciplinary team. Conclusions: The results of this research demonstrate that a gap in RD accessibility for children (aged 0-5) exists and many high-risk vulnerable pediatric patients are unable to access much needed community RD services. Pediatricians and allied HCPs identified dietetic services were difficult to obtain, inaccessibility was challenging and RDs provide unique, specialized and valuable information to patients. Parents have expressed frustration, echoing the difficulty with access. Further research is needed to identify barriers and solutions to community dietetic inaccessibility. Financial Support received from: none

S105—Patterns in Nutritional Provision Following Stage 1 Palliation for Hypoplastic Left Heart Syndrome Christine L. French, MS, RD, LDN1; Kimberly Mills, MD2; Catherine Baird1; John N. Kheir, MD2 1Gastroenterology and Nutrition, Boston Children's Hospital, Boston, MA, USA; 2Cardiology, Boston Children's Hospital, Boston, MA, USA Purpose: The ideal management of nutrition and fluid following cardiac surgery in neonates is complex and dynamic. Nutritional support in the form of parenteral (PN) and enteral nutrition (EN) is essential to promote post-operative recovery and neurodevelopmental growth. The breakdown of macronutrients provided to neonates following stage 1 palliation (S1P) for hypoplastic left heart syndrome (HLHS) has previously not been characterized. Methods: Retrospectively, we reviewed a cohort of 56 consecutive neonates who underwent S1P for HLHS between September 2012 and February 2015. Patient characteristics, mode of nutritional support and proportion of macronutrient supplementation were extracted from the electronic medical record and a detailed nutrition record. Results: The average daily caloric intake from POD 0-30 was 86 ± 31.4 kcal/kg/d. Enteral nutrition was initiated on the second post-operative day (POD) and the percentage of enteral calories increased from 15% of total calories on POD 0-7 to 93% of total calories from POD 24-30 (Figure A). The primary source of energy was carbohydrate from POD 0 – 7 (65% of total calories from carbohydrates) with a decrease in carbohydrate provision from POD 24-30 (49% of total calories from carbohydrates) (Figure B). Protein intake was persistently below A.S.P.E.N. recommendations through POD 30 at 1.8 ± 0.9 g/kg/d. The study population’s birth weight was 3.1 ± 0.5 kg with a weight-for-age z-score (WAZ) of -0.47 ± 0.8. Minimal growth was obtained while admitted to the hospital with a mean discharge weight of 3.2 ± 0.4 kg and a WAZ of -2.5 ± 0.8. Conclusions: Among newborns with HLHS, protein intake is deficient during the critical early postoperative period, placing them at risk for protein calorie malnutrition. Specific clinical practice changes should be introduced which optimize protein intake in the setting of the fluid restrictions imposed upon these patients. Financial Support received from: none


S106—Gastric Residual Volume Does Not Predict Delayed Gastric Emptying in Critically Ill Children Enid E. Martinez, MD1; Nicole Stenquist, BA1; Luis M. Pereira, PhD1; Mark D. Kellogg, PhD2; Kathleen M. Gura, PhD3; Samuel Nurko, MD3; Nilesh M. Mehta, MD1,3 1Department of Anaesthesia, Perioperative and Pain Medicine, Boston Children's Hospital, Boston, MA, USA; 2Division of Pathology, Boston Children's Hospital, Boston, MA, USA; 3Division of Gastroenterology, Hepatology and Nutrition, Boston Children's Hospital, Boston, MA, USA Purpose: Delayed gastric emptying (GE), a cause for nutrition intolerance and poor nutrition delivery, affects 50% of critically ill children. High gastric residual volume (GRV) is commonly used as a measure of nutrition intolerance and an indirect measure of delayed GE in the pediatric intensive care unit (PICU). However, GRV does not predict delayed GE in critically ill adults, and has not been studied in children. Reliable detection of delayed GE in children is essential for safe and optimal enteral nutrition (EN) delivery. We aimed to compare GRV measurements in critically ill children with and without delayed gastric emptying, as determined by the paracetamol* absorption test (PAT). We hypothesized that current thresholds of GRV are not predictive of delayed GE and nutrition intolerance in critically ill children. * Alternative trade name for paracetamol is acetaminophen Methods: Prospective pilot study in children 1 year and older, admitted to the PICU and eligible to initiate EN via the gastric route. Demographic, clinical and nutritional data, including time to initiation of EN, time to goal, perceived intolerance, were recorded. GE was directly measured by the PAT. A 15 mg/kg or max 650mg one time dose of paracetamol was administered via nasogastric tube or gastrostomy. Blood samples for plasma paracetamol concentration testing were obtained prior to administering paracetamol and at 4 additional time points after administration of paracetamol over 4 hours. A truncated AUC60, area under the curve at 60 minutes, less than 600 mg*min/L determined by the trapezoidal rule was diagnostic of delayed GE. The GRV measurement at time of GE testing by PAT, and clinical signs of intolerance (emesis, diarrhea, abdominal distension and abdominal discomfort)


were compared between subjects with delayed and normal GE. Continuous data presented as median (IQR); categorical data as frequency (%). Results: We enrolled 9 subjects, median age 11.4 (5.3,15.6) years, 55.6% male. Six subjects had delayed GE and 3 had normal GE (median AUC60 118.1mg*min/L (80.4,151.35) vs. 850.8 mg*min/L (791,855), p=0.01, Figure 1). Study procedures were completed within a median of 21 hours (15,35.5) from EN initiation. Median rate of EN at time of study was 20ml/kg (10, 37.5). Seven of nine (77.8%) subjects had a primary respiratory diagnosis; 8/9 (88.9%) were on mechanical ventilation. There were no differences in severity of illness scores or the frequency of administration of sedative drips between subjects with delayed and normal GE. In subjects with delayed GE, EN was initiated on average 28 hours later than in subjects with normal GE; and the median time to reaching goal feeds was longer, 115 hours (29,168.5) vs. 50.5 hours (26,192). There was no difference between the mean GRV measurements in the delayed GE group at time of GE testing by PAT, median 0.31mL/kg (0.16, 2) vs. 1.6 mL/kg (0.02, 2), p >0.05. GRV was <150mL in all subjects; median in the delayed GE group was 10mL (4.5, 47.3) and 20 mL (1, 90) in those with normal GE. There was no difference in the incidence of other signs of EN intolerance between subjects with delayed GE and normal GE: emesis (0/6 vs. 1/3), diarrhea (1/6 vs.1/3), abdominal distension (0 vs. 0) and abdominal discomfort (1/6 vs. 0/3). Conclusions: GRV measurements as well as clinical signs of intolerance did not differentiate critically ill children with delayed GE from those with normal GE in our current study. GRV measurements were not higher in patients with delayed initiation and advancement of EN. Novel accurate measures of GE are needed to guide nutrition intolerance assessments and safe EN delivery in the PICU. Future studies examining serum biomarkers of intolerance are desirable. Financial Support received from: Boston Children's Hospital, Office of Faculty Development/Clinical and Translational Research Executive Committee/ Research Recruitment and Resource Committee Career Development Fellowships

Figure 1. Mean plasma paracetamol concentrations in critically ill children with delayed (squares) gastric emptying and normal (diamonds) gastric emptying over the duration of the paracetamol absorption test. * p-value


S107—Critically Ill Overweight and Obese Children Are at Risk for Caloric Overfeeding and Suboptimal Protein Intake Enid E. Martinez, MD1; Nicole Stenquist, BA1; Daniela Anderson, MD2; Eliana B. Muñoz, MD3; Nilesh M. Mehta, MD1,4 1Division of Critical Care Medicine, Department of Anaesthesia, Perioperative and Pain Medicine, Boston Children's Hospital, Boston, MA, USA; 2University of Sao Paulo-Ribeirao Preto School of Medicine, Sao Paulo, Brazil; 3Universidad de Chile, Hospital Dr. Luis Calvo Mackenna, Providencia, Chile; 4Department Gastroenterology, Hepatology, and Nutrition, Boston Children's Hospital, Boston, MA,USA Purpose: Thirteen percent of children admitted to pediatric intensive care units (PICU) are estimated to be obese. Obesity in critically ill patients is associated with altered metabolism and predominant loss of lean body mass despite excess fat stores. Best practices for the obese critically ill adult have been published and demonstrated improved outcomes. Current nutrition practices and best management strategies for the obese critically ill child have not been adequately studied. We aimed to describe nutrition practices in critically ill children who were overweight or obese and highlight practice and knowledge deficits. Methods: We performed a retrospective cohort study to record nutrition practices in children, 2 years and older, admitted between 2010 to 2015 for at least 2 days to two PICUs in a single-center, and with body mass index (BMI) ≥ 85th percentile consistent with CDC criteria for overweight and obesity. Data, including demographics, and details of clinical, and nutritional variables were recorded. Continuous data are presented as mean (standard deviation) and categorical data are presented as frequency (percentage). In a subgroup of patients who were tested with indirect calorimetry (IC), we examined the agreement between measured resting energy expenditure (MREE) and estimated energy expenditure (EEE) from standard equations. Estimated values that were within 20% of MREE were deemed as accurate. We simulated a 7-day cumulative energy imbalance in this subgroup based on the difference between EEE and MREE values. Results: Data from 80 PICU admissions for 50 eligible patients [50% male, mean age 10.1 (5.7) years] were included. Mean BMI percentile was 97.7 (3.3). Admission diagnoses included respiratory insufficiency 36/80 (45%), post-operative care 22/80 (28%), and gastrointestinal disease including pancreatitis 11/80 (14%). Median (IQR) PICU length of stay was 8 days (5,15.25). A nutrition evaluation with specified energy goal was available in 49/80 (61.3%) admissions within a mean of 2.5 days (1.9) after admission. Energy expenditure was estimated using the Schofield equation in 19/49 (38.8%) and World Health Organization (WHO) equation in 9/49 (18.4%) admissions. In 33/49 (67.3%) admissions an average stress factor of 1.15 (range 0.6-2) was applied to the energy expenditure estimate. Mean prescribed energy was 36.6 kcal/kg (15.7). Protein goal was clearly identified in the nutrition evaluation of 24/80 (30%) admissions, with a mean protein prescription of 1.3 g/kg/d (0.63); a third (8/24) of the protein prescriptions were within the age-based recommendations by the A.S.P.E.N. IC was completed during 9 admissions for 7 patients. Mean MREE was 21 kcal/kg (10.6) and was significantly lower than that estimated by Schofield equation [EEEs = 34.4 kcal/kg (10.2)] and WHO [EEEw = 37.7 kcal/kg (14.1)] in this subcohort, p=0.003. Mean bias (95% limits) for agreement between EEEs and MREE was +51.62 (-139.1 to 35.9)% and between EEEw and MREE was +57.5% (-146.6 to 31.7)%, Figure 1. Schofield and the WHO equations estimated the EE within 20% of MREE in 3/9 (33.3%) and 2/9 (22.2%) admissions, respectively. The average simulated 7-day cumulative energy imbalance was +3630 kcals (2847) for the Schofield equation and +4163 kcals (2893) for WHO equation (Figure 2). Conclusions: Obese critically ill children are at risk for caloric overfeeding due to the use of inaccurate equations, and inadequate protein intake due to suboptimal prescription. Indirect calorimetry should be prioritized, when available, to determine energy expenditure in this vulnerable cohort. The use of stress factors should be avoided given the risk for overfeeding. Protein intake must be optimized to avoid unintended loss of lean body mass and potential for poor outcome. Financial Support received from: No financial support to report for this study.


Figure 1. Bland-Altman plots depicting agreement between measured resting energy expenditure (MREE) by indirect calorimetry and estimated energy expenditure by the Schofield equation (EEEs, Panel A) and the World Health Organization (WHO) equation (EEEw, Panel B). The dashed lines represent the mean bias and the dotted lines represent the limits of agreement.

Figure 2. Seven day simulated cumulative energy imbalance with the use of predictive equations to estimate energy expenditure in 9 overweight and obese critically ill children. Difference between estimated energy expenditure (EEE) as determined by two predictive equations and measured resting energy expenditure (MREE) as determined with indirect calorimetry (EEE-MREE) was summed over a simulated seven day period. The black bars represent the difference between EEE and MREE determined by the Schofield equation. The gray bars represent the difference between EEE and MREE determined by the World Health Organization (WHO) equation.


ENCORE ABSTRACT: Previously presented at NACFC (North American Cystic Fibrosis Conference), October 2015. Abstract not published at request of the authors.

S108—Growth Parameters in Infants With Cystic Fibrosis: A Comparative Registry Analysis Michael Schechter, MD MPH1; Suzanne Michel, MPH RD2; Jerry Hall, MD3; Shufang Liu, PhD3; Rupal Khurmi, MD3; Mark Haupt, MD3 1Children's Hospital of Richmond, Virginia Commonwealth University, Richmond, VA, USA; 2Medical University of South Carolina, Charleston, SC, USA; 3AbbVie Inc., North Chicago, IL, USA Financial Support received from: AbbVie Inc. funded the research and writing of this abstract.

S109—Nutrition Initiation and Advancement Pathway in the Pediatric Critical Care Unit Monica Nagle, RD, CNSC, LDN1; Stephanie Seiple, RD, CNSC, LDN1; Shiela Falk, RD, CNSC, LDN1; Vijay Srinivasan, MD2; Maria Mascarenhas, MBBS 3; Judy Verger, PhD, CRNP4; Beth Connor, RN5; Lena Gayraud, RN5; Jennifer Hewlett, PharmD6; Martha Sisko, RN2; Ruoxi Li, MBA7; Blair Kraus, MSN, RN7; Sharon Y. Irving, PhD, CRNP4 1Clinical Nutrition, Children's Hospital of Philadelphia, Philadelphia, PA, USA; 2Anesthesiology and Critical Care Medicine, Children's Hospital of Philadelphia, Philadelphia, PA, USA; 3Gastroenterology, Hepatology and Nutrition, Children's Hospital of Philadelphia, Philadelphia, PA, USA; 4School of Nursing, University of Pennsylvania, Philadelphia, PA, USA; 5Nursing, Children's Hospital of Philadelphia, Philadelphia, PA, USA; 6Pharmacy Services, Children's Hospital of Philadelphia, Philadelphia, PA, USA; 7Center for Clinical Quality and Improvement, Children's Hospital of Philadelphia, Philadelphia, PA, USA

Purpose: There has not been a standardized approach to initiation and advancement of enteral nutrition (EN) for patients admitted to the critical care units at CHOP. Historically the clinical care team has been responsible for ordering and advancing EN based on attending physician judgment, patient acuity and Registered Dietitian (RD) recommendation. Evidence suggests nutrition support early in the course of illness reduces morbidity and mortality, decreases healthcare acquired infections and may improve patient outcomes. EN has been shown to be superior to parenteral nutrition (PN) for patients with a functioning gastrointestinal system. The American Society for Parenteral and Enteral Nutrition (A.S.P.E.N.) guidelines for nutrition support of the critically ill child recommend use of a feeding protocol to guide nutrition support in this population. An interdisciplinary, collaborative group consisting of providers from critical care units, clinical nutrition, pharmacy, and gastroenterology, hepatology and nutrition examined the historical scope of duration without feedings or NPO time in patients admitted to the PICU, and established the necessity for a feeding pathway. The overall goals of the pathway include: 1) decreased time from the critical care unit admission to initiation of nutrition support; 2) decreased time from initiation of EN to attainment of goal nutrition; 3) reduction in unnecessary and prolonged enteral feeding interruptions and 4) reduction in unnecessary use of parenteral nutrition. Methods: The multidisciplinary ICU-Pediatric Nutrition Team (ICU-PNuT) examined NPO time in critical care patients. The pathway was developed through the expertise of the ICU-PNuT group with input from nursing staff, critical care executive leadership, electronic medical record programmers and other key stakeholders. The Critical Care Nutrition Pathway was introduced through widespread communication, including presentation at weekly CQI meeting, and email distribution. An eLearning module was developed and disseminated to all critical care providers and nursing staff for pathway education. All applicable patients are initiated on the pathway within 48 hours of admission. Patients excluded from the pathway include those with special nutritional needs, such as metabolic disorders or patients in the intestinal rehabilitation program. Results: The Critical Care Nutrition Pathway was successfully published in December 2014. Providers are guided through initiation and advancement of EN and given recommendations for initiation of PN, if necessary. In addition, the pathway includes extensive resources on prescribing bowel regimens, anti-reflux medications, and provides definitions and potential resolution of feeding intolerance. The feeding pathway has highlighted nutrition as an important factor in providing care to the critically ill child. Preliminary data on the initial 30 patients started on the pathway demonstrate an average age of 4.1 years, with 53% male and 85% mechanically ventilated. The average time from critical care unit admission to start of EN was 2.2 days. The primary mode of delivery was via nasogastric tube (77%). Conclusions: The Critical Care Nutrition Pathway has highlighted nutrition as an important component in providing pediatric critical care. Providing an initiation and advancement pathway for nutrition support may prove to enhance nutrition delivery and improve outcomes in these patients. Providers may also benefit from the information listed in the pathway to address challenges that arise related to nutrition support.



ENCORE ABSTRACT: Previously presented May 2014 – Pediatric Academic Society Meeting.

S110—Introduction of a Feeding Protocol in the Pediatric Intensive Care Unit to Improve Caloric Intake of Critically Ill Children Helen C. Allen, MD1,2; Paul Watkins, RN2; Michael Anderson, PhD3; Summer G. Frank, MPH3 1Pediatric Critical Care, University of Oklahoma, Oklahoma City, OK, USA; 2Pediatric Intensive Care Unit, Children's Hospital at OU Medical Center, Oklahoma City, OK, USA; 3College of Public Health, Biostatistics, University of Oklahoma, Oklahoma City, OK, USA Purpose: Critically ill patients frequently do not receive adequate caloric intake during their hospital stay. Adult studies have demonstrated that critically ill patients who receive early enteral nutrition have decreased infection rates, increased ventilator free days, and decreased mortality. Recently, pediatric studies have shown that early enteral nutrition can be achieved and is well tolerated in Pediatric Intensive Care Units (PICU) with the use of feeding protocols. In an effort to decrease the time it takes to initiate feeds in the PICU, as well as improve the total caloric intake for PICU patients throughout their length of stay (LOS) we implemented a feeding protocol in our PICU. We hypothesized that the initiation of a feeding protocol in the PICU would increase the percent of goal calories fed to patients during their LOS in the PICU as well as decrease the time for initiation of feeds. Methods: This was a retrospective and concurrent chart review of patients admitted to the PICU and enterally fed during their PICU stay. We collected data on two separate patient populations: those who received enteral feeds before the initiation of the feeding protocol (retrospective group) and those who were enterally fed per the feeding protocol (prospective group). Results: A total of 74 patients were included, 50 in the prospective group and 24 in the retrospective group. Patients on the enteral feeding protocol had a decrease in the median time from admission to the initiation of feeds (18 vs. 34.5 hours p<0.0035) and from admission to time goal calorie feeds were reached (27.5 vs. 89 hours p<0.0002). In addition, PICU patients on the enteral feeding protocol received a significantly greater mean percent of their goal calories for their LOS (77.74 vs. 53.38% p<0.0001). In multiple regression analysis, the feeding protocol remained a significant predictor of the percent of goal calories for LOS when controlling for age, ventilator status, number of times feeds were held, time to initiation of feeds and time to reach full feeds. Conclusions: The introduction of an enteral feeding protocol in the PICU resulted in a significant increase in the percent of goal calories provided to patients during their PICU LOS as well as shorter times for the initiation of feeds. Financial Support received from: none

S111—Does Persistent Inflammatory Catabolic Syndrome Exist in Critically Ill Neonates? Bodil Larsen, PhD1,2; Vera Mazurak, PhD2; Chloe Joynt, MD3; Nicole Ellis, BSc1,2; Caitlin Hughes, BSc1,2 1Nutrition Services, Alberta Health Services, Edmonton, Canada; 2Faculty of Agriculture, Life and Environmental Sciences, University of Alberta, Edmonton, Canada; 3Department of Pediatrics, University of Alberta, Edmonton, Canada Purpose: Background: Persistent inflammatory catabolic syndrome (PICS) has not been described in the critically ill infant population. The objective of this study was to propose a definition of PICS in critically ill infants based on an adult definition. Methods: A previously published adult criteria of PICS was modified by applying anthropometric and biochemical reference ranges for infants. A prospective chart review of all admissions to a tertiary surgical neonatal intensive care unit (NICU) was performed over 65 days. Daily data was collected on anthropometrics, biochemistries (C reactive protein, lymphocytes, neutrophils, and prealbumin), demographics, and clinical variables. Infants were categorized as having or not having PICS and differences between groups determined. Results: Results: Infants (n=75) were admitted to NICU with a total of 44 infants eligible for the study (35.6± 4.7 weeks; aged 0-88 days at admission).Twenty percent of infants admitted to the NICU developed PICS using the proposed criteria. Infants with PICS were more likely to be classified as failure to thrive (53%) and, on average, met only 75% of their anticipated weight gain. Significantly more infants with PICS had undergone surgery (100%; P = 0.01), were on inotropic medication (40%; P = 0.05), and had a longer NICU and total hospital length of stay than infants without PICS (P < 0.001 and P < 0.001). Infants with PICS had higher peak glucose levels (11.8mmol/L ± 7.3) and elevated urea concentrations (7.9mmol/L ± 4.6).


Conclusions: Conclusions: PICS may be identified using the definition proposed in this study and does exist in a critically ill neonatal population. Compared to infants without PICS, infants with PICS had metabolic dysregulation, a profound decrease in growth velocity, and longer length of clinical stay despite no significant differences in clinical severity scores or diagnosis between the two groups. Further validation of this definition is required, and research into nutritional therapies for PICS should be explored. Financial Support received from: None

S112—Nutritional Status and Clinical Outcomes in Critically Ill Children Admitted to a Large Multidisciplinary Pediatric Intensive Care Unit Tito J. Velasquez, RD, LD1; Guisela Mackey, BS2,3; Jennifer L. Lusk, RN, CPNP-AC2,3; Ursula G. Kyle, MS, RD,LD2,3; Tonita P. Fontenot, RN, MBA3; Pamela A. Marshall, BSN, RN3; Lara S. Shekerdemian, MD2,3; Jorge Coss-Bu, MD2,3 1Clinical Nutrition Services, Texas Children's Hospital, Houston, TX, USA; 2Pediatrics, Baylor College of Medicine, Houston, TX, USA; 3Critical Care, Texas Children's Hospital, Houston, TX, USA Purpose: Critically ill children in the pediatric intensive care unit (PICU) are at high risk for developing nutritional deficiencies and undernutrition is known to be a risk factor for morbidity and mortality. Malnutrition represents a continuous spectrum ranging from marginal nutrient status to severe metabolic and functional alterations and this in turn, affects clinical outcomes. The aim of the study was to assess nutritional status of critically ill children admitted to the PICU and its relationship to clinical outcomes. Methods: Critically ill children age 6 months to 18 years were prospectively enrolled on PICU admission. Their nutritional status was assessed by weight for age (WFA: underweight), weight for height (WFH: wasting), height for age (HFA: stunting) z-scores and mid upper arm circumference (MUAC: wasting) according to the WHO. (1,2) Malnutrition was defined as mild, moderate, and severe if z-scores were > -1, >-2, and >-3, respectively. Hospital and PICU length of stay (LOS), duration of mechanical ventilation (MV), and risk of mortality (ROM) by the Pediatric Index of Mortality 2 (PIM2) were obtained. Sensitivity and specificity of the MUAC to identify children with wasting (WFH) were calculated. Results: Two hundred and two children (112 males), aged 63 months (18-157; median (25-75th IQR)), were included in the study. The hospital LOS was 7.5 (4-14) days; PICU LOS: 2 (1-4) days; duration of MV, 0 (0-2) days; PIM2 ROM 2.64±0.28%. WFA, WFH, and HFA z-scores of -0.53±0.16; 0.13±0.14; and -1.03±0.14 respectively; MUAC, 16.2±0.21 cm (6 to 59 months, n=101); 24.2±0.58 cm (5 to 18 years, n=101). The prevalence of underweight, wasting and stunting was 29.20%, 46.03%, and 19.8% respectively. The sensitivity and specificity for MUAC vs. WFH to identify wasting was: 35.7% (21.5-51.9; 95% CI) and 96.9% (92.8-98.9), respectively. Patients with a wasting condition (n=40) were more likely to have a hospital LOS > 1 week vs. children with no wasting (n=162); Odds Ratio 2.91 (1.05-8.07 95% CI); p < 0.05. Values are mean ± SE. Conclusions: Malnutrition in critically ill children is prevalent with half of the patients being stunted, reflecting the chronic nature of the disease process and its effects on the nutritional status. The performance of MUAC as a screening tool in this population was poor, but identified correctly almost all children with wasting. Patients who were underweight and wasted had a longer duration of mechanical ventilation. There was an association between nutritional status and risk of mortality.1 WHO: Technical Report Series, No. 854, 19952 Bulletin of the WHO, 1997, 75:11-18 Financial Support received from: None

Underweight Wasting Stunting NO (n=143) YES (n=59) NO (n=162) YES (n=40) NO (n=109) YES (n=93)

Hospital LOS days 13±2 13±3 12±1 16±4 14±2 12±2 PICU LOS days 3.2±1 4±1 3.2±0.4 5.2±2 3.8±1 3.3±1

MV days 1.4±0.3 4±1a 1.4±0.3 4.6±2b 1.6±0.4 2.6±1 PIM2 ROM % 1.9±0.1 4.2±0.8b 2.2±0.2 4.2±1.2b 1.9±0.2 3.4±0.6b

Values are mean±SE; a p < 0.05, b p < 0.01, by unpaired t-test.


S113—Nutritional Status and Somatic Growth in Children after Lung Transplantation Guisela Mackey, BS1,2; Tito J. Velasquez, RD, LD3; Ursula G. Kyle, MS, RD, LD1,2; Brittany R. Pearo, RD, LD3; Suzanne T. Culhane, RD, LD3; M. Carolina Gazzaneo, MD1,2; George B. Mallory, MD4,2; Lara S. Shekerdemian, MD1,2; Jorge Coss-Bu, MD1,2 1Critical Care, Texas Children's Hospital, Houston, TX, USA; 2Pediatrics, Baylor College of Medicine, Houston, TX, USA; 3Clinical Nutrition Services, Houston, TX, USA; 4Pulmonary Medicine, Texas Children's Hospital, Houston, TX, USA Purpose: Lung transplantation is the final therapeutic option for children with end-stage lung disease. Malnutrition increases risk of mortality post-operatively through impaired immunity, delayed wound healing, and decreased muscular and respiratory function. Several studies in adults have indicated an association between nutritional status at time of transplantation and outcomes. The aim of this study was to evaluate nutritional status in children admitted to the pediatric intensive care unit (PICU) at the time of lung transplantation (LTx) and somatic growth during the first year thereafter. Methods: Retrospective chart review of patients who underwent lung transplant (1/11-8/14). Nutritional status on admission to the PICU and for the first year after lung transplant were assessed by weight for age (WFA: underweight), weight for height (WFH: wasting), height for age (HFA: stunting) z-scores according to the WHO/CDC growth charts. Malnutrition was defined as mild, moderate, and severe if z-scores were > -1, >-2, and >-3, respectively (JPEN 2013;37:460). Hospital and PICU length of stay (LOS), and duration of mechanical ventilation were obtained. Results: Fifty-three children (19 males) were included in the review: age 10.9 years (3.8-16; median (25th-75th IQR)); Hospital LOS, 21 (12-71) days; PICU LOS: 4.2 (3-5) days; duration of mechanical ventilation, 1 (1-2) days. The prevalence of underweight, wasting and stunting at time of LTx was: 66.0%, 33.9%, and 69.8%, respectively. Patients with cystic fibrosis (n=25) had on admission a WFA, HFA and WFH z-scores of -1.83±0.93 vs. -0.84±1.66 (p< 0.05), -1.98±1.10 vs -1.03±1.56 (p < 0.05), and -0.27±1.40 vs -0.85±1.84 (p =0.21) for non cystic fibrosis patients, respectively. There was no association between presence of underweight, wasting and stunting and longer hospital and PICU LOS and duration of mechanical ventilation. Two patients died during the first year after LTx. Values are mean±SD. Conclusions: Malnutrition was highly prevalent in this cohort of children at the time of lung transplantation. Patients with a diagnosis of cystic fibrosis had a higher prevalence of underweight and stunting compared to non cystic fibrosis patients. There was an improvement on anthropometric variables after lung transplantation without significant change on underweight and wasting. Patients increased their height after LTX, but at a lower than expected rate for age leading to worsening stunting. Financial Support received from: None

LTx n=53

3 months n=52

6 months n=52

9 months n=51

12 months n=51

Weight kg 29.1±18 31.3±18* 32.8±19* 32.9±19* 33.8±19* Height cm 122±34 124±34* 125±33* 126±33* 127±32*

WFA z-score -1.29±1.46 -1.45±2.70 -1.25±1.98 -1.32±2.19 -1.43±2.01 WFH z-score -0.60±1.67 -0.29±1.55 -0.25±1.40 -0.16±1.48 -0.16±1.55 HFA z-score -1.46±1.46 -1.94±1.49* -1.91±1.48* -1.81±2.46 -2.04±1.40*

Values are mean ± SD; * p < 0.05 vs time of LTx by paired t-test.


S114—A Randomized Controlled Trial Assessing Growth of Infants Fed a 100% Whey Extensively Hydrolyzed Formula Compared To a Casein-Based Extensively Hydrolyzed Formula Laura Czerkies, MS, RD1; David Fields, PhD2; Shumei Sun, PhD3; Heidi Storm, MS, RD1; Jose Saavedra, MD4; Ricardo Sorensen, MD5 1Nestle Nutrition, Florham Park, NJ, USA; 2University of Oklahoma, Oklahoma City, OK, USA; 3Virginia Commonwealth University, Richmond, VA, USA; 4Nestle Nutrition, Vevey, Switzerland; 5Louisiana State University, New Orleans, LA, USA Purpose: Infants with cow’s milk allergy who are not exclusively breastfed require the use of formulas whose protein source meets the American Academy of Pediatrics criteria for hypoallergenicity and supports adequate growth. A new, 100% whey-based, hypoallergenic extensively hydrolyzed infant formula with Bifidobacterium lactis (B.lactis CNCM I-3446) and 49% of its fat from medium chain triglycerides (EHF-W) designed for the dietary management of cow’s milk protein allergy has been developed. A study comparing the growth of healthy infants fed EHF-W to infants fed a commercially available casein-based extensively hydrolyzed formula containing 55% of fat from medium chain triglycerides (EHF-C) was conducted to evaluate the nutritional adequacy of the EHF-W. Methods: Healthy, formula fed full-term infants were randomized at 14±3 days of age to either EHF-W or EHF-C until 112 days of age. Anthropometrics were assessed at 14, 28, 56, 84, and 112 days of age. Daily records documenting tolerance-related outcomes were kept for 2 days prior to study visits. A subset of 42 infants had a blood sample taken at 84 days for assessment of serum albumin and plasma amino acids. Results: A total of 282 infants were randomized (124 EHF-W, 158 EHF-C). Significantly more infants dropped out of the EHF-C (56%) as compared to the EHF-W (41%) group. Mean daily weight gain was significantly higher in the EHF-W group compared to the EHF-C group (27.95 ± 5.91 g/d in EHF-W, 25.93 ± 6.12 g/d in EHF-C; p = 0.027) and the EHF-W group reported significantly fewer stools (2.2 stools/d in EHF-W, 3.6 stools/d in EHF-C; p <0.0001). A significantly higher percentage of infants in the EHF-C group experienced vomiting than in the EHF-W formula group (13% in EHF-C, 7% in EHF-W; p=0.034). There were no significant differences in gas, mood, or sleep between the groups. Serum albumin was similar between the two groups. Plasma arginine (67 ± 16 µmol/L in EHF-W, 95 ± 56 µmol/L in EHF-C; p<0.05) and valine (174 ± 48 µmol/L in EHF-W, 216 ± 67 µmol/L in EHF-C; p<0.05) were significantly lower in the EHF-W group than EHF-C, while leucine (157 ± 35 µmol/L in EHF-W, 131 ± 45 µmol/L in EHF-C; p< 0.05) and lysine (217 ± 40 µmol/L, EHF-W vs 174 ± 58 µmol/L in EHF-C; p<0.05) were significantly higher. Values for those amino acids were all within normal limits for both groups. Infants assigned to the EHF-W formula had significantly fewer adverse events than infants assigned to the EHF-C formula (39% in EHF-W, 61% in EHF-C; p<0.001). In the EHF-W group, 14% of the adverse events were reported to be definitely or probably related to the formula by the site investigator; in the EHF-C group, 29% of the adverse events were reported to be definitely or probably related to the formula. Conclusions: The new, 100% whey-based hypoallergenic extensively hydrolyzed formula containing Bifidobacterium lactis and medium chain triglycerides supported growth of healthy infants. Future studies on the application of this formula in clinically indicated populations are warranted. Financial Support received from: Nestle Nutrition

S115—Identification Factors in Pediatric Malnutrition Assessment Stacey S. Beer, MPH, RD, LD1; Marisa D. Juarez, MPH, RD, LD2; Molly R. Vega, MS, RD, LD2; Nicole M. Smith, MS, RD, LD1; Poyyapakkam R. Srivaths, MD2 1Gastroenterology, Hepatology and Nutrition, Texas Children's Hospital, Houston, TX, USA; 2Renal Department, Texas Children's Hospital, Houston, TX, USA Purpose: The purpose was to identify which specific domains contribute to malnutrition in the pediatric hospitalized patient. Methods: To facilitate a standardized approach to the identification of pediatric malnutrition, a tool was developed which included anthropometrics, physical assessment of wasting, growth velocity, risk factors associated with malnutrition, and impact on functional status. The risk factors associated with malnutrition included in our tool were energy intake, altered absorption, hyercatabolism, increased losses, inflammation, micronutrient deficiency, and physical function. All patients admitted to a tertiary-care pediatric hospital during a 16 day period were screened using this tool. The exclusion criteria included patients less than 45 cm and 44 weeks post menstrual age and adults identified by chronological age of >/= 19 years. General demographic information was obtained as well.


Results: A total of 540 patients were screened, 19 were excluded from the analysis due to insufficient data (272 male, 249 female). The age range was 1-227 months with a mean age of 106 months (8.8 years) (SD 69.8 months). The presence of wasting and decelerated growth or weight loss were significantly associated with prevalence as well as severity of malnutrition (all p<0.000). A lower ZS was associated with having up to 3 risk factors present (p<0.003). Of these, altered absorption (p<0.004) and increased losses (p<0.042) were found to be independently associated with malnutrition (p=0.002). The presence of both risk factors increased the odds ratio (OR) of the prevalence of malnutrition from 2.05 to 2.62 (p=0.002). Conclusions: Malnutrition represents a significant burden on hospitalized patients. Diagnosis is crucial to implement an appropriate nutrition intervention and prevent complications associated with malnutrition. Apart from the evaluation of BMI and Wt/L, growth velocity and physical assessment of wasting should be an integral part of pediatric malnutrition assessment. Malnutrition associated risk factors, including increased losses and altered absorption increase the OR of prevalence of malnutrition and should also be considered at time of nutrition assessment. Financial Support received from: none

S116—Iron Status in Pediatric Celiac Disease: A Retrospective Review Jelena Popov, BSc1; Michelle Gould, MD2; Heather Mileski, RD1; Herbert Brill, MD MBA FRCPC1; Nikhil Pai, MD FRCPC FAAP1 1Department of Pediatrics, Division of Gastroenterology & Nutrition, McMaster University, Hamilton, Canada; 2Department of Pediatrics, University of Toronto, Toronto, Canada Purpose: Celiac disease (CD) causes an autoimmune enteropathy that leads to varying degrees of intestinal injury. The extent of small-bowel involvement in celiac disease may be associated with variability in its clinical presentation and associated nutritional abnormalities. Proximal duodenal integrity is critical for dietary iron


absorption, suggesting a potential role for serum iron studies in the diagnosis of CD; however, current diagnostic standards only recommend measurement of serum IgA antibodies to tissue-transglutaminase (TTG-IgA), and histopathologic assessment of villous damage from proximal duodenal biopsies (Marsh classification). The purpose of this study was to determine whether a relationship exists between serum ferritin levels, TTG-IgA and Marsh classification in a pediatric celiac disease outpatient clinic population. Methods: Patients ages 17 years or younger diagnosed and followed by the pediatric celiac disease clinic from 2007 to 2015 were included for retrospective analysis if they had measures of: serum ferritin, TTG-IgA, and duodenal biopsy-proven celiac disease. Results were analyzed by Marsh classification. Primary outcomes were correlation of serum ferritin with TTG-IgA, and Marsh classification. Secondary outcomes included: change in ferritin levels after CD diagnosis, and correlation with other common intestinally-absorbed nutritional markers. Results: 193 patients were identified from the retrospective chart review. 65 patients had serum ferritin levels performed prior to histopathologic CD diagnosis (Marsh 3A = 18, 28%; Marsh 3B = 23, 35%; Marsh 3C = 22, 34%; Marsh 4 = 3, 5%). 651 paired ferritin / TTG-IgA levels were obtained. Serum ferritin levels had a significant, weak inverse correlation with TTG-IgA (r^2 = 0.013, p = 0.0038). Ferritin levels significantly rose from baseline after CD diagnosis and institution of a gluten-free diet (RR = 2.51; 95% CI: 1.89-3.13). There was no significant correlation between ferritin level and Marsh classification at diagnosis (r^2 = 0.022; p > 0.05). Secondary outcomes included: relationship between 25-OH Vitamin D and TTG-IgA level at diagnosis (n = 29), and assessment of C-reactive protein (CRP) and TTG-IgA level at diagnosis, as a potential confounder of serum ferritin levels. No significant correlation was found between 25-OH Vitamin D and TTG-IgA (n = 29, p > 0.05), or CRP and TTG-IgA (n = 49, p > 0.05). Conclusions: This study suggests that serum ferritin levels in pediatric patients have a significant correlation with serum markers of CD, but this relationship is weak and should not be considered in the diagnostic assessment. Further, there was no correlation between Marsh classification and serum ferritin level, suggesting that the clinical-pathologic correlation between iron status and CD may involve other factors. Pediatric CD patients did show a significant increase in mean serum ferritin levels after their diagnosis of CD, but TTG-IgA level, the biomarker conventionally used in pediatric celiac disease to assess resolution of enteropathy, may not be a strong predictor of improvement in iron stores. Prospective studies in pediatric CD patients, involving assessment of broader nutritional indices at time of diagnosis, may provide further information about optimal nutritional monitoring in this patient population. Financial Support received from: None


S117—The Prevalence of Vitamin D Deficiency in Children With Short Bowel Syndrome Anita Nucci, PhD, MPH, RD, LD1; Jane Anne Yaworski, MSN, RN2; Robyn Sechler, MS, RD, LDN2; Jeffrey Rudolph, MD2 1Department of Nutrition, Georgia State University, Atlanta, GA, USA; 2Intestinal Care and Rehabilitation Center, Children's Hospital of Pittsburgh of UPMC, Pittsburgh, PA, USA Purpose: Many children with short bowel syndrome (SBS) undergo progressive adaptation of their remaining bowel and are able to wean from parenteral nutrition (PN). Previous studies have reported an increase in the prevalence of micronutrient deficiencies after full discontinuation of PN. The risk of nutrient deficiency after achieving enteral autonomy is dependent upon the type of bowel resected. Children who have had an ileal resection are predisposed to deficiencies of fat-soluble vitamins due to bile acid malabsorption. The primary function of vitamin D is calcium homeostasis. However, maintenance of sufficient concentrations of serum 25-hydroxyvitamin D [25(OH)D] may also be necessary for proper immune function. The purpose of this investigation is to describe the serum 25(OH)D status of a large population of children with SBS by mode of nutrition therapy (PN only, PN and enteral nutrition [EN] or full EN) as well as examine the demographic and clinical factors related to serum vitamin D sufficiency in patients who achieved enteral autonomy. Methods: Registry data for patients treated at our interdisciplinary Intestinal Care and Rehabilitation Center between 2009 and 2015 were reviewed. Severe malnutrition was determined using the -3 length/height z score primary indicator as defined by the Academy of Nutrition and Dietetics/American Society for Parenteral and Enteral Nutrition. Logistic regression was used to identify independent predictors of 25(OH)D sufficiency (normal range 25-100 ng/mL). Results: The population included 59 patients (58% male, 64% Caucasian) with a median age of 0.3 years (interquartile range [IQR], 0.1-0.7 years) who received medical management for SBS. The most common diagnoses were necrotizing enterocolitis (27%), midgut volvulus (15%), small bowel atresia (10%), and gastroschisis (10%). Median gestational age was 35 weeks (IQR, 31-39 weeks) and percent small bowel remaining at the time of diagnosis was 20% (IQR, 10-25%). 32% had a functional ileocecal valve (ICV). 49% were successfully weaned from PN (median time to wean = 1.4 years; IQR, 0.95-2.5 years) during the treatment period (median time followed = 3.4 years; IQR, 1.9-5.2 years). The serum vitamin D and supplementation status of the population by mode of nutrition therapy is shown in Table 1. Eleven patients had no 25(OH)D levels drawn while being treated at our center. Of patients in transition from PN to EN, the median percent of enteral kilocalories (including oral diet) was 48% (IQR, 30-61%). The prevalence of severe malnutrition in children receiving PN only, PN and EN, and full EN was 2 (20%), 6 (24%) and 3 (14%), respectively. Vitamin D sufficiency was not predicted by gestational age, race, bowel length, an intact ICV, presence of severe malnutrition, or duration of PN. Conclusions: Lower serum levels of 25(OH)D and a higher prevalence of vitamin D deficiency were observed in children with SBS who had achieved enteral autonomy compared with those who were receiving full PN or transitioning from PN to EN. Half of patients receiving full EN had been prescribed an oral vitamin D supplement. Vitamin D sufficiency in children receiving full EN was not predicted by their demographic or clinical characteristics. A monitoring protocol should be implemented to identify and subsequently treat vitamin D deficiency in children with SBS. Financial Support received from: none

Table 1. Serum vitamin D and supplementation status by mode of nutrition therapy

PN only (n=10)

PN and EN (n=25)

Full EN (n=22)

Serum 25(OH)D (ng/mL)* 40 (24-44.5) 35 (29.5-39) 28 (20-35) Number of patients with vitamin D deficiency (%)** 2 (20) 2 (8) 7 (32)

Number of patients receiving vitamin D supplementation (%) 2 (20) 3 (12) 11 (50) *Median (IQR), **25(OH)D


ENCORE ABSTRACT: Previously presented at Pancreas Fest. Abstract text is not published at the request of the author.

S118—Early Enteral Nutrition in Children With Acute Pancreatitis Maisam Abu-El-Haija, MD1; Rebecca J. Wilhelm, MS, RD,LD1; Christie Heinzman, NP1; Bruna Nabuco Freire Siqueria, RD1; Yuanshu Zou, PhD1; Lin Fei, PhD1; Conrad R. Cole, MD1 1Gastroenterology, Hepatology and Nutrition, Cincinnati Children's Hospital Medical Center, Cincinnati, OH, USA Financial Support received from: none

These abstracts have not been copyedited. Any errors contained within the abstracts are the sole responsibility of the authors. A.S.P.E.N. does not assume liability for any errors herein. This content does not constitute medical or other professional advice. These and all of the Clinical Nutrition Week 2016 abstracts can be accessed online at http://pen.sagepub.com/supplemental.


sunday parenteral posters s1 to s33 final€¦ · parenteral nutrition s1–s33 1 critical care...

Documents