OBJECTIVE: Although low socioeconomic status(SES) hasbeen associated with increased mortality in children with Cys-tic Fibrosis(CF), little is known regarding SES or pulmonaryoutcomes in children with CF and rhinosinusitis. Our objec-tives were to evaluate SES and long-term pulmonary functionin children with CF and rhinosinusitis, and to assess the asso-ciation of endoscopic sinus surgery (ESS) on these outcomes.METHOD: We conducted a retrospective cohort study of CFchildren ages 0-21 from an urban tertiary CF center, referred toOtolaryngology between 1998-2008. Children were analyzedbased on SES [Medicaid (MA)] or surgery status (ESS). Pul-monary function tests (PFTs): FVC% predicted, FEV1% pre-dicted] were recorded over time. Mean comparison and multi-variate regression analyses were performed to compare groupsand test the association of ESS or MA on long-term PFTs.RESULTS: Of 59 patients evaluated, 21 (36%) underwentESS, and 16 (27%) had MA. Age at evaluation and genderwere similar regardless of MA status or ESS. Polyps weremore common in the ESS group (86% vs 34%, P�0.001). At2 years, improvement in FVC% predicted, but not FEV1%predicted, was associated with ESS (P�0.056). ConcerningSES, all baseline PFTs were lower for children with MA(P�0.0001), however long-term changes were similar in so-cioeconomic groups. Multivariate analysis showed that chil-dren with MA were more likely to undergo ESS (P�0.0001).Lower FEV1% predicted at 2-year follow-up was associatedwith MA status but not with ESS (P�0.0001).CONCLUSION: ESS is not clearly associated with improvedlong-term pulmonary outcomes in children with CF and rhi-nosinusitis. Low SES, however, is associated with low baselinePFTs and may impact long-term outcomes in these children.

Pediatric CRS: Lund-Mackay Validity and Surgery

Outcome

John Manoukian, MD, FRCSC (presenter); Talal Al-Khatib, MBBS

OBJECTIVE: 1) To explore the possibility of predictingwhich of the children suffering from chronic sinusitis, whoundergo a first FESS surgery, are likely to need a secondintervention. 2) To also explore the possibility of a modifiedLund-Mackay score that takes into account undeveloped si-nuses in children.METHOD: This is a consecutive case series of 81 childrendiagnosed with chronic sinusitis and operated on at our chil-dren’s hospital between 2005 and 2008. Their medical recordsand CT scans were reviewed. The CT scans were evaluated fortheir overall sinus opacity, then graded using a modified Lund-Mackay score that takes into account the presence of undevel-oped sinuses in children. Fifteen variables, including the mod-ified Lund-Mackay score and the patient gender, weredocumented and compared.RESULTS: An AUC and regression analysis revealed sixrelevant variables that correlated with revision surgery, includ-

ing male gender, cystic fibrosis, asthma, polyps, allergies, andour modified Lund-Mackay score. A new formula to predictrevision surgery was constructed by assigning a certain numberof points per variable. This formula can predict revision sur-gery in 9 out of 10 patients, even before their first surgery, with89% sensitivity and 92% specificity. Also, the difference be-tween the classical Lund-Mackay scale and the modified onewas statistically significant with p�0.0001.

CONCLUSION: 1) A formula is introduced to project thelikelihood of revision surgery for chronic sinusitis in children.2) The role of the male gender is explored. 3) A modifiedLund-Mackay scale has been proposed in children.

Pediatric Ear Infections: Racial Disparities in Health

Care

Neil Bhattacharyya, MD (presenter); Nina Shapiro,MD; Kalpesh Vakharia, MSc, MD

OBJECTIVE: 1) Understand the impact of racial/ethnic dis-parities on access to care for children with frequent ear infec-tions. 2) Identify disparities to target for future interventions.

METHOD: The National Health Interview Survey (1997 to2006) was utilized to identify children suffering from frequentear infections (FEI), defined as 3 or more ear infections in thepreceding year. Age, sex, race/ethnicity, income level andinsurance status were extracted. Access to care was measuredby ability to afford medical care and prescription medications,specialist visitation, and emergency room visits. Multivariateanalyses determined the influence of demographic variables onthe ability to access health care resources.

RESULTS: An annualized population of 4.65 plus/minus 0.08million children reported FEI. Overall, 3.7% could not affordcare, 5.6% could not afford prescriptions, and only 25.8% sawa specialist. A larger percent of the Black (42.7%) and His-panic children (34.5%) with FEI were below the poverty levelversus White children (12.4%, p�0.001). 18.2% of Hispanicchildren were uninsured versus 6.5% of White children(p�0.001). On multivariate analysis, children with FEI thatwere Black or Hispanic had increased odds ratios relative toWhite children for (1) not being able to afford prescriptionmedications (odds ratios, 1.76 and 1.47, respectively,p�0.002), (2) not being able to see a specialist (OR, 1.62 and1.86, p�0.001) and (3) visiting the emergency room (OR, 2.50and 1.32, p�0.001).

CONCLUSION: Racial/ethnic disparities among childrenwith FEI significantly influence healthcare resource access andutilization. These disparities should be targeted for intervention.

Pediatric Tonsillectomy: Cold Steel vs Harmonic

Scalpel

Muhammad Shakeel, MBBS, MRCSEd, DOHNS(presenter); Aaron Trinidade, MRCS; AhmedAl-Adhami, MB, ChB; Haytham Kubba, MD, FRCS

P116 Otolaryngology-Head and Neck Surgery, Vol 143, No 2S2, August 2010

OBJECTIVE: 1) To learn about post-tonsillectomy morbiditybased on the technique of surgery. 2)To identify risk factorsassociated with secondary post-tonsillectomy bleeding.METHOD: Retrospective chart review of all children whounderwent harmonic scalpel and cold steel tonsillectomy withor without adenoidectomy between 2004 and 2006. Patientswere identified from theatre log book. Data included demo-graphics, indication and type of surgery, duration of hospitalstay, re-attendance and readmission, and management of com-plications. SPSS was used for data collection and analysis.RESULTS: A total of 933 (65%) patients underwent cold steeland 497 (35%) patients had harmonic scalpel tonsillectomy.There were 738 (51.6%) females and 692 (48.4%) males witha median age of 6.6 years (IQR 4.5 9.6). Main indications forsurgery were recurrent tonsillitis, obstructive sleep apnea, andobstructive symptoms such as snoring and nasal blockage.Primary hemorrhage occurred in 10 patients in the cold steelgroup and they required a general anesthetic for hemostasis.The secondary post-tonsillectomy bleeding rate was 5.5% incold steel compared to 8.0% for harmonic scalpel technique.When added to a stepwise logistic regression model, age wasthe only significant factor influencing the risk of secondarybleeding. The odds ratio indicates that the older patients aremore likely to develop secondary bleeding (OR 1.09; 95% CI,1.02 1.16).CONCLUSION: Our post-tonsillectomy bleeding rates arecomparable to the published results. Based on our data, ton-sillectomy technique does not appear to statistically influencethe secondary post-tonsillectomy bleeding risk.

Predicting Bronchoscopy Findings in Recurrent

Croup

Noel Jabbour, MD (presenter); Noah Parker, MD;Timothy Lander, MD; James Sidman, MD

OBJECTIVE: To develop an evidence-based model for pre-dicting negative bronchoscopy findings in patients with recur-rent croup.METHOD: Retrospective chart review was performed on 124patients who received consultations for recurrent croup be-tween 2000 and 2009. Bronchoscopy findings were catego-rized as normal, mildly abnormal (non-obstructing subglotticcysts or grade I subgottic stenosis), moderately abnormal(grade II subglottic stenosis), or severely abnormal (grade III toIV subglottic stenosis).RESULTS: Of the 124 consultations for recurrent croup, 81patients (average age � 3.5 years) proceeded with diagnosticbronchoscopy. Normal examinations occurred in 33/81 (41%).Abnormal findings were encountered with the following fre-quency: mildly abnormal 40/81 (49%), moderately abnormal6/81 (7.5%), and severely abnormal 2/81 (2.5%). Relative risk(RR) of either moderately abnormal or severely abnormalfindings was increased for patients who had a history of pre-

vious intubation (RR�9.8; p�.002), prematurity (RR�6.4;p�.01), or inpatient consultation (RR�3.9; p�.049). The rateof moderately or severely abnormal findings in patients with-out the risk factors of prematurity, previous intubation, orsignificant comorbidity was 0/45 (0%; CI�0 to 9.3%). How-ever, mild abnormalities in this group were encountered in23/45 (51%).CONCLUSION: A high index of suspicion for significantabnormalities should exist for patients with risk factors ofprevious intubation, prematurity, and current hospitalization.Mild airway abnormalities are common in children with recur-rent croup and cannot be ruled out based on history. However,for patients without a history of previous intubation, prematu-rity or comorbidities, the likelihood of finding a significantabnormality is quite low. A predictive model based on thisevidence is discussed.

Predictors of Prolonged PACU Stay After

Adenotonsillectomy

Eric Jaryszak, MD, PhD (presenter); Lina Lander,ScD; Anju Patel; Sukgi Choi, MD; Rahul Shah, MD,FAAP

OBJECTIVE: 1) To understand factors resulting in prolongedrecovery room times in patients undergoing outpatient adeno-tonsillectomy. 2) To identify areas of improvement within theambulatory system to facilitate efficient patient care.METHOD: One-hundred ninety consecutive patients undergo-ing adenotonsillectomy by one of two attending surgeons at anoutpatient ambulatory surgery center were enrolled. In eachpatient we measured time spent in the recovery room. Upperand lower deciles of recovery room times were isolated and 21variables were analyzed to determine predictors of prolongedrecovery time. Univariate and multivariate analyses were per-formed.RESULTS: Of the 190 patients, 22 patients were in the lowerdecile (mean recovery room time of 63 6 min) and 17 patientswere in the upper decile (155 40 min, P � 0.0001). Of the 21variables analyzed, post-anesthesia care unit (PACU) nursingstaff was the only significant predictor of prolonged recoveryroom time. Compared with one PACU nurse, other nurses(N � 5) predicted a longer recovery time (OR � 10.8, 95% CI2.0 59.5, P � 0.0017). This association remained significantwhen controlling for anesthesiologist and surgeon (OR � 8.8,95% CI 1.5 50.9, P � 0.0072). There were no patient compli-cations.CONCLUSION: Recovery room times after outpatient adeno-tonsillectomy vary significantly. Of 21 potential predictors,only the human factor (PACU nursing staff) predicts pro-longed recovery room times, independent of surgeon and an-esthesiologist. Development of standardized protocols fornurses to use for discharge has the potential to increasethroughput for adenotonsillectomy patients in an outpatientsurgery center setting.

P117Oral Presentations

OR

ALS