american journal of kidney disease

K/DOQITM DisclaimerThese guidelines are based on the best information available at the time of publication. They are designed

to provide information and assist in decision making. They are not intended to define a standard of care, andshould not be construed as one. Neither should they be interpreted as prescribing an exclusive course ofmanagement. Variations in practice will inevitably and appropriately occur when clinicians take into accountthe needs of individual patients, available resources, and limitations unique to an institution or type ofpractice. Every healthcare professional making use of these guidelines is responsible for evaluating theappropriateness of applying them in the setting of any particular clinical situation.

K/DOQI is a trademark of the National Kidney Foundation.

FOREWORD

FROM ITS RUDIMENTARY beginnings inthe 1960s, renal replacement therapy hasbecome a lifesaving treatment that can provideend-stage renal disease (ESRD) patients with agood quality of life. As a result, the number ofESRD patients who receive renal replacementtherapy has risen, and their survival has in-creased, but considerable geographic variabilityexists in practice patterns and patient outcomes.It was this realization, and the belief that substan-tial improvements in the quality and outcomes ofrenal replacement therapy were achievable withcurrent technology, that prompted several organi-zations to seek to reduce variations in ESRDtreatment with the goal of a more uniform deliv-ery of the highest possible quality of care todialysis patients. Notable among these effortswere the report on Measuring, Managing andImproving Quality in the ESRD Treatment Set-ting issued by the Institute of Medicine inSeptember 1993; the Morbidity and Mortalityof Dialysis report issued by the National Insti-tute of Diabetes, Digestive and Kidney Diseases(NIDDK) in November 1993; the Core IndicatorProject initiated by the ESRD Networks and theHealth Care Financing Administration (HCFA)in 1993; the Clinical Practice Guidelines on theAdequacy of Hemodialysis issued by the RenalPhysicians Association in December 1993; andthe Dialysis Outcomes Quality Initiative (DOQI)initiated by the National Kidney Foundation(NKF) in 1995.

In keeping with its longstanding commitmentto the quality of care delivered to all patientswith kidney and urologic diseases, the NKFconvened a Consensus Conference on Controver-sies in the Quality of Dialysis Care in March

1994. Following a series of nationwide town hallmeetings held to obtain input into the recommen-dations made at the Consensus Conference, theNKF issued an Evolving Plan for the Contin-ued Improvement of the Quality of DialysisCare in November 1994. A central tenet of theplan was recognition of an essential need forrigorously developed clinical practice guidelinesfor the care of ESRD patients that would beviewed as an accurate and authoritative reflec-tion of current scientific evidence. It was to thisend that the NKF launched the Dialysis Out-comes Quality Initiative (DOQI) in March 1995,supported by an unrestricted grant from Amgen,Inc.

The objectives of DOQI were ambitious: toimprove patient survival, reduce patient morbid-ity, improve the quality of life of dialysis pa-tients, and increase efficiency of care. To achievethese objectives, it was decided to adhere toseveral guiding principles that were consideredto be critical to that initiatives success. The firstof these principles was that the process used todevelop the DOQI guidelines should be scientifi-cally rigorous and based on a critical appraisal ofall available evidence. Such an approach was feltto be essential to the credibility of the guidelines.Second, it was decided that participants involvedin the development of the DOQI guidelines shouldbe multidisciplinary. A multidisciplinary guide-line development process was considered to becrucial, not only to the clinical and scientificvalidity of the guidelines, but also to the need formultidisciplinary adoption of the guidelines fol-lowing their dissemination, in order for them tohave maximum effectiveness. Third, a decisionwas made to give the DOQI guideline develop-

American Journal of Kidney Diseases, Vol 35, No 6, Suppl 2 (June), 2000: pp S1-S3 S1

The Official Journal of the

National Kidney FoundationVOL 35, NO 6, SUPPL 2, JUNE 2000

AJKD American Journal ofKidney Diseases

ment Work Groups final authority over the con-tent of the guidelines, subject to the requirementthat guidelines be evidence-based whenever pos-sible. By vesting decision-making authority in agroup of individuals, from multiple disciplinesand with diverse viewpoints, all of whom areexperts with highly regarded professional reputa-tions, the likelihood of developing sound guide-lines was increased. Moreover, by insisting thatthe rationale and evidentiary basis of each DOQIguideline be made explicit, Work Group partici-pants were forced to be clear and rigorous informulating their recommendations. The finalprinciple was that the guideline development pro-cess would be open to general review. Thus, thechain of reasoning underlying each guideline wassubject to peer review and available for debate.

Based on the NKF Evolving Plan for theContinued Improvement of the Quality of Dialy-sis Care and criteria recommended by theAgency for Health Care Research and Quality(AHCRQ; formerly known as the Agency forHealth Care Policy and Research [AHCPR]),four areas were selected for the initial set ofclinical practice guidelines: hemodialysis ad-equacy, peritoneal dialysis adequacy, vascularaccess, and anemia. Each Work Group selectedwhich topics were considered for guideline cre-ation. During the DOQI guideline developmentprocess, nearly 11,000 potentially relevant pub-lished articles were subjected to evaluation, andboth the content and methods of approximately1,500 articles underwent formal, structured re-view. Although labor-intensive and costly, theprocess resulted in an intensive, disciplined, andcredible analysis of all available peer-reviewedinformation. When no evidence existed, or theevidence was inadequate, guidelines were basedon the considered opinion of the Work Groupexperts. In all cases the rationale and the eviden-tiary basis of each recommendation was statedexplicitly.

Draft guidelines were then subjected to a three-stage review process. In the first stage, an Advi-sory Council, consisting of 25 experts and lead-ers in the field, provided comments on the initialdraft of the guidelines. In the second stage, avariety of organizations (ESRD Networks, profes-sional and patient associations, dialysis provid-ers, government agencies, product manufactur-ers, and managed care groups) were invited to

review and comment on a revised draft of theguidelines. After considering these commentsand suggestions, the Work Groups produced athird draft of the Guidelines. In the final stage,this draft was made available for public reviewand comment by all interested individuals orparties. Following consideration of the com-ments submitted during this open review period,the guidelines were revised again and then pub-lished as supplements to the September andOctober 1997 issues of the American Journal ofKidney Diseases was made available on theInternet and widely distributed.

The four sets of DOQI guidelines published in1997 addressed only part of the Evolving Planfor the Continued Improvement of the Quality ofDialysis Care adopted by the NKF in 1994. Inthat plan, as well as in the early DOQI prioritiza-tion process, nutrition was considered to be animportant determinant of ESRD patient out-come. Consequently, a Nutrition Work Groupwas convened in 1997 to review the key clinicalnutrition literature and to define topics for whichguidelines related to the nutritional managementof patients should be developed. Supported pri-marily by a grant from Sigma Tau Pharmaceuti-cals, Inc, the Nutrition Work Group began towork intensively on those topics in January 1998,and the Nutrition Guidelines that they have devel-oped constitute this fifth set of the original DOQIguidelines.

NKF-DOQI achieved many, but not all of itsgoals. The guidelines have been well receivedand are considered by many to reflect the stateof the art of medical practice in their fields. Thefrequency with which the DOQI guidelines havebeen cited in the literature and have served as thefocus of local, national, and international scien-tific and educational symposia is one measure oftheir influence. The guidelines also have beentranslated into more than 10 languages and havebeen adopted in countries across the globe. Inaddition, DOQI has spawned numerous educa-tional and quality improvement projects in virtu-ally all relevant disciplines, as well as in dialysistreatment corporations and individual dialysiscenters. Furthermore, the Health Care FinancingAdministration has responded to a Congres-sional mandate to develop a system for evalua-tion of the quality of care delivered in dialysiscenters by developing a series of Clinical Perfor-

S2 FOREWORD

mance Measures (CPMs) based on selected DOQIguidelines.

It is encouraging that two of the ESRD Net-works have developed a guideline prioritizationtool and embarked on a Prioritization and Imple-mentation Project that would link selected DOQIguidelines into the Health Care Quality Improve-ment Project proposed by HCFA in the ESRDNetworks most recent Scope of Work. Thisproject would involve a collaborative effort ofprofessional organizations, local practitioners,and patients. In fact, it is this collaborative spiritand total commitment to patient care that ac-counts for the success that DOQI has achievedheretofore.

As we begin the new millennium, the DOQIclinical practice guideline initiative will moveforward into a completely new phase, in whichits scope will be enlarged to encompass thespectrum of chronic kidney disease well beforethe need for dialysis, when early intervention andprevention measures can delay or prevent theneed for dialysis and improve its outcomes. Thisenlarged scope increases the potential impact ofimproving outcomes of care from hundreds ofthousands to millions of individuals with kidneydisease. To reflect this expansion, the referenceto Dialysis in DOQI will be changed to Dis-ease and the new initiative will become knownas Kidney Disease Outcomes Quality Initiative(K/DOQI).

The dissemination and implementation strate-gies that have proven so effective for NKF-

DOQI have been adapted and expanded to reflectthe new mission of K/DOQI and its multidisci-plinary focus. Relevant material from the Nutri-tion Guidelines and future K/DOQI Guidelineswill be developed into implementation tools ap-propriate not just for nephrology, but also thespecialties most likely to encounter those at riskfor chronic kidney disease early in the course oftheir illness, including cardiology, hypertension,diabetes, family practice, pediatrics, and internalmedicine.

On behalf of the National Kidney Foundation,we would like to acknowledge the tremendouscontributions of all the volunteers who gave somuch of their time and effort to the success ofDOQI in order to improve the quality of life andoutcomes of dialysis patients. The NutritionGuidelines extend the DOQI objectives evenfurther into the new and broader K/DOQI goals.Since the effort that went into preparing theNutrition Guidelines was under the aegis of theoriginal DOQI Advisory Council and SteeringCommittee, these two bodies are acknowledged.The new K/DOQI Advisory Board now willassume the charge of disseminating and imple-menting the Nutrition Guidelines.

Garabed Eknoyan, MDK/DOQI Co-Chair

Nathan W. Levin, MDK/DOQI Co-Chair

K/DOQI Advisory Board MembersGeorge Bailie, PharmD, PhDGavin Becker, MD, MBBSJerrilynn Burrowes, MSN,

RD, CDNDavid N. Churchill, MD, FACPAllan Collins, MD, FACPWilliam Couser, MDDick DeZeeuw, MDGarabed Eknoyan, MDAlan Garber, MD, PhDThomas Golper, MDFrank A. Gotch, MDAntonio Gotto, MDJoel W. Greer, PhD

Richard Grimm, Jr, MDRamon G. Hannah, MD, MSJaime Herrera Acosta, MDRonald Hogg, MDLaurence Hunsicker, MDCynda Ann Johnson, MDMichael Klag, MD, MPHSaulo Klahr, MDNathan W. Levin, MD, FACPCaya Lewis, MPHEdmund Lowrie, MDArthur Mattas, MDSally McCulloch, MSN, RN, CNNMaureen Michael, BSN, MBA

Rosa A. Rivera-Mizzoni,MSW, LCSW

Joseph V. Nally, MDJohn M. Newmann, PhD, MPHAllen Nissenson, MDKeith Norris, MDWilliam Owen, Jr, MDGlenda Payne, RNDavid SmithRobert Star, MDMichael Steffes, MD, PhDTheodore Steinman, MDProfessor John WallsNanette Wenger, MD

FOREWORD S3

K/DOQITMNUTRITION WORK GROUP MEMBERSHIP

Joel D. Kopple, MD, FACPChair

Harbor-UCLA Medical Center and UCLA Schools of Medicine and Public HealthTorrance, CA

Marsha Wolfson, MD, FACPVice-Chair, Protein-Energy GuidelinesBaxter Healthcare CorporationMcGaw Park, IL

Glenn M. Chertow, MD, MPHVice-Chair, Carnitine GuidelinesUCSF School of MedicineSan Francisco, CA

Isidro B. Salusky, MD, FAAPVice-Chair, Pediatric GuidelinesUCLA School of MedicineLos Angeles, CA

Adult Work Group Members

Suhail Ahmad, MDUniversity of WashingtonSeattle, WA

Jerrilynn D. Burrowes, MS, RD, CDNBeth Israel Medical CenterNew York, NY

David B. Cockram, MS, RD, LDAbbott LaboratoriesColumbus, OH

Charles J. Foulks, MD, FACP, FACNScott & White ClinicTemple, TX

Denis Fouque, MD, PhDUniversite Claude Bernard, Hopital E. HerriotLyon, France

Bradley Maroni, MDAmgen, Inc.Thousand Oaks, CA

Linda W. Moore, RDSangStat Medical Corporation

Germantown, TN

S4

Pediatric Work Group Members

James C.M. Chan, MDMedical College of VirginiaRichmond, VA

Richard N. Fine, MDSUNY at Stony Brook School of MedicineStonybrook, NY

Craig B. Langman, MDNorthwestern University Medical SchoolChildrens Memorial HospitalChicago, IL

Bruce Morgenstern, MDMayo ClinicRochester, MN

Pauline Nelson, RDUCLA Medical CenterLos Angeles, CA

Bradley A. Warady, MDChildrens Mercy HospitalKansas City, MO

S5

NKF-DOQI ADVISORY COUNCIL*Garabed Eknoyan, MD

Co-chairNathan W. Levin, MD, FACP

Co-chair

William H. Bennett, MDAnatole Besarab, MDWendy W. Brown, MDGlenn Chertow, MD, MPHDavid N. Churchill, MD, FACPPeter DeOreo, MDHarriet O. Ellis, MS, RPHFrank A. Gotch, MDJoel Greer, PhDRamon G. Hannah, MD, MSWilliam E. Harmon, MDPhilip J. Held, PhDSteven Helgerson, MD, MPHAlan Hull, MDKeith Johnson, MDDerrick L. Latos, MD

Edmund Lowrie, MDLinda M. McCann, RD, LDMaureen A. Michael, RNAllen R. Nissenson, MDKarl D. Nolph, MDNeil R. Powe, MDChristy Price, RN, MSNRosa A. Rivera-Mizzoni, MSW, LCSWJimmy L. Roberts, MDCharlie RodriguezMark Rolston, CHTIsidro B. Salusky, MDGary E. Striker, MDJoseph D. White, MBAMarsha Wolfson, MD

NKF-DOQI STEERING COMMITTEE*Garabed Eknoyan, MD

Co-chairNathan W. Levin, MD, FACP

Co-chair

Sally Burrows-Hudson, RNBrenda DysonMarguerite Hartigan, MSN, RN, CNNAlan Kliger, MDJoel D. Kopple, MDAndrew S. Levey, MD

Donna Mapes, DNSc, RNMaureen Michael, BSN, MBAEdith Oberley, MAEarl Steinberg, MD, MPPBradley A. Warady, MD

*Participants in the guideline review process.

S6

Acronyms and Abbreviations List

Abbreviation Term

a1-AG a1-Acid GlycoproteinaBWef Adjusted Edema-Free Body WeightAMA Arm Muscle AreaAPD Automated Peritoneal DialysisBCG Bromcresol GreenBCP Bromcresol PurpleBIA Bioelectrical Impedance AnalysisBMI Body Mass Index, also called Quetelets IndexBUN Blood Urea NitrogenCAD Coronary Artery DiseaseCANUSA Canada/United States Peritoneal Dialysis StudyCAPD Continuous Ambulatory Peritoneal DialysisCCPD Continuous Cyclic Peritoneal DialysisCoA Coenzyme ACPD Chronic Peritoneal DialysisCrCl Urinary Creatinine ClearanceCRF Chronic Renal Failure (GFR less than 20 mL/min)CRI Chronic Renal Insufficiency (GFR less than normal but greater than 20 mL/min)CRP C-Reactive ProteinCVVHD Continuous Venovenous Hemofiltration with HemodialysisDEI Dietary Energy IntakeDPI Dietary Protein IntakeDRI Dietary Reference IntakeDXA Dual Energy X-Ray AbsorptiometryESRD End-Stage Renal DiseaseGH Growth HormoneGFR Glomerular Filtration RateHD HemodialysishGH Human Growth HormoneIDWG Interdialytic Weight GainIDPN Intradialytic Parenteral NutritionIGF-I Insulin-Like Growth Factor-IIPAA Intraperitoneal Amino AcidsKt/Vurea A measure of dialysis where K is the dialyzing membrane clearance, t is the time of

dialysis delivered in minutes, and Vurea is the volume of distribution of ureaMAC Mid-Arm CircumferenceMAMA Mid-Arm Muscle AreaMAMC Mid-Arm Muscle CircumferenceMD Maintenance Dialysis (ie, maintenance hemodialysis or chronic peritoneal dialysis)MHD Maintenance HemodialysisNHANES National Health and Nutrition Evaluation SurveynPCR Protein Catabolic Rate normalized to body weightnPNA Protein Equivalent of Total Nitrogen Appearance normalized to body weightPCR Protein Catabolic RatePEM Protein-Energy Malnutrition

S7

PNA Protein Equivalent of Total Nitrogen AppearancePredialysis serum Serum obtained from an individual immediately before the initiation of a

hemodialysis or intermittent peritoneal dialysis treatmentPTH Parathyroid HormoneRD Registered DietitianRDA Recommended Dietary AllowanceREE Resting Energy ExpenditureRTA Renal Tubular AcidosisSBW Standard Body WeightSDS Standard Deviation ScoreSGA Subjective Global AssessmentStabilized serum Serum obtained for performance of a specific measurement after the measurement

has stabilized on a given dose of CAPDSUN Serum Urea NitrogenTBW Total Body WaterTNA Total Nitrogen AppearanceTPN Total Parenteral NutritionTSF Triceps Skinfold ThicknessUBW Usual Body WeightUNA Urea Nitrogen AppearanceUSRDS United States Renal Data SystemV Volume of Distribution

S8

Introduction

PROTEIN-ENERGY malnutrition (PEM) isvery common among patients with advancedchronic renal failure (CRF) and those undergo-ing maintenance dialysis (MD) therapy world-wide. Different reports suggest that the preva-lence of this condition varies from roughly 18%to 70% of adult MD patients. In adults, thepresence of PEM is one of the strongest predic-tors of morbidity and mortality. However, in thepoorly nourished pediatric patient, mortality isless common, and growth retardation is an addi-tional and greater concern. Impaired linear growthpersists despite ongoing renal replacementtherapy with either hemodialysis (HD) or perito-neal dialysis, and improvements in linear growthafter successful renal transplantation usually failto fully correct pre-existing growth retardationunless growth hormone (GH) is administered.Although several factors contribute to the im-paired skeletal growth in pediatric patients withchronic renal disease, protein and energy malnu-trition play a critical role, particularly during thefirst few years of life. Additional factors thatcontribute to impaired growth in pediatric pa-tients include anemia, acidemia, calcitriol defi-ciency, renal osteodystrophy, and tissue resis-tance to the actions of GH and insulin-like growthfactor-I (IGF-I).

There are many causes of PEM in patientswith advanced CRF. These include:

(a) inadequate food intake secondary to: anorexia caused by the uremic state altered taste sensation intercurrent illness emotional distress or illness impaired ability to procure, prepare, or

mechanically ingest foods unpalatable prescribed diets

(b) the catabolic response to superimposedillnesses

(c) the dialysis procedure itself, which maypromote wasting by removing such nutri-ents as amino acids, peptides, protein,glucose, water-soluble vitamins, and otherbioactive compounds, and may promoteprotein catabolism, due to bioincompatibil-ity

(d) conditions associated with chronic renalfailure that may induce a chronic inflam-

matory state and may promote hyperca-tabolism and anorexia

(e) loss of blood due to: gastrointestinal bleeding frequent blood sampling blood sequestered in the hemodialyzer

and tubing(f) endocrine disorders of uremia (resistance

to the actions of insulin and IGF-I, hyper-glucagonemia, and hyperparathyroidism)

(g) possibly the accumulation of endogenouslyformed uremic toxins or the ingestion ofexogenous toxins.

Notwithstanding the many causes of PEM inpatients with CRF, provision of adequate nutri-tion is a key component of the prevention andtreatment of PEM in adults and children receiv-ing MD. These K/DOQI Nutrition Clinical Prac-tice Guidelines provide recommendations regard-ing the nutritional assessment of protein-energynutritional status and the desirable dietary energyand protein intake for adults and children under-going MD. Guidelines were developed for chil-dren treated with MD concerning their nutri-tional needs for vitamins, zinc, and copper andfor their treatment with recombinant human GH.Guidelines are also provided regarding the nutri-tional intake of L-carnitine for adult MD pa-tients, the nutritional management of the nondia-lyzed adult patient with advanced CRF, and themanagement of the acutely ill pediatric and adultpatient. For logistical reasons, recommendationsfor the nutritional management of nondialyzedpediatric patients with advanced CRF were notdeveloped. The decision was made to not addressvitamin and mineral needs or the use of anabolicagents in the adult MD patient, because the scopeof the subject matter and the volume of scientificliterature was considered to be too large forinclusion in this set of guidelines.

The guidelines are based on a structured re-view of the medical literature and, where insuffi-cient evidence exists, on the expert opinion ofthe Work Group members. In each case, the

2000 by the National Kidney Foundation, Inc.0272-6386/00/3506-0201$3.00/0doi:10.1053/kd.2000.6669


guidelines are intended to serve as starting pointsfor clinical decision making, and it is empha-sized that the clinical judgment of the health carepractitioner must always be included in the deci-sion making process and the application of theseguidelines. The guidelines are not to be consid-ered as rules or standards of clinical practice. Atthe end of each guideline, recommendations aremade for research studies that may enhance thescientific evidence base concerning the subjectmatter of that guideline. In keeping with theK/DOQI objectives, it is hoped that the informa-tion provided in these guidelines and the re-search recommendations will improve the qual-ity of care provided to children and adults whohave chronic kidney disease or are receiving

chronic dialysis therapy and will stimulate addi-tional research that will augment and refine theseguidelines in the future.

The K/DOQI Nutrition Work Group expressesits indebtedness and appreciation to ThomasGolper, MD, and John Burkhart, MD, for theircontributions to Guideline 27; to Tom Greene,PhD, and Thomas Depner, MD, for their assis-tance with the development of Appendix V; toPaul Shekelle, MD, and Erin Stone, MD, for thestructured review and guidance in the guidelinedevelopment process; and to Donna Fingerhut,MSEd, for the innumerable hours she devoted tothe overall administration of the project. Theefforts and expertise of these individuals wereinvaluable.

S10 INTRODUCTION

METHODS

The Guideline Development ProcessCLINICAL PRACTICE GUIDELINES DEFINED

THE INSTITUTE OF Medicine has definedpractice guidelines as systematically devel-oped statements to assist practitioner and patientdecisions about appropriate health care for spe-cific clinical circumstances. The AmericanMedical Association endorsed this definition bydescribing practice guidelines as systematicallydeveloped statements, based on current profes-sional knowledge, that assist practitioners andpatients to make decisions about appropriatehealth care for specific clinical circumstances.Put simply, practice guidelines constitute an ef-fort to advise health-care providers and patientsas to what constitutes optimal clinical practice,based on the best information available. As aresult, practice guidelines can not only improveboth quality and cost-effectiveness of care, butcan also facilitate continuous improvement inclinical practice as new information becomesavailable.

K/DOQI GUIDING PRINCIPLESFour principles guided decision-making in the

conduct of the NKF-DOQI and will be retainedfor the K/DOQI guidelines:

1. K/DOQI practice guidelines will be devel-oped using a scientifically rigorous pro-cess, and the rationale and evidentiary ba-sis for each guideline will be clearlyexplained.

2. K/DOQI guidelines will be developed bymultidisciplinary Work Groups with exper-tise in the topic of interest.

3. The Work Group members will work inde-pendently of any organizational affiliationsand would have final responsibility for de-termining guideline content.

4. K/DOQI guidelines will undergo wide-spread critical review before being final-ized.

EVIDENTIARY BASIS FOR GUIDELINES

The guidelines were developed using an evi-dence-based approach similar to the one used byThe Federal Agency for Health Care Researchand Quality (AHCRQ). That is, before formulat-

ing recommendations, the Work Groups re-viewed all published evidence pertinent to thetopics being considered and critically appraisedthe quality and strength of that evidence. Formany issues that the Work Groups chose toaddress, there either was no pertinent literatureavailable or available evidence was flawed orweak. As a result, in many instances the WorkGroups formulated their recommendations basedon the opinions of the Work Group members andcomments received from the peer reviewers. Inall instances, the Work Groups have documentedthe rationale for their recommendations. That is,they have articulated each link in the chain oflogic they used as the evidentiary or opinion-related basis for their recommendation. This ap-proach helps readers of the guidelines determinethe quantity and quality of evidence underlyingeach recommendation.

Although some of the DOQI guidelines areclearly based entirely on evidence or entirely onopinion, many are based in part on evidence andin part on opinion. Such hybrid guidelinesarise when some (or even most) of the links inthe chain of logic underlying a guideline arebased on empirical evidence, but some (ie, atleast one) are based on opinion. The opinion ofthe Work Group members can enter the chain oflogic that supports a guideline either to fill in agap in available evidence on some scientific orclinical issue, or in the form of a value judgmentregarding what they feel is appropriate clinical prac-tice based on available evidence.Thus, many opinion-based guidelines may have substantial empiricalevidence underlying them. These guidelines weredeveloped using a seven-stage process.

Phase I: Work Group Member Selection

The DOQI Steering Committee selected aChair to lead the Adult and Pediatric NutritionWork Group and suggested names of individualswith particular expertise to serve on the WorkGroup. Final decisions on the membership of the



Work Group were made by the Work GroupChair. In recognition of the different bodies ofliterature and expertise for nutrition issues inadult and pediatric ESRD and MD patients, theWork Group Chairs appointed separate nutritionWork Groups for adult and pediatric patients.Two Vice Chairs, for protein-energy nutritionand for carnitine, were appointed for the AdultWork Group, and one Vice Chair was appointedfor the Pediatric Work Group.

Support for the Work Groups in coordinatingand performing the systematic literature review,synthesizing data abstracted from the literatureinto evidence tables, facilitation of the guidelinedevelopment process, conducting meetings ofthe Work Groups, and analyzing results of theguideline development meetings was providedby personnel from the RAND Corporation andCedars-Sinai Medical Center. Both of these insti-tutions are associated with the Southern Califor-nia Evidence-Based Practice Center.

Phase II: Targeting

The Work Groups defined the specific topicson which guidelines would focus and the specificquestions on which the systematic literaturewould focus. The following clinical questionswere formulated:

Question 1. Which of the following measuresof nutritional status best predicts patient morbid-ity/mortality (and growth rate in children) in MDpatients?

Serum albumin, serum prealbumin, anthropomet-ric measures (height, weight, skinfold thickness,body mass index [BMI], percent of normal bodyweight, percent of desirable body weight, postdialy-sis body weight), bioelectrical impedance (BIA),urea nitrogen appearance, serum creatinine and cre-atinine index, subjective global nutritional assess-ment (SGA), dietary diaries and interviews, serumcholesterol, serum transferrin, serum IGF in pediat-ric patients, protein equivalent of total nitrogen ap-pearance (PNA/PCR), prognostic nutrition index,serum acute-phase proteins (C-reactive protein), se-rum alpha-1 glycoprotein, dual energy x-ray absorp-tiometry (DXA), a combination of more than one ofthese measures.

Question 2. Which of the following measuresis the best diagnostic test for protein/energynutritional status in MD patients?

Serum albumin, serum prealbumin, anthropomet-ric measures (height, weight, skinfold thickness,

BMI, percent of normal body weight, percent ofideal body weight, postdialysis body weight), BIA,urea nitrogen appearance, serum creatinine and cre-atinine index, SGA, dietary diaries and interviews,serum cholesterol, serum transferrin, serum IGF,PNA, prognostic nutrition index, serum acute phaseproteins (C-reactive protein), serum alpha-1 glyco-protein, DXA, a combination of more than one ofthese measures.

Question 3. What is the effect of acid/basestatus on nutritional measures in MD patients?

Question 4. Which levels of intake of proteinand energy in MD patients produce the following:

The lowest morbidity/mortality, the most opti-mum changes in nutritional status using mea-sures from Question 1 above, positive nitrogenbalance, the most optimal growth in children?

Question 5. Which levels of protein and en-ergy intake in predialysis patients produce thelowest morbidity at the initiation of dialysis?(This question was included because of evidencethat nutritional status at the onset of MD therapyis a strong predictor of nutritional status andmortality during the course of MD therapy.)

Question 6. What is the energy expenditure ofMD patients during resting and other activities,and how does it compare with energy expendi-ture in normal individuals?

Question 7. Is interdialytic weight gain a goodmeasure for dietary compliance or a good prog-nostic indicator?

Question 8. Does carnitine supplementation inadult MD patients improve morbidity or mortality?

Question 9. What are the toxic/adverse effectsof L-carnitine, if any, in adult MD patients?

Question 10. Which nutritional interventionsproduce the lowest morbidity/mortality (and bestgrowth in children) or the most optimum changesin nutritional status in MD patients using mea-sures from Question 1 above?

Question 11. Does GH therapy improve growthor morbidity/mortality in pediatric MD patients?

Question 12. Does vitamin or mineral supple-mentation (exclusive of calcium, magnesium,and vitamin D) improve morbidity/mortality inpediatric MD patients?

Phase III: Literature Review, Selection,and Abstraction

A structured database search of two computer-ized bibliographic databases (MEDLINE andEMBASE) was performed with the following

S12 K/DOQI NUTRITION IN CHRONIC RENAL FAILURE

specifications: language: English and non-En-glish articles; dates: 1966 through 1997; sub-jects: human; article types: letters, editorials,reviews, case reports, and abstracts of meetingproceedings were excluded. The literature searchwas performed in collaboration with a librarianexperienced in searching computerized biblio-graphic databases and performing evidence-based systematic reviews. The Journal of RenalNutrition was hand-searched, because, at thetime, it was not indexed in the bibliographicdatabases listed above. Additionally, referralsfrom DOQI Work Group members through Au-gust 1999 were reviewed.

After loading articles from MEDLINE, EM-BASE, Work Group referrals, and the Sigma Taubibliography into an electronic database, onereviewer performed an initial title review ofthese articles. Two independent reviewers thenreviewed the abstracts of articles whose titleswere selected. Selection disagreements were re-solved by consensus. English language articlesfor which the abstracts were selected were thenobtained and categorized based on the clinicalquestion the article addressed. Two independentreviewers then reviewed these articles. Informa-tion was abstracted from the articles (see below)by one abstracter and verified by a second. Dis-agreements were resolved by consensus. Articlesthat were rejected at this stage were coded usingthe following codes:

R1: Editorial, letter, review, case report, ar-ticle published as abstracts

R2: Article does not answer clinical questionof interest

R3: Article does not have study design ofinterest

R4: Pediatric article (if adult section)R5: Not humanR6: Adult article (if pediatric section)In order to increase precision and reduce sys-

tematic errors, the language of manuscripts wasnot limited to English.1,2 The English titles andEnglish abstracts of foreign language articles,when available, were sent to all Work Groupmembers for review. The abstracts of foreignlanguage manuscripts were translated into En-glish if any Work Group member thought that thepaper might contribute positively to the evidencebase. Selections were further based on studydesign. For prognostic articles, only those withprospective cohort or historical prospective co-

hort designs were included for further analysis.For assessment of nutritional status, only manu-scripts in which a nutritional parameter wascompared to a recognized standard nutritionalmeasure or to a clinical outcome were includedfor further analysis. For manuscripts examiningnutritional treatment, only those with a prospec-tive design with concurrent controls were ana-lyzed further. Because there were smaller num-bers of these types of studies for carnitinetreatment or pediatric renal nutrition, these re-quirements were not as rigidly applied for thisliterature.

After article abstraction (see below), evidencetables were produced from a subset of abstracteddata elements and evaluated by the Work Groupduring meetings in Los Angeles in August 1998(Adult Work Group), in October 1998 (PediatricWork Group), and during a series of subsequentconference calls. The Work Group accepted orrejected articles based on the study design andmethods and the adequacy with which it ad-dressed the clinical questions. The final selectedarticles are indicated by an asterisk in the refer-ence section. Other citations, that are not aster-isked, were not used for guideline development,but were used to more fully explain the back-ground or rationale for a guideline.

Critical Appraisal Method for Articles Con-cerning Prognosis. For each prognostic article,the following characteristics were ascertained3:(1) the study type; (2) the three main co-morbidconditions; (3) whether there was a representa-tive and well-defined sample of patients at asimilar phase in the course of disease; (4) thecharacteristics of the study population and dialy-sis procedures that might have affected the studyresults; (5) the duration of the follow-up period;(6) whether the outcomes were objective and theinterpretation of the outcomes was unbiased; (7)whether adjustment was made for importantknown prognostic factors; and (8) the results ofthe study.

Critical Appraisal Methods for Articles Con-cerning Nutritional Assessment. For each ar-ticle concerning nutritional assessment, the fol-lowing information was obtained4,5: (1) the typeof study; (2) the three main co-morbid condi-tions; (3) whether there was an independentblinded comparison with a reference (gold) stan-dard; (4) the characteristics of the study popula-tion and the dialysis procedures that might have

K/DOQI NUTRITION IN CHRONIC RENAL FAILURE S13

affected the study results; (5) whether the resultsof the nutritional measure that was studied influ-enced the decision to measure the referencestandard; (6) whether characteristics and varietyof the patients standard is similar to those foundin dialysis centers; (7) whether the test methodol-ogy are described well enough to be reproduc-ible; and (8) the results of the study.

Critical Appraisal Methods for Articles Con-cerning Nutritional Treatment. For each treat-ment article, the following information was ob-tained6,7: (1) the type of study; (2) the three mainco-morbid conditions; (3) the Jadad qualityscores8; (4) the randomization score; (5) thedouble blind score; (6) the score for whether allpatients were accounted for; (7) an intention-to-treat score; (8) whether the treatment groupswere similar at baseline; (9) the characteristics ofthe study population, dialysis procedure, andother ancillary treatment that might have affectedthe study results; (10) whether the treatmentgroups were treated similarly except for thestudy intervention; and (11) the results of thestudy.

The Jadad quality scores address issues mostimportant in demonstrating the validity of ran-domized clinical trials and have been demon-strated to reflect methodological quality. Empiri-cal evidence demonstrates that when these qualityfeatures are not met in clinical trials, bias and anexaggeration of the effect sizes often result.8-12

Results of the Systematic Review. The initialliterature search identified 19,272 MEDLINEand 4,943 EMBASE articles. In addition, theWork Groups referred 134 articles for review,and the Sigma Tau Pharmaceutical Corporationsubmitted a bibliography that contained 138 ad-ditional references that were included in theanalysis. Of these 24,487 references, 22,362 titleswere rejected as not meeting the inclusion crite-ria, leaving 2,125 titles. Abstracts of these ar-ticles were reviewed and 1,021 were rejected asnot meeting the inclusion criteria, thus leaving1,104 articles. One hundred and seventy of thesewere foreign language articles whose titles andabstracts were sent to the Adult or PediatricWork Groups. Of these, 102 were not selectedfor further evaluation. Two were selected butcould not be translated, and 66 were furtherevaluated. Of the remaining 1,000 manuscripts(including the 66 mentioned above), 29 were

unobtainable, leaving 971 to be abstracted. Ofthese, 640 were rejected because they were clas-sified as an editorial, letter, review, case report,or abstract, did not answer a clinical question ofrelevance, did not have a valid study design, ordid not involve humans. The remaining 331articles were sent to the Adult or Pediatric WorkGroups along with evidence tables for thesearticles created from the abstraction forms. TheWork Groups rejected 81 additional articles forone or more of the same reasons indicated above,leaving 250 accepted articles.

Phase IV: Formulation of Guidelines

The group process used to develop the guide-lines is a modification of the RAND/UCLA Ap-propriateness Method. This group process methodhas the following essential features: multidisci-plinary, iterative, quantitative, and each panelisthas equal weight in determining the final result.13

In conjunction with the Work Groups, RANDand Cedars-Sinai staff developed draft guide-lines based on the results of the systematic re-view. The draft guidelines corresponded to thekey questions developed by each Work Group.The draft guidelines included all possible topicsarticulated by the Work Groups during the target-ing phase and at the Work Group meetings todiscuss the evidence. These draft guidelines werethen transmitted to the Work Group members,who used the evidence tables and their expertjudgment to rate each guideline statement forvalidity on a 1-to-9 scale. The RAND staff thencompiled summaries for the face-to-face meet-ings of the Work Groups. At these meetings,Work Group members were provided with thesummaries of these first round ratings of validity.These summary ratings were used to key a point-by-point discussion of the evidence and opinionsurrounding each potential guideline statement.After each discussion, the Work Group membersprivately re-rated each guideline statement forvalidity. These votes form the basis for the finalguidelines. Statements were accepted as valid ifthe median panel rating on validity was 7 orgreater on the 1-to-9 scale. Complete agree-ment was defined as occurring when all WorkGroup members rated a guideline statementwithin the same three-point range of the scale(for example, all members ratings were in therange of 7, 8, or 9). After determining the final


guideline statements, Work Group members wentthrough a similar two-step rating process toassess the level of evidence. A rating of Evi-dencewas defined as mainly convincing scien-tific evidence, limited added opinion; Opin-ion was defined as mainly opinion, limitedscientific evidence; and Evidence plus Opin-ion was defined as about equal mixtures ofscientific evidence and opinion.

Phase V: Draft Report With Supporting Rationale

Following the development of the guidelines,the Work Group drafted a report that included thesupporting rationale for each guideline. Whilewriting the rationale for each guideline, WorkGroup members cited additional references thathad either not been identified previously in theliterature search efforts, or had been identifiedbut rejected. These citations contained informa-tion that was felt to be important either as back-ground material, or to further explain the ratio-nales. However, these additional references werenot part of the evidence base that was used toeither formulate the guideline statements or thevotes on the validity or the rating of evidenceversus opinion for each guideline.

Phase VI: Peer Review

The purpose of the peer review process was toidentify: unclear wording in the draft guidlines substantive concerns regarding the content

of specific guidelines important but uncited data relevant to spe-

cific draft guidelines guidelines that may be difficult to imple-

ment or that would benefit from specificstrategies to facilitate compliance such aseducational programs, tools, etc.

The nutrition guidelines were subjected to athree-stage peer review process:

Stage One: Primary Review. NKF-DOQIsmultidisciplinary Steering Committee was as-signed to review the draft report. Drafts weredistributed to the committee in August 1999 andmembers had the opportunity to offer oral com-ments at a face-to-face meeting in mid-Septem-ber. The draft report was also sent to the NKF-DOQI Advisory Council, the NKF ScientificAdvisory Board, and selected experts in the field.Many substantive comments were received, and

this resulted in substantive changes in the organi-zation and content of some of the guidelines andrationales. Given the large volume of commentsreceived, the Work Group vice-chairs reviewedthe comments first and entered them into a com-puter database separating these according towhether they had a potential minor or substan-tive impact. Comments were sorted by guidelinetopic and then provided to the Work Groups foranalysis and response.

Stage Two: Organizational Review. Close to200 individuals representing nearly 50 end-stagerenal disease (ESRD)-related organizations re-viewed the second draft of the guidelines inDecember 1999. Organizations that were invitedto participate in the second round of peer reviewwere selected by the Steering Committee basedon suggestions from the Advisory Council andthe Work Groups. Organizations included vari-ous nephrology professional societies (eg, RenalPhysicians Association, American Society of Ne-phrology, American Nephrology Nurses Associa-tion, and American Renal Administrators Asso-ciation), the American Association of KidneyPatients, the ESRD Networks, NKF Councils,dialysis chains, managed care organizations, andprivate industry organizations selected their ownreviewers.

Stage Three: Open Review. In the final roundof review, in December 1999, approximately 400individuals received copies of the revised draftguidelines. Within 3 weeks, 30% of these review-ers provided comments. The Work Group vice-chairs sorted and organized these comments andthe Work Group analyzed the responses.

Phase VII: Issue Final Guidelines

The Work Group and staff performed severaltasks to complete the guidelines. The guidelineswere edited to ensure clarity and consistency.The Work Group carefully reviewed the finaldraft and made the indicated changes. Accuracyof the literature citations for each guideline docu-ment were also verified.

K/DOQI IMPLEMENTATION PLANNINGThe NKF plans to undertake three types of

activities to promote implementation of theserecommendations.

1. Translating recommendations into prac-tice. K/DOQI will develop core patient and

K/DOQI NUTRITION IN CHRONIC RENAL FAILURE S15

professional education programs and toolsto facilitate the adoption of their recommen-dations.

2. Building commitment to reducing practicevariations. K/DOQI will work with provid-ers and insurers to clarify the need for andthe benefits of changes in practice patterns andto encourage the adoption of the guidelines.

3. Evaluation. K/DOQI, in collaboration withother relevant organizations, will partici-pate in the development of performancemeasures that can be used to assess compli-ance with the K/DOQI practice guidelines.

In addition, the association between compli-ance with the K/DOQI guidelines and pa-tient outcomes will be evaluated in an ef-fort to validate and improve the guidelinesover time.

The development of the K/DOQI practice guide-lines is a cooperative, rewarding, and unifying effortfor the participants and the community of health careworkers who are involved in the care of the indi-vidual with kidney disease. We hope this spirit ofcooperation and commitment to improvement ofdialysis patient outcomes will help the K/DOQI inefforts to put its quality improvements into practice.


I. ADULT GUIDELINES

A. MAINTENANCE DIALYSIS

1. Evaluation of Protein-Energy Nutritional Status

R A T I O N A L E

Optimal monitoring of protein-energy nutri-tional status for maintenance dialysis (MD) pa-tients requires the collective evaluation of mul-tiple parameters, particularly using measures thatassess different aspects of protein-energy nutri-tional status. No single measure provides a com-plete overview of protein-energy nutritional sta-tus. Each of the valid indicators described inGuidelines 2 and 23 has a role in the overallnutritional assessment of dialysis patients.

There are ample data suggesting that comple-mentary indicators of nutritional status exhibitindependent associations with mortality and mor-bidity in maintenance hemodialysis (MHD)and chronic peritoneal dialysis (CPD) patients.For example, the serum albumin, serum creati-nine, and body weight-for-height are indepen-dently associated with survival.14 Data from the

USRDS confirm these findings, using the serumalbumin and body mass index (BMI; kg/m2).15 Inthe CANUSA study, both the serum albumin andSGA were independent predictors of death ortreatment failure.16 A discussion of why serumtransferrin concentrations and bioelectrical im-pedance studies are not recommended for thenutritional assessment of MD patients in clinicalpractice is given in Appendix VIII.

RECOMMENDATIONS FOR RESEARCH

1. Studies are needed to determine the mosteffective combination of measures of nutritionalstatus for evaluating protein-energy malnutri-tion.


GUIDELINE

1

Use of Panels of Nutritional Measures

Nutritional status in maintenance dialysis patients should be assessedwith a combination of valid, complementary measures rather than anysingle measure alone. (Opinion)

There is no single measure that provides a comprehensive indication ofprotein-energy nutritional status. Measures of energy and protein intake, visceral protein pools, musclemass, other dimensions of body composition, and functional status identifydifferent aspects of protein-energy nutritional status. Malnutrition may be identified with greater sensitivity and specificityusing a combination of factors.


R A T I O N A L EThe advantages to using these individual nutri-

tional measures are discussed in Guidelines 3and 8 through 10 and in Appendices III, V, andVII. The combination of these measurementsprovides an assessment of visceral and somaticprotein pools, body weight and hence fat mass,and nutrient intake.

Serum albumin is recommended for routinemeasurement because there is a large body ofliterature that defines the normal serum albuminvalues, characterizes the nutritional and clinicalfactors affecting serum albumin concentrations,and demonstrates the relationship between se-rum albumin concentrations and outcome. Bodyweight, adjusted for height, is proposed becauseof the clear association between body weight andbody fat mass and because body weight is corre-lated with clinical outcome. SGA is recom-mended because it gives a comprehensive over-

view of nutritional intake and body composition,including a rough assessment of both musclemass and fat mass, and because it is correlatedwith mortality rates. Assessment of nutrient in-take is essential for assessing the probability thata patient will develop PEM, for evaluating thecontribution of inadequate nutrient intake to ex-isting PEM, and for developing strategies toimprove protein-energy nutritional status. Also,nutrient intake is correlated with clinical out-come. nPNA provides an independent and lesstime consuming assessment of dietary proteinintake (DPI). Dietary interviews and diaries canbe used to assess intake not only of protein andenergy but also of a variety of other nutrients aswell as the pattern and frequency of meals (infor-mation that may aid in identifying the cause ofinadequate nutrient intake). A low predialysis orstabilized serum urea level may indicate a lowintake of protein or amino acids.


1. Research is necessary to identify and vali-date the following:

(a) The optimal panel of measures to screenfor disorders in nutritional status.

*A predialysis serum measurement is obtained from anindividual immediately before the initiation of a hemodialy-sis or intermittent peritoneal dialysis treatment. A stabilizedserum measurement is obtained after the patient has stabi-lized on a given dose of CAPD.

GUIDELINE

2

Panels of Nutritional Measures for Maintenance Dialysis Patients

For maintenance dialysis patients, nutritional status should be rou-tinely assessed by predialysis or stabilized* serum albumin, percent ofusual body weight, percent of standard (NHANES II) body weight,subjective global assessment, dietary interviews and diaries, and nPNA.(Opinion)

These parameters should be measured routinely (as indicated in Table 1)because they provide a valid and clinically useful characterization of theprotein-energy nutritional status of maintenance dialysis patients


(b) The optimal panel of measures for a com-prehensive assessment of nutritional status.

(c) The optimal frequency with which thesenutritional measures should be employed.

2. More information is needed concerning theappropriate parameters to be used for assessmentof body composition (eg, for expressing dual

energy x-ray absorptiometry [DXA] measure-ments, anthropometry, and the creatinine index).

3. Patient subgroups should be identified (eg,elderly, obese, severely malnourished, or physi-cally very inactive individuals) for whom the useof specialized combinations of body composi-tion measures are beneficial.

Table 1. Recommended Measures for Monitoring Nutritional Status of Maintenance Dialysis Patients

Category MeasureMinimum Frequency of

Measurement

I. Measurements that should beperformed routinely in all patients

Predialysis or stabilized serumalbumin

% of usual postdialysis (MHD) orpost-drain (CPD) body weight

Monthly

Monthly

% of standard (NHANES II) bodyweight

Every 4 months

Subjective global assessment (SGA) Every 6 months Dietary interview and/or diary Every 6 months nPNA Monthly MHD; every 3-4 months

CPDII. Measures that can be useful to

confirm or extend the data obtainedfrom the measures in Category I

Predialysis or stabilized serum pre-albumin

Skinfold thickness Mid-arm muscle area, circumfer-

ence, or diameter Dual energy x-ray absorptiometry

As needed

As needed As needed

As needed

III. Clinically useful measures, which,if low, might suggest the need fora more rigorous examination ofprotein-energy nutritional status

Predialysis or stabilized serumCreatinineUrea nitrogenCholesterol

Creatinine index

As needed As needed As needed

ADULT GUIDELINES S19

R A T I O N A L E

Serum albumin levels have been used exten-sively to assess the nutritional status of individu-als with and without chronic renal failure (CRF).17Malnutrition is common in the end-stage renaldisease (ESRD) population,18 and hypoalbumin-emia is highly predictive of future mortality riskwhen present at the time of initiation of chronicdialysis as well as during the course of mainte-nance dialysis (MD).14,19-27 It follows that nutri-tional interventions that maintain or increaseserum albumin concentrations may be associatedwith improved long-term survival, although thishas not been proven in randomized, prospectiveclinical trials. Serum albumin levels may fallmodestly with a sustained decrease in dietaryprotein and energy intake and may rise withincreased protein or energy intake.28 Conversely,serum albumin levels may fall acutely with in-

flammation or acute or chronic stress and in-crease following resolution or recovery.

Despite their clinical utility, serum protein (eg,albumin, transferrin, and prealbumin) levels maybe insensitive to changes in nutritional status, donot necessarily correlate with changes in othernutritional parameters, and can be influenced bynon-nutritional factors.29-32 Some of these non-nutritional factors, which are frequently presentin this population, include infection or inflamma-tion, hydration status, peritoneal or urinary albu-min losses, and acidemia.33-36 Hence, hypoalbu-minemia in MD patients does not necessarilyindicate protein-energy malnutrition (PEM). Thepatients clinical status (eg, comorbid conditions,dialysis modality, acid-base status, degree ofproteinuria) must be examined when evaluatingchanges in the serum albumin level. Serum albu-min concentrations are inversely correlated withserum levels of positive acute-phase pro-

GUIDELINE

3

Serum Albumin

Serum albumin is a valid and clinically useful measure of protein-energy nutritional status in maintenance dialysis (MD) patients.(Evidence)

The predialysis or stabilized serum albumin is a measure of visceralprotein pool size. The serum albumin at the time of initiation of chronic dialysis therapy orduring the course of maintenance dialysis is an indicator of future mortalityrisk. A predialysis or stabilized serum albumin equal to or greater than thelower limit of the normal range (approximately 4.0 g/dL for the bromcresolgreen method) is the outcome goal. Individuals with a predialysis or stabilized serum albumin that is lowshould be evaluated for protein-energy malnutrition. The presence of acute or chronic inflammation limits the specificity ofserum albumin as a nutritional marker.


teins.33,34,37 An elevated C-reactive protein hasbeen reported to negate the positive relationshipbetween serum albumin and nPNA.34 However,some studies suggest that serum albumin is inde-pendently affected by both inflammation andnutritional intake.34

As indicated above, positive acute-phase pro-teins (eg, C-reactive protein [CRP], alpha-1 acidglycoprotein [a1-AG], ferritin, and ceruloplas-min) are not nutritional parameters but may beused to identify the presence of inflammation38in individuals with low serum albumin or preal-bumin (Guideline 4) levels and possibly forpredicting outcome. a1-AG may be more spe-cific than CRP for detecting inflammation in MDpatients.37 Serial monitoring of serum concentra-tions of positive acute-phase proteins (CRP, a1-AG) during episodes of inflammation in MDpatients indicate that serum levels follow pat-terns similar to those found in acutely ill individu-als who do not have CRF.39

Although no single ideal measure of nutri-tional status exists, the serum albumin concentra-tion is considered to be a useful indicator ofprotein-energy nutritional status in MD patients.The extensive literature, in individuals with or

without renal failure, relating serum albumin tonutritional status, and the powerful associationbetween hypoalbuminemia and mortality risk inthe MD population, strongly support this conten-tion. In addition, the measurement of serumalbumin levels is inexpensive, easy to perform,and widely available. Methods for measuringserum albumin are discussed in Appendix I.


1. More information is needed concerning therelative contributions of nutritional intake andinflammatory processes to serum albumin concen-trations.

2. There is a need for a better understanding ofthe mechanisms by which hypoalbuminemia orthe factors causing hypoalbuminemia lead toincreased morbidity and mortality in MD pa-tients.

3. Studies are needed to assess whether andunder what conditions nutritional interventionincreases serum albumin concentrations in hypo-albuminemic MD patients.

4. Will an increase in serum albumin levelsinduced by nutritional support reduce morbidityand mortality in persons undergoing MD?


R A T I O N A L ESerum prealbumin (transthyretin) has been used

in individuals with or without CRF as a marker ofprotein-energy nutritional status.40 It has been sug-gested that serum prealbumin may be more sensitivethan albumin as an indicator of nutritional status,since it has a shorter half-life than albumin (2 to 3days versus 20 days, respectively).25,41 However,prealbumin is limited by many of the same factorsdescribed for albumin. Prealbumin may not correlatewith changes in other nutritional parameters31,32 andit is a negative acute-phase reactant (ie, serum levelsdecline in response to inflammation or infection43).In addition, recommendations for the routine use ofserum prealbumin levels as a marker are temperedby the fact that prealbumin levels are increased inrenal failure, presumably due to impaired degrada-tion by the kidney.17,42 Although fewer studies havebeen published relating prealbumin levels to out-comes in MD patients than have been publishedregarding albumin levels, several studies have dem-onstrated that prealbumin levels less than 30 mg/dLare associated with increased mortality risk andcorrelate with other indices of PEM.25,41,42a,44

Based on available evidence, serum prealbumin is

considered to be a valid measure of protein-energynutritional status in individuals undergoing MD.There is insufficient evidence to conclude that preal-bumin is a more sensitive or accurate index ofmalnutrition than is serum albumin. If the predialysisor stabilized serum prealbumin level is used tomonitor nutritional status, it is recommended that theoutcome goal for prealbumin is a value greater thanor equal to 30 mg/dL.


1. What range of serum prealbumin concentra-tions is associated with optimal outcome?

2. More information is needed concerning therelative contributions of nutritional intake and inflam-matory processes to serum prealbumin levels.

3. Data are needed concerning the mecha-nisms by which low serum levels of prealbuminlead to increased mortality in MD patients.

4. Will nutritional intervention in malnour-ished hypoprealbuminemic MD patients increaseserum prealbumin concentrations?

5. Will an increase in serum prealbumin levelsinduced by nutritional support reduce morbidityand mortality in individuals undergoing MD?

GUIDELINE

4

Serum Prealbumin

Serum prealbumin is a valid and clinically useful measure of protein-energy nutritional status in maintenance dialysis (MD) patients.(Evidence and Opinion)

The predialysis or stabilized serum prealbumin is a measure of visceralprotein pool size. The serum prealbumin level at the time of initiation of dialysis or duringmaintenance dialysis is an indicator of future mortality risk. An individual with predialysis or stabilized serum prealbumin less than30 mg/dL should be evaluated for protein-energy malnutrition. The presence of acute or chronic inflammation limits the specificity ofserum prealbumin as a nutritional marker. There is insufficient evidence to conclude that prealbumin is a moresensitive index of nutritional status than albumin.


R A T I O N A L E

In MHD patients with little or no renal func-tion who are receiving a constant dose of dialy-sis, the predialysis serum creatinine level will beproportional to dietary protein (muscle) intakeand the somatic (skeletal muscle) mass.17,45,46 Inchronic peritoneal dialysis (CPD) patients withlittle or no residual renal function, the stabilizedserum creatinine level with a given dialysis dosewill be proportional to skeletal muscle mass anddietary muscle intake. Thus, a low predialysis orstabilized serum creatinine level in an MD pa-tient with negligible renal function suggests de-creased skeletal muscle mass and/or a low di-etary protein intake (DPI).17 Among nonanuricindividuals, this relationship persists, but themagnitude of the urinary creatinine excretionmust be considered when interpreting the predi-alysis or stabilized serum creatinine as a nutri-tional parameter. This is particularly relevant to

CPD patients, who are more likely to maintainresidual renal function for longer periods.

The creatinine index is used to assess creati-nine production and, therefore, dietary skeletalmuscle protein intake and muscle mass. Thecreatinine index estimates fat-free body massrather accurately in individuals with ESRD.46,48Appendix II discusses creatinine metabolism ingreater detail and describes methods for calculat-ing the creatinine index and, from this value, thefat-free body mass.

In individuals in whom loss of skeletal musclemass is suspected on the basis of low or declin-ing serum creatinine levels, this observation maybe confirmed using the creatinine index. Directrelationships between serum creatinine and theserum albumin29,33,42a and prealbumin concen-trations42a are reported. Among individuals under-going CPD, the creatinine index is lower inindividuals with protein-energy malnutrition asdetermined by a composite nutritional index.30

GUIDELINE

5

Serum Creatinine and the Creatinine Index

The serum creatinine and creatinine index are valid and clinicallyuseful markers of protein-energy nutritional status in maintenancedialysis (MD) patients. (Evidence and Opinion)

The predialysis or stabilized serum creatinine and the creatinine indexreflect the sum of dietary intake of foods rich in creatine and creatinine (eg,skeletal muscle) and endogenous (skeletal muscle) creatinine productionminus the urinary excretion, dialytic removal, and endogenous degradationof creatinine. Individuals with low predialysis or stabilized serum creatinine (less thanapproximately 10 mg/dL) should be evaluated for protein-energy malnutri-tion and wasting of skeletal muscle. A low creatinine index and, in the absence of substantial endogenousurinary creatinine clearance, a low serum creatinine concentration suggestlow dietary protein intake (DPI) and/or diminished skeletal muscle massand are associated with increased mortality rates.


Serum creatinine and the creatinine index arepredictors of clinical outcome. In individualsundergoing maintenance HD (MHD), predialysisserum creatinine14,25,42,44,45,49-52 and the molarratio of serum urea to creatinine are both predic-tive of and inversely related to survival. Thisrelationship persists even after adjusting for pa-tient characteristics (age, sex, diagnosis, anddiabetic status) and dialytic variables.14,25,44,45,50,52The serum creatinine at the onset of MHD distin-guishes between short-term ( 12 months) andlong-term ( 48 months) survival in incidentpatients.25 In longitudinal studies of PD patients,initial serum creatinine levels are inversely re-lated to mortality.25,44,52 The creatinine index isdirectly related to the normalized protein equiva-lent of total nitrogen appearance (nPNA) andindependent of the dialysis dose (Kt/Vurea).53 Alow or declining creatinine index correlates withmortality independently of the cause of death,although people with catabolic diseases mayhave larger and faster declines in the creatinineindex before death.53 Some research has notshown a clear association between the serumcreatinine concentration and outcome.23,42,54

The serum creatinine concentration that indi-cates malnutrition has not been well defined.The mortality risk associated with low serumcreatinine increases at levels below 9 to 11 mg/dLin individuals on MHD or PD.14,25,30,44,51 Inindividuals with negligible urinary creatinineclearance (CrCl), the nutritional status of indi-viduals undergoing MHD or CPD who havea predialysis or stabilized serum creatinine of

less than approximately 10 mg/dL should beevaluated.


1. The degree of correlation of the serumcreatinine and creatinine index with skeletalmuscle mass and DPI, and the sensitivity tochange in these parameters of creatinine metabo-lism, need to be better defined.

2. The relationship between the creatinine in-dex and the edema-free lean body mass or skel-etal muscle protein mass needs to be defined forESRD patients.

3. The rate of creatinine degradation in ESRDpatients needs to be defined more precisely.

4. The level of serum creatinine and the creati-nine index associated with optimal nutritionalstatus and lowest morbidity and mortality ratesneed to be defined.

5. The relationships between other markers ofprotein-energy nutritional status (eg, serum albu-min, prealbumin, or anthropometry) and serumcreatinine or creatinine index are limited, some-what contradictory, and need to be further exam-ined.

6. Whether nutritional interventions that in-crease serum creatinine or creatinine index willimprove morbidity or mortality in malnourishedMD patients should be tested.

7. The effects of age, gender, race, and size ofskeletal muscle mass on the relationship betweenthe serum creatinine and the creatinine index onmorbidity and mortality need to be examined.


R A T I O N A L EThe predialysis or stabilized serum cholesterol

concentration may be a useful screening tool fordetecting chronically inadequate protein-energyintakes. Individuals undergoing MHD who havea low-normal (less than approximately 150 to180 mg/dL) nonfasting serum cholesterol havehigher mortality than do those with highercholesterol levels.14,25,47,50,55 As an indicator ofprotein-energy nutritional status, the serum cho-lesterol concentration is too insensitive and non-specific to be used for purposes other than fornutritional screening, and MD patients with se-rum cholesterol concentrations less than approxi-mately 150 to 180 mg/dL should be evaluated fornutritional deficits as well as for other comorbidconditions.

Serum cholesterol is an independent predictorof mortality in MHD patients.14,19,47,55 The rela-tionship between serum cholesterol and mortal-ity has been described as either U-shaped orJ-shaped, with increasing risk for mortality asthe serum cholesterol rises above the 200 to 300mg/dL range14 or falls below approximately 200mg/dL.19,25,47,50 The mortality risk in most stud-

ies appears to increase progressively as the se-rum cholesterol decreases to, or below, the nor-mal range for healthy adults (200 mg/dL).14,19,25,50,55 Not all studies of MHD patientsshow that serum cholesterol levels predict mortal-ity, however.19,23,42 The relationship between lowserum cholesterol and increased mortality is notobserved in the CPD population,14,25,42,44,52 possi-bly because sample sizes in studies of individu-als undergoing CPD are smaller and possibly dueto confounding by greater energy (glucose in-take) and/or hypertriglyceridemia. In one study,higher serum cholesterol concentrations (250mg/dL) were associated with increased mortalityin CPD patients.56

Predialysis serum cholesterol is generally re-ported to exhibit a high degree of collinearitywith other nutritional markers such as albumin,42prealbumin,42 and creatinine,44 as well as age.44In MHD patients, the predialysis serum choles-terol level measured may be affected by non-nutritional factors. Cholesterol may be influ-enced by the same comorbid conditions, such asinflammation, that affect other nutritional mark-ers (eg, serum albumin).42 In one study there was

GUIDELINE

6

Serum Cholesterol

Serum cholesterol is a valid and clinically useful marker of protein-energy nutritional status in maintenance hemodialysis patients. (Evi-dence and Opinion)

Low or declining serum cholesterol concentrations are predictive ofincreased mortality risk. Hypocholesterolemia is associated with chronic protein-energy deficitsand/or the presence of comorbid conditions, including inflammation. Individuals with low, low-normal (less than approximately 150 to 180mg/dL), or declining serum cholesterol levels should be investigated forpossible nutritional deficits.


no difference in serum cholesterol in CAPDpatients whose serum albumin level was lessthan 3.5 g/dL as compared with those with levels3.5 g/dL.33


1. What are the conditions under which serumcholesterol is a reliable marker of protein-energynutrition? What can be done to increase thesensitivity and specificity of the serum choles-terol as an indicator of protein-energy nutritionalstatus?

2. The relationships between other markers ofprotein-energy nutritional status (eg, serum albu-min or anthropometry) and serum cholesterol arelimited, somewhat contradictory, and need to bebetter defined.

3. How does nutritional intervention in mal-nourished MD patients affect their serum choles-terol concentrations?

4. Recent data suggest that serum cholesterolexhibits a negative acute-phase response to in-flammation.42 The relationship among serum cho-lesterol, nutritional status, and inflammation needsto be further investigated.

5. Why does mortality increase when the se-rum cholesterol falls outside the 200 to 250mg/dL range?

6. More information is needed about the pat-terns of morbidity and mortality associated withabnormal serum cholesterol concentrations inMD patients. For example, in these individuals,is cardiovascular mortality directly related to theserum cholesterol level and are malnutrition andmortality from infection inversely related to theserum cholesterol level?

7. Additional data investigating the relation-ships among serum cholesterol, protein-energynutritional status, morbidity, and mortality areneeded for persons undergoing CPD.


R A T I O N A L EPatients undergoing MHD or CPD frequently

have low protein and energy intake. Evidenceindicates that for patients ingesting low proteinor energy intakes, increasing dietary protein orenergy intake improves nutritional status.57-60 Itis important, therefore, to monitor the dietaryprotein and energy intake of MHD and CPDpatients. A number of studies in individuals with-out renal disease indicate that dietary diaries andinterviews provide quantitative information con-cerning intake of protein, energy, and other nutri-ents.61,62 It is recommended, therefore, that indi-viduals undergoing MHD or CPD periodicallymaintain 3-day dietary records followed by di-etary interviews conducted by an individualtrained in conducting accurate dietary interviews

and calculating nutrient intake from the diariesand interviews, eg, a registered dietitian, prefer-ably with experience in renal disease (see Appen-dices III and IV). When staffing conditions limitthe time available to conduct more formal assess-ments of nutritional intake, a 24-hour dietaryrecall may be substituted for dietary interviewsand/or diaries in nutritionally stable patients.


1. Techniques to improve the reliability andprecision of dietary interviews or diaries for MDpatients are needed.

2. Other less laborious and more reliable meth-ods to estimate nutrient intake, particularly en-ergy intake, are needed.

GUIDELINE

7

Dietary Interviews and Diaries

Dietary interviews and/or diaries are valid and clinically useful formeasuring dietary protein and dietary energy intake in maintenancedialysis patients. (Evidence and Opinion)


R A T I O N A L EDuring steady-state conditions, nitrogen in-

take is equal to or slightly greater than nitrogenassessed as total nitrogen appearance (TNA).63TNA is equal to the sum of dialysate, urine, fecalnitrogen losses, and the postdialysis increment inbody urea-nitrogen content. Because the nitro-gen content of protein is relatively constant at16%, the protein equivalent of total nitrogenappearance (PNA) can be estimated by multiply-ing TNA by 6.25 (PNA is mathematically identi-cal to the protein catabolic rate or PCR). In theclinically stable patient, PNA can be used toestimate protein intake. Because protein require-ments are determined primarily by fat-free,edema-free body mass, PNA is usually normal-ized (nPNA) to some function of body weight(eg, actual, adjusted, or standardized [NHANESII] body weight [SBW] or body weight derivedfrom the urea distribution space [Vurea/0.58]).63Because urea nitrogen appearance (UNA; ie, thesum of urea nitrogen in urine and dialysate and

the change in body urea nitrogen) is highlycorrelated with TNA and measurement of totalnitrogen losses in urine, dialysate, and stool isinconvenient and laborious, regression equationsto estimate PNA from measurements of ureanitrogen in serum, urine, and dialysate have beendeveloped. The estimation of PNA from measure-ments of urea nitrogen is readily performed fromthe routine urea kinetic modeling session in HDpatients and, at least in theory, should be subjectto less measurement error than dietary diariesand recall. The equations used to estimate PNAare discussed in Appendix V.

There are several important limitations to PNAas an estimate of DPI. First, PNA approximatesprotein intake only when the patient is in nitro-gen equilibrium (steady-state).63 In the catabolicpatient, PNA will exceed protein intake to theextent that there is net degradation and metabo-lism of endogenous protein pools to form urea.Conversely, when the patient is anabolic (eg,growth in children, recovering from an intercur-

GUIDELINE

8

Protein Equivalent of Total Nitrogen Appearance (PNA)

PNA or PCR is a valid and clinically useful measure of net proteindegradation and protein intake in maintenance dialysis (MD) patients.(Evidence)

When nitrogen balance is zero in the steady state, the difference betweennitrogen intake and total nitrogen losses is zero or only slightly positive (ie,up to about 0.5 g nitrogen/d because of unmeasured nitrogen losses).Hence, in the clinically stable patient, PNA provides a valid estimate ofprotein intake. The protein equivalent of total nitrogen appearance (PNA) can beestimated from interdialytic changes in urea nitrogen concentration inserum and the urea nitrogen content of urine and dialysate. Because both net protein breakdown under fasting conditions and dietaryprotein requirements are strongly influenced by body mass, PNA (or PCR)is often normalized to a function of body weight (Guideline 12).


rent illness, or during the last trimester of preg-nancy) dietary protein is utilized for accrual ofnew body protein pools, and PNA will underesti-mate actual protein intake. Second, UNA (andhence PNA) changes rapidly following varia-tions in protein intake. Hence, PNA may fluctu-ate from day to day as a function of proteinintake, and a single PNA measurement may notreflect usual protein intakes. Third, when DPI ishigh, TNA underestimates protein intake (ie,nitrogen balance is unrealistically positive).64,65This is probably caused by increased nitrogenlosses through unmeasured pathways of excre-tion (eg, respiration and skin).66 Fourth, PNAmay overestimate DPI when the protein intake isless than 1 g/kg/d (possibly due to endogenousprotein catabolism).67-69 Finally, normalizing PNAto body weight can be misleading in obese,malnourished, and edematous patients. There-fore, it is recommended that for individuals whoare less than 90% or greater than 115% of SBW,the adjusted edema-free body weight (aBWef) beused when normalizing PNA to body weight(Guideline 12).

Notwithstanding these limitations, when con-sideration is given to the caveats discussed above,the nPNA is a valid and useful method forestimating protein intake. However, PNA shouldnot be used to evaluate nutritional status inisolation, but rather as one of several indepen-dent measures when evaluating nutritional sta-tus.


1. There are still a number of technical prob-lems with measuring PNA in individuals under-going HD or peritoneal dialysis that engendererrors and increase the costs of measurement.Research to decrease these sources of error wouldbe useful.

2. The mathematical relationship between PNAand protein intake in MHD patients has not beenwell defined. A larger database to examine theserelationships more precisely would be useful.

3. More research into optimal methods fornormalizing PNA to body mass would be valu-able.


R A T I O N A L ESubjective global assessment (SGA) is a repro-

ducible and useful instrument for assessing thenutritional status of MD patients.16,29,70-72 It is asimple technique that is based on subjective andobjective aspects of the medical history andphysical examination. SGA was initially devel-oped to determine the nutritional status of pa-tients undergoing gastrointestinal surgery73,74 andsubsequently was applied to other popula-tions.16,29,70-72,74-77

Among the benefits of using the SGA are thatit is inexpensive, can be performed rapidly, re-quires only brief training, and gives a globalscore or summation of protein-energy nutritionalstatus. Disadvantages to the SGA include the factthat visceral protein levels are not included in theassessment. SGA is focused on nutrient intakeand body composition. It is subjective, and itssensitivity, precision, and reproducibility overtime have not been extensively studied in MHDpatients.

Many cross-sectional studies have used theSGA to assess nutritional status in individuals

undergoing CPD.16,29,71,75,78 Correlations amongSGA and other measures of protein-energy nutri-tional status are well described.29,71 SGA hasbeen less well studied in MHD patients.72 In theCanada-USA (CANUSA) study, a prospectivecohort study of 680 continuous ambulatory peri-toneal dialysis (CAPD) patients, SGA was modi-fied to four items (weight loss, anorexia, subcuta-neous fat, and muscle mass). Subjectiveweightings were assigned to each of the fouritems representing nutritional status (eg, 1 to 2represented severe malnutrition; 3 to 5, moderateto mild malnutrition; and 6 to 7, normal nutri-tion).16

It is recommended that SGA be determined bythe 4-item, 7-point scale used in the CANUSAStudy,16 because this method may provide greatersensitivity when assessing nutritional status andmore predictive power in MD patients than theoriginal 3-point ordinal scale.73,74 The CANUSAstudy, using the 7-point scale, showed with mul-tivariable analysis that a higher SGA score wasassociated with a lower relative risk of death andfewer hospitalized days per year.16 Also, small

GUIDELINE

9

Subjective Global Nutritional Assessment (SGA)

SGA is a valid and clinically useful measure of protein-energy nutri-tional status in maintenance dialysis patients. (Evidence)


changes in the SGA score correlated with clinicaloutcomes.79 Methods for performing SGA arediscussed in Appendix VI.


1. The most effective technique for performingSGA needs to be identified. Is the currentlyrecommended 4-item scale optimal? Should vis-ceral proteins (eg, serum albumin, transferrin,

and/or prealbumin) be added to the SGA? Shoulda standard reference of body mass be included(eg, BMI or %SBW)?

2. The technique of SGA needs greater valida-tion with regard to sensitivity, specificity, accu-racy, intraobserver and interobserver variability,correlation with other nutritional measures, andpredictability of morbidity, mortality, or otherclinical outcomes.


R A T I O N A L EAnthropometry quantifies body mass, pro-

vides a semiquantitative estimate of the compo-nents of body mass, particularly the bone, muscle,and fat compartments, and gives informationconcerning nutritional status.31,80-83 The anthropo-metric parameters that are generally assessedinclude body weight, height, skeletal frame size,skinfold thickness (an indicator of body fat),mid-arm muscle circumference (MAMC; an indi-cator of muscle mass), area, or diameter, orpercent of the body mass that is fat, percent ofusual body weight (%UBW), percent of standard(NHANES II ) body weight (%SBW), and BMI.The various anthropometric measures providedifferent information concerning body composi-tion; therefore, there are advantages to measur-ing all of the parameters indicated above. Hence,the emphasis given to different anthropometricparameters and their relative precision should betaken into consideration. Anthropometry re-quires precise techniques of measurement andthe use of proper equipment to give accurate,reproducible data; otherwise, the measurementsmay give quite variable results.82 Some measures

of anthropometry are more precise, such as%UBW, %SBW, and BMI, than are skinfoldthickness and MAMC. Methods for performinganthropometry and calculating body composi-tion from these measurements and referencetables are presented in Appendix VII.

In adult MD patients, height is not a validmethod for measuring protein or energy nutri-tional status. However, it must be measuredbecause it is used in height-adjusted referencetables for weight (including SBW and BMI).Because height may decrease with aging, particu-larly in MD patients who have bone disease,height should be measured annually. Skeletalframe size must also be determined to calculatean individuals %SBW (see Appendix VII).

Muscle area, diameter, or circumference isused to estimate muscle mass and, by inference,the fat-free mass and somatic protein pool. Sig-nificant changes in these measurements reflectchanges in body muscle and somatic proteinmass and may indicate a nutritionally compro-mised state. Anthropometry has been used toassess nutritional status in MHD and CPD pa-tients.29,31,32,71,75,84 These studies indicate that

GUIDELINE

1 0

Anthropometry

Anthropometric measurements are valid and clinically useful indi-cators of protein-energy nutritional status in maintenance dialysispatients. (Evidence and Opinion)

These measures include percent usual body weight, percent standardbody weight, body mass index (BMI), skinfold thickness, estimated per-cent body fat, and mid-arm muscle area, circumference, or diameter.


muscle mass is decreased, often markedly, inmany, if not the majority, of MD patients.

Anthropometric monitoring of the same pa-tient longitudinally may provide valuable infor-mation concerning changes in nutritional statusfor that individual. The desirable or optimalanthropometric measures for MD patients havenot been defined. There is evidence that MHDpatients who have larger body-weight-for-height(eg, BMI) measurements are more likely to sur-vive, at least for the subsequent 12months.15,50,85,86 Patients in the lower 50th percen-tile of weight-for-height clearly have a reducedsurvival rate.15,85-87 One study indicates that MHDpatients who are in the upper 10th percentile ofbody weight-for-height have the greatest 12-month survival rate.85

In contrast to these findings, virtually all stud-ies of normal populations indicate that lowweight-for-height measures are associated withgreater survival, especially if the analyses areadjusted for the incidence of cigarette smokingin individuals with low BMI.88 Interpretation ofthese disparate findings among individuals under-going MD and the normal population is alsoconfounded by the lack of interventional trials inwhich a change in anthropometric measurementsis correlated with clinical outcome.

Anthropometric measurements in MD patientscan be compared with normal values obtainedfrom the NHANES II data89 or with values fromnormal individuals who have the greatest longev-

ity.88,90-97 Anthropometric norms for patientstreated with HD are published and generally aresimilar to the values available for the generalpopulation.98 Differences in anthropometric mea-surements among MD patients and normal indi-viduals may indicate a nutritional disorder orother clinical abnormality (eg, edema or amputa-tion). The use of currently available anthropomet-ric norms obtained from MD patients is of ques-tionable value since age-, sex-, and race- orethnicity-specific reference data are not availablefor this population. Furthermore, it has not beenshown that the norms for MHD patients aredesirable or healthy values.


1. Age-, sex-, and race- or ethnic-specificdesirable reference values for anthropometry ob-tained in large numbers of MD patients areneeded.

2. The risk of morbidity and mortality associ-ated with different anthropometric measure-ments in MD patients should be determined.

3. To determine whether anthropometry mightbe an acceptable intermediate outcome in nutri-tion intervention trials.

4. Will improvement in anthropometric valuesthrough nutritional intervention be associatedwith decreased morbidity and mortality and en-hanced quality of life in individuals undergoingMD?


R A T I O N A L EAssessment of body composition, particularly

with serial evaluation, can provide informationconcerning the long-term adequacy of protein-energy nutritional intake.58,99 Most clinically use-ful techniques for measuring body compositionare not very precise unless obtained by trainedanthropometrists using standardized methods,such as in Guideline 10. Whole body dual energyx-ray absorptiometry (DXA) is a reliable, nonin-vasive method to assess the three main compo-nents of body composition (fat mass, fat-freemass, and bone mineral mass and density). Theaccuracy of DXA is less influenced by the varia-tions in hydration that commonly occur in ESRDpatients.100-102 In vivo precision and accuracy offat mass estimates by DXA are approximately2% to 3% and 3%, respectively, in MHD101 andCPD patients. Studies of DXA in CRF, MHD,and CPD patients have demonstrated the supe-rior precision and accuracy of DXA as comparedwith anthropometry, total body potassium count-ing, cr

american journal of kidney disease

Documents

quality of dialysis

andimproving quality

quality of dialysiscare

care of esrd patients

agood quality of life

standard of care

patient outcomes

individual patients