Potential Application of the National Kidney Foundation's Chronic Kidney Disease Guidelines in a Managed Care Setting

,
The American Journal of Managed Care, July 2004 - Part 1, Volume 10, Issue 7 Pt 1

Chronic kidney disease (CKD) is a public health problem in theUnited States with rising incidence and prevalence. The diseasemay progress to end-stage renal disease, which is associated withsignificant morbidity and mortality. The National Kidney Foundationrecently established clinical practice guidelines to aid in theidentification and stratification of CKD. Central to these guidelinesis the use of glomerular filtration rate estimation equations.Integrated healthcare systems can use the guidelines to identifypatients with CKD early in its course. Early identification allows fortimely nephrology referral, institution of measures to slow the progressionof CKD, and treatment of CKD-specific complications.Administration of these efforts early in the course of CKD improvessubsequent morbidity and mortality of affected patients. By identifyingCKD early in its course, patients can be referred to a coordinatedcase management team that can help administer care. Theteam can follow up a patient from the early diagnosis of CKD toend-stage renal disease and renal replacement therapy. Finally,continuous quality improvement can be instituted to assist inimproving patient care.

(Am J Manag Care. 2004;10:417-422)

More than 19 million Americans have chronickidney disease (CKD).1 Adverse outcomes andconditions are associated with CKD, includingcoronary heart disease,2 congestive heart failure,3 andpremature death.4 Patients who survive the conditionsassociated with CKD develop end-stage renal disease(ESRD) and require dialysis or transplantation. The USRenal Data System5 estimated in a 2002 report thatthere are more than 450 000 persons with ESRD in theUnited States and that the rate of ESRD is increasing 5%to 6% per year.

Several risk factors predispose patients to CKD,including diabetes mellitus, hypertension, and age. Withrising rates of diabetes mellitus and hypertension in anaging population, it is projected that the prevalence ofCKD will continue to increase.5

Little is known about the progression of CKD toESRD. The estimated 19 million persons in the UnitedStates who have CKD is a significantly larger populationthan the 450 000 with ESRD. The large number ofpatients with CKD relative to the smaller number withESRD has led some to conclude that many patients withCKD die before ESRD or never progress.6 Common diseasesassociated with CKD (ie, diabetes mellitus andcardiac disease) are associated with high mortalityrates, which may explain some of the apparent lack ofCKD progression. The concept that diseases causingCKD lead to death before ESRD remains a theory, as thenatural history of CKD remains unknown.

Although little is known about the progression ofCKD in the aggregate, progression and treatment of specifickidney diseases causing CKD are well understood.Most notably, diabetic nephropathy, the underlying etiologyof nearly 40% of cases of ESRD,5 can be amelioratedwith the use of angiotension-converting enzymeinhibitors,7 aggressive glycemic control,8 and appropriatemanagement of hypertension.9 Early identificationof CKD among diabetic patients, combined with theseinterventions, has been shown to decrease the rate ofprogression of diabetic nephropathy (and thus ESRD).10Similar approaches to the management of nondiabetickidney diseases have been shown to slow the progressionof CKD to ESRD as well.11

In addition to slowing the course of CKD, treatmentof its many adverse effects (eg, anemia, hyperparathyroidism,etc) is important in ensuring quality patientcare.12 The appropriate addition of erythropoietin therapy,13 management of calcium and phosphorous,14 andtimely referral to a nephrologist15 are some of theimportant measures that improve outcomes amongpatients with CKD.

With little understood about the progression of CKD,many patients with CKD may not survive long enoughto develop ESRD. Therefore, interventions to improvethe health of patients with CKD may improve patientoutcomes before the development of ESRD and thusincrease the number of patients with ESRD. Conversely,earlier intervention among patients may slow or haltprogression and decrease the number of patients withESRD. In either case, intervention at an early stageseems appropriate.

The National Kidney Foundation (NKF) has establishedclinical practice guidelines that define and stratifyCKD based on the glomerular filtration rate (GFR).16The guidelines have been developed to help physiciansidentify CKD among individual patients and subsequentlymanage its many complications. The use of theNKF guidelines is, however, not limited to managementof individual patients. The guidelines offer a significantopportunity to manage populations of patients, especiallywithin large integrated healthcare systems. Theuse of patient demographic information, laboratoryfindings, patient education, and comprehensive supportiveservices can maximize the effect of these guidelineswithin a large population. The resulting effect ofpopulation-based CKD management may help curtailthe epidemic of CKD and ever-growing numbers ofpatients requiring renal replacement therapy. This articlewill discuss the potential roles the NKF guidelinescan play in the population-based management of CKDwithin integrated healthcare systems.

Identifying Patients With CKD

A central feature of the NKF guidelines is the use ofthe GFR to identify and stratify CKD. The GFR is consideredthe best means of assessing total kidneyfunction.17 Although other means of evaluatingkidney function are often used (eg, serum creatinineand creatinine clearance), they are lessreflective of functional nephron capacity than theGFR. Serum creatinine and creatinine clearancelevels are problematic because they overestimatethe GFR (because of secreted creatinine) at lowlevels.18 The GFR is difficult and expensive tomeasure directly. Although GFR measurementsusing iothalamate or inulin are accurate, the arealso cumbersome and expensive.

Many physicians do not regularly use GFRestimation equations unless they suspect apatient has decreased renal function. As a result,patients with relatively normal serum creatininesbut decreased GFRs may remain undiagnosed.

The NKF recommends that laboratoriesreporting serum creatinine levels, when possible,calculate a GFR estimation equation.19 Most suchequations rely on a combination of laboratoryand demographic information. For example, acommonly used GFR estimation equation wasderived for the Modification of Diet in RenalDisease Study.20 In its simplest form, it requires serumcreatinine level, age, sex, and whether or not the patientis African American. Greater precision can be achievedby adding variables that include serum albumin andblood urea nitrogen levels. A benefit of using GFR estimationequations is a more accurate assessment of kidneyfunction at low levels.20 With such a calculationalready made, physicians would be less likely to overestimaterenal function among patients with a seeminglynormal serum creatinine and a low GFR.

A problem with some GFR estimation equations isthe uncertainty of an estimated GFR (eGFR) when thetrue GFR is greater than about 60 mL/min per 1.73 m2.The previously mentioned Modification of Diet in RenalDisease equation is an example. As the study20 examinedpatients with decreased GFRs, the Modification ofDiet in Renal Disease estimation equation was not thoroughlyassessed among patients with normal renal function.As a result, a patient with normal kidney functionmay appear to have decreased function. To address thisproblem, the National Kidney Disease EducationProgram21 has recommended that values greater than60 mL/min per 1.73 m2 be reported simply as "above 60mL/min per1.73 m2," without being specific. Thisapproach would seem reasonable when considering theNKF definition of CKD. The diagnostic criteria defineCKD as a GFR greater than 60 mL/min per 1.73 m2 andthe presence of "radiological, serological, or urinary evidence,"while a lower GFR is CKD without other criteria.22 Therefore, patients with radiological, serological,or urinary evidence of CKD have CKD regardless of theGFR and (at least in theory) should be diagnosed basedon other findings.

Application of the NKF definition of CKD and use ofthe Modification of Diet in Renal Disease estimationequations allow for an automated system of CKD identification.When a patient has his or her serum creatininelevel checked, demographic information can beobtained and entered into the laboratory computerreporting system. When the serum value is obtained, aneGFR can be automatically calculated and reported.

To ensure that patients with acute renal insufficiencyor failure are not diagnosed with CKD, the NKF criteriafor the diagnosis require measurement of at least 2eGFRs 3 months apart.22 Therefore, laboratory reportingof eGFRs is only the first step in establishing a diagnosis.Healthcare systems with electronic medicalrecords that integrate laboratory information may beable to automate the identification of CKD by using multipleeGFRs, calculating the interval, and reporting tophysicians whether a specific patient meets the criteriafor the diagnosis of CKD. Healthcare systems that do notintegrate laboratory data with an electronic medicalrecord may be limited in their ability to take this secondstep and automatically identify patients with CKD onthe basis of laboratory values. Such systems mayinstead rely on physician recognition of CKD.

CKD Stratification and Population Management

Once a GFR has been established, the NKF guidelinesrecommend stratifying patients into 1 of 5 stages: stage1 (GFR, &#8805;90 mL/min per 1.73 m2), stage 2 (GFR, 60-89mL/min), stage 3 (GFR, 30-59 mL/min), stage 4 (GFR,15-29 mL/min), and stage 5 (GFR, <15 mL/min).22-24Additional criteria for patients in stages 1 and 2 includekidney damage for longer than 3 months manifest bypathological abnormalities or markers of kidney damagein blood or urine samples or on imaging tests. Markersof kidney damage include proteinuria, abnormalities ofthe urinary sediment, or abnormal radiological findings.

Identifying patients with stage 1 or 2 CKD requiresmore than a GFR estimation. The most common radiological,serological, or urinary evidence of CKD in mostpopulations is proteinuria. As proteinuria is often transient,multiple measurements may be required.25 Threeapproaches commonly used to identify proteinuriainclude measurement of the protein-creatinine ratio ona random urine sample, protein measurement from a24-hour urine collection, or protein identified by urinedipstick. The ratio from a spot urine sample is the mosteasily obtained and is less frequently associated withcollection error. A ratio greater than 200 mg/g is consideredan elevated level of proteinuria on a randomurine sample. Elevations of proteinuria due to transientconditions (exercise, urinary tract infections, etc) makeit necessary to confirm an initial test result with repeatedmeasures. There are few guidelines regarding howoften proteinuria needs to be detected before it is recognizedas a chronic problem (and thus providing forthe diagnosis of CKD). One guideline has been suggestedby the American Diabetes Association for assistancein identifying microalbuminuria among diabeticpatients with potential renal disease. The guideline suggeststhat 2 of 3 test results for microalbumin need to bepositive during 3 to 6 months to suggest a patient hasdiabetes mellitus with renal disease.26 The NKF recommendsthat patients with positive findings on 2 or morequantitative tests 1 to 2 weeks apart be diagnosed withchronic proteinuria.25

Healthcare systems with integrated electronic medicalrecord and laboratory computer systems may beable to establish protocols to ensure that patients withproteinuria (or albuminuria) on a single sample undergothe appropriate repeat testing. Computerized systemscan flag physicians regarding the need for possiblefollow-up testing once a positive result has beenobtained.

A final benefit of laboratory identification of CKD isthe potential creation of an automated surveillance systemreporting lists of patients with CKD at variousstages to a centralized monitor (perhaps a nephrologistor nephrology oversight group) that can review individualcases, alert primary care physicians, or directlyrefer patients to a nephrologist. A surveillance systemthat monitors laboratory results and cross-checks withpatient records to ensure that patients who have CKDare receiving appropriate treatment and referral canprevent patients from being overlooked who may benefitfrom specific intervention. Such a system may relyon monitoring GFRs of patients. Those in stages 3, 4, or5 who have not yet seen a nephrologist can be contactedthrough their primary care physician and advisedthat they should see a nephrologist. In addition,patients who may benefit from early interventions,including the use of angiotension-converting enzyme(ACE) inhibitors, lipid-lowering agents, or tight diabetesmellitus control can be identified early in the course oftheir disease.

The NKF recommends that the management ofpatients with CKD in stages 1 or 2 should focus ondetermination of a diagnosis and developing a treatmentplan that includes comorbid conditions and the underlyingcause of kidney disease.25 Comorbid conditionsoften associated with CKD include hypertension, diabetesmellitus, and hypercholesterolemia.27 Slowingthe progression of CKD includes controlling diabetesmellitus,8 treatment of hypertension,28 and use of anACE inhibitor11 or angiotension receptor blocker.29 AsCKD advances to GFRs less than 60 mL/min per 1.73m2, the NKF recommends additional focus on the specificcomplications of CKD (ie, development of bone disease,anemia, and protein malnutrition).16

Anemia develops among patients with CKD becauseof a relative deficiency of erythropoietin. It is recommendedthat patients with stage 3 CKD have hemoglobinlevels checked and, if less than 12.5 g/dL, undergoan evaluation of their anemia.24 If the workup is unremarkableand the patient is not iron deficient, treatmentwith erythropoietin is undertaken.

Bone disease occurs as a result of vitamin D deficiencyand hyperparathyroidism. Bone disease in CKDmay manifest as high-turnover bone disease with highparathyroid hormone levels or as low-turnover bone diseasewith low or normal parathyroid hormone levels.14It is recommended that patients with GFRs less than60 mL/min per 1.73 m2 be evaluated for bone diseasewith longitudinal measurements of parathyroid hormone,phosphorus, and ionized calcium levels.30 Treatmentof bone disease involves vitamin D replacement oruse of phosphate binders.

Protein malnutrition is common among dialysispatients and those with CKD. Patients with CKD withprotein malnutrition have greater morbidity and mortalitythan those who do not.31 To prevent protein malnutrition,early and ongoing dietary assessments,usually conducted by a dietician, are necessary.

Complicating treatment of protein malnutrition arecontinued questions regarding potential benefits of alow protein diet during early CKD. The results of theModification of Diet in Renal Disease Study32 (whichstudied whether low protein diets slow the rate of progressionof CKD) were inconclusive, further fueling thecontroversy. As a result, the NKF has recommendedthat patients with a GFR less than 25 mL/min per 1.73m2 consume 0.60 g/kg of body weight per day of protein,but does not make recommendations for patientswith higher GFRs. As the recommended dietaryallowance of protein for adults with a normal GFR is0.75 g/kg per day, it seems reasonable that patientswith GFRs of 25 mL/min or higher consume thisamount.33

Patients with stage 4 CKD (GFR, 15-29 mL/min per1.73 m2) should begin the preparing for renal replacementtherapy. At this point, patients should see or haveseen a nephrologist, been educated regarding renal replacementoptions, and (if hemodialysis is likely to be administered)have a dialysis access graft placed. Patients withstage 5 CKD (GFR, <15 mL/min per 1.73 m2) shouldreceive renal replacement therapy as determined bytheir nephrologist.

Nephrology Referral

Timely referral to a nephrologist is correlated withimproved morbidity and mortality after the initiation ofdialysis.34 Ensuring appropriate referral to a nephrologistcan be accomplished within health systems thathave integrated laboratory, primary care, and specialtyservices (Figure). Application of the NKF guidelines to asystem that is programmed to automatically stage CKDcan determine which patients are more likely to need anephrology referral. An integrated healthcare system,particularly one with an electronic medical record (onethat contains laboratory, radiological, pharmacy, andmedical history information) can help to further determinewhich patients should be closely monitored andwhich can be followed up less frequently.

Once patients who need to be seen by a nephrologisthave been referred, case management can be an importantmeans of further managing a population of patientswith CKD. Case management allows for coordination ofCKD- and ESRD-specific issues.

Coordinated Care Management

Patients with CKD have specific problems thatrequire therapy from nurses and physicians with specialknowledge of kidney disease. The development of a multidisciplinarycare management group within largehealthcare systems is helpful in administering qualitycare among patients with CKD and ESRD. Elements of acare management team often include a nurse care manager,social worker, vascular access coordinator, anddietician. Care should begin before ESRD and be maintainedfrom the time of diagnosis of CKD through renalreplacement therapy (dialysis or transplantation). Atleast 1 study35 has shown that patients seen by such ateam before initiation of dialysis had fewer cerebrovascularevents, decreased hospital days, and reducedoverall mortality compared with those not followed upbefore dialysis.

Patient education is an important part of care managementand should be initiated early in the course ofCKD. Patients involved in predialysis education had aslower rate of decline in renal function than those notexposed to an educational system.36 Patient educationmay take place in the form of a group class, individualmeetings with members of the care managementteam, mailings, videos, and so on. The focus of educationshould include an understanding of kidney function,modification of diet and lifestyle to slow diseaseprogression, coping skills, and participation in healthcarediscussions. Educational activities should bestarted early and continued throughout the course ofthe patient's illness. As CKD progresses, specific informationshould be presented about the different dialysisoptions available, dialysis access placement,transplantation, and decisions to forgo renal replacementtherapy. Referral for education is an excellentpoint at which to begin case management of patientswith CKD.

Within the multidisciplinary team, individual teammembers should address specific issues. A social workerassists in finding placement for dialysis treatment,helps to arrange transportation, and provides supportduring the transition to dialysis. Efforts to facilitate dialysisin a manner that allows patients to continue towork are a part of dialysis planning. The renal nutritionistmay begin monitoring laboratory results to managebone disease and promote dietary compliance anddiabetes mellitus management in the face of changingrenal function. Nurse managers can assist with vascularaccess placement, transplant workup, and erythropoietinadministration.

The period immediately preceding and immediatelyfollowing the initiation of dialysis is often stressful forpatients and requires close monitoring. The benefits ofpatient management early in the course of CKD oftenmanifest at the time of dialysis initiation. A well-preparedpatient, with the support of a multidisciplinaryteam, can undergo an easier transition to dialysis.

Once a patient starts dialysis, many of the functionsof the multidisciplinary team are continued within thedialysis unit. Integrated healthcare systems that havetheir own dialysis units may continue to use the sameteam to manage patient populations before and afterdevelopment of ESRD. Those systems that contractwith outside dialysis units to provide dialysis forpatients with ESRD need to consider the role of ongoingcare management once the patient is receiving similarservices through the dialysis unit. The continuation ofsome services may be appropriate (such as dialysisaccess and transplant coordination), while others maybe duplicative.

Continuous Quality Improvement

The NKF guidelines can be applied to continuousquality improvement within an integrated healthcaresystem. Specific outcome measurements for each stageof CKD and ESRD can be measured and followed up.Identifying problems within individual dialysis units oramong elements of care management should allow systemsto improve population health and ultimatelydecrease the morbidity of patients with CKD.

Efforts to slow the progression of CKD amongpatients with specific kidney diseases, such as diabeticnephropathy, might also benefit from the NKF guidelines.By following up the progression of CKD amongspecific populations, healthcare systems can ensurethat important interventions (eg, the use of ACEinhibitors among patients with diabetic nephropathy)are administered to a maximum number of patientswithin the population.

Other continuous quality improvement targetsinclude preparation for dialysis or transplantation,beginning when the eGFR falls below 30 mL/min per1.73 m2. This includes dialysis access placement, which(for hemodialysis patients) ideally involves placement ofan arteriovenous fistula, an access requiring up to 4months to mature. Arteriovenous fistulas are the leastlikely form of dialysis access to be associated with complicationsand are thus preferred. The NKF KidneyDialysis Outcomes Quality Initiative suggests a targetgoal of 50% of new kidney failure patients having an arteriovenousfistula placed before starting hemodialysis.37

Conclusions

The NKF guidelines for the identification, evaluation,and management of CKD provide integratedhealthcare systems an opportunity to more efficientlyand precisely manage patients with the disease.Patients with CKD can be identified using automatedcalculations of GFRs based on laboratory and demographicinformation. Once identified, patients may bestratified using the NKF criteria and appropriate interventionundertaken. Integrated healthcare systems canuse the NKF criteria to automate the identification andreferral of patients with CKD. As CKD progresses, coordinatedcare management should be instituted andcontinued as the disease progresses to ESRD. Finally,continuous quality improvement can be facilitated bydata collection.

From the Lake Road Nephrology Center, Kaiser Kidney Program, Kaiser PermanenteNorthwest, Milwaukie, Ore.

Address correspondence to: Micah L. Thorp, DO, MPH, Lake Road NephrologyCenter, Kaiser Kidney Program, Kaiser Permanente Northwest, 6902 SE Lake Road,Milwaukie, OR 97267. E-mail: micahthorp@comcast.net.

Am J Kidney Dis.

Am J Kidney Dis

1. Jones CA, McQuillan GM, Kusek JW, et al. Serum creatinine levels in the USpopulation: Third National Health and Nutrition Examination Survey [publishedcorrection appears in 2000;35:178]. .1998;32:992-999.

Kidney Int

2. Culleton BF, Larson MG, Wilson PW, Evans JC, Parfrey PS, Levy D.Cardiovascular disease and mortality in a community-based cohort with mild renalinsufficiency. . 1999;56:2214-2219.

Ethn Dis

3. McCullough PA. Scope of cardiovascular complications in patients with kidneydisease. . 2002;12:S3-44-8.

Am J Kidney Dis

4. Drey N, Roderick P, Mullee M, Rogerson M. A population-based study of theincidence and outcomes of diagnosed chronic kidney disease. .2003;42:677-684.

USRDS 2002 Annual Data Report: Atlas of End-StageRenal Disease in the United States

5. US Renal Data System. . Bethesda, Md: National Institute of Diabetesand Digestive and Kidney Diseases, National Institutes of Health; 2002.

J Am Soc Nephrol

6. Keith D, Nichols G, Gullion C, Brown J, Smith D. Mortality of chronic kidneydisease (CKD) in a large HMO population [abstract]. .2002;13:620A. Abstract SU-P0740.

Lancet.

7. EUCLID Study Group. Randomised placebo-controlled trial of lisinopril in normotensivepatients with insulin-dependent diabetes and normoalbuminuria ormicroalbuminuria. 1997;349:1787-1792.

N Engl J Med.

8. Diabetes Control and Complications Trial Research Group. The effect of intensivetreatment of diabetes on the development of long-term complications ofinsulin-dependent diabetes mellitus. 1993;329:977-986.

Lancet

9. Parving HH, Anderson AR, Smidt UM. Early aggressive antihypertensive treatmentreduces rate of decline of kidney function in diabetic nephropathy. .1983;1:1175-1179.

AmJ Kidney Dis

10. Bakris GL, Williams M, Dworkin L, et al, National Kidney FoundationHypertension and Diabetes Executive Committees Working Group. Preservingrenal function in adults with hypertension and diabetes: a consensus approach. . 2000;36:646-661.

Curr Opin Nephrol Hypertens

11. Ruggenenti P, Remuzzi G. Angiotension-converting enzyme inhibitor therapy fornondiabetic progressive renal disease. . 1997;6:489-495.

SeminDial

12. Barrett BJ. Applying multiple interventions in chronic kidney disease. . 2003;16:157-164.

Kidney Int

13. McGonigle RJ, Wallin JD, Shadduck RK, Fisher JW. Erythropoietin deficiencyand inhibition of erythropoiesis in renal insufficiency. . 1984;25:437-444.

Kidney Int

14. Sherrard DJ, Hercz G, Pei Y, et al. The spectrum of bone disease in end-stagerenal failure: an evolving disorder. . 1993;43:436-442.

Ann Intern Med

15. Kinchen KS, Sadler J, Fink N, et al. The timing of specialist evaluation inchronic kidney disease and mortality. . 2002;137:479-486.

Am J Kidney Dis

16. National Kidney Foundation. K/DOQI Clinical Practice Guidelines for ChronicKidney Disease: evaluation, classification, and stratification: part 1: executive summary.. 2002;39(suppl 1):S17-S31.

Ann Intern Med

Ann Intern Med

17. Levey AS, Coresh J, Balk E, et al. National Kidney Foundation practice guidelinesfor chronic kidney disease: evaluation, classification, and stratification [publishedcorrection appears in . 2003;139:605]. .2003;139:137-147.

Kidney Int

18. Shemesh O, Golbetz H, Kriss JP, Myers BD. Limitations of creatinine as a filtrationmarker in glomerulopathic patients. . 1985;28:830-838.

Am J Kidney Dis.

19. National Kidney Foundation. K/DOQI Clinical Practice Guidelines for ChronicKidney Disease: evaluation, classification, and stratification: part 5: evaluation oflaboratory measurements for clinical assessment of kidney disease. Estimation ofGFR. 2002;39(suppl 1):S76-S92.

Ann Intern Med

20. Levey AS, Bosch JP, Lewis JB, Greene T, Rogers N, Roth D, Modification of Dietin Renal Disease Study Group. A more accurate method to estimate glomerular filtrationrate from serum creatinine: a new prediction equation. .1999;130:461-470.

21. National Kidney Disease Education Program. Meeting on creatinine assay andreporting of estimated GFR 2003. Available at: http://www.nkdep.nih.gov.Accessed May 10, 2004.

AmJ Kidney Dis

22. National Kidney Foundation. K/DOQI Clinical Practice Guidelines for ChronicKidney Disease: Part 4-Definition and Classification of Chronic Kidney Disease. . 2002;39(suppl 1):S46-S64.

23. National Kidney Foundation Web site. Available at: http://www.kidney.org.Accessed May 10, 2004.

Am J KidneyDis

24. National Kidney Foundation. K/DOQI Clinical Practice Guidelines for ChronicKidney Disease: Guideline 8-Association of level of GFR with Anemia. . 2002;39(suppl 1):S120-S127.

Am J Kidney Dis

25. National Kidney Foundation. K/DOQI Clinical Practice Guidelines for ChronicKidney Disease: Guideline 5-Assessment of Proteinuria. .2002;39(suppl 1):S93-S102.

Diabetes Care.

26. Molitch ME, DeFronzo RA, Franz MJ, Keane WF, Mogensen CE, Parving HH.Diabetic nephropathy. 2003;26(suppl 1):S94-S98.

Diabetologia

27. Lam KSL, Cheng IKP, Janus ED, Pang RWC. Cholesterol-lowering therapy mayretard the progression of diabetic nephropathy. . 1995;38:604-609.

Am J Kidney Dis.

28. Walker WG. Hypertension-related renal injury: a major contributor to endstagerenal disease. 1993;22:164-173.

N Engl J Med.

29. Lewis EJ, Hunsicker LG, Clarke WR, et al. Renoprotective effect of theangiotensin-receptor antagonist irbesartan in patients with nephropathy due to type2 diabetes. 2001;345:851-860.

Am J Kidney Dis.

30. National Kidney Foundation. K/DOQI Clinical Practice Guidelines for ChronicKidney Disease: Guideline 10-Association of level of GFR with Bone Disease andDisorders of Calcium and Phosphate Metabolism. 2002;39(suppl1):S143-S169.

Nephrol Dial Transplant

31. Holland D, Lam M. Predictors of hospitalization and death amongst predialysispatients: a retrospective study. . 2000;15:650-658.

J Am Soc Nephrol

32. Levey AS, Greene T, Beck GJ, et al, Modification of Diet in Renal DiseaseStudy Group. Dietary protein restriction and the progression of chronic renal disease:what have all the results of the MDRD Study shown? .1999;10:2426-2439.

Am JKidney Dis

33. National Kidney Foundation. K/DOQI Clinical Practice Guidelines for ChronicKidney Disease: Guideline 9-Association of GFR with Nutritional Status. . 2002;39(suppl 1):S128-S142.

Am J Kidney Dis

34. Stack AG. Impact of timing of nephrology referral and pre-ESRD care on mortalityrisk among new ESRD patients in the United States. .2003;41:310-318.

Nephrologie

35. Jungers P, Nguyen KT, Massey ZA, Man NK. Incidence of atherosclerotic cardiovascularevents in patients with chronic anemia: epidemiologic studies in Ile deFrance. . 1998;19:489-494.

J NervMent Dis

36. Binik YM, Devins GM, Barre PE, et al. Live and learn: patient education delaysthe need to initiate renal replacement therapy in end-stage renal disease. . 1993;181:371-376.

Am J Kidney Dis.

37. National Kidney Foundation. III. NKF–K/DOQI Clinical Practice Guidelines forVascular Access: update 2000. 2001;37(suppl 1):S137-S181.