Serum albumin measurement in dialysis patients: Should it be a measure of clinical performance?

Despite many years of work by renal nutritionists, protein-energy malnutrition remains very common among patients with advanced chronic kidney disease and those treated with dialysis.1 Adults with chronic kidney disease and protein-energy malnutrition have high mortality, and hypoalbuminemia at the initiation of chronic dialysis is highly predictive of future mortality.2, 3, 4 The Kidney/Disease Outcomes Quality Initiative (K/DOQI) Clinical Practice Guidelines for Nutrition of Chronic Renal Failure make recommendations for dialysis patient dietary protein intake.1 These guidelines outline the evidence indicating that the serum albumin level is a valid and useful measure of nutritional status in dialysis patients. It would seem logical, therefore, that the serum albumin level should be used to assess the clinical performance of the nutritionist and the care team in improving patient nutrition and outcome.

Why do we need clinical performance measures (CPMs)? In our new information era, the efforts of care teams to improve patient outcomes can be measured and compared. Consumers and payers of health care services and government regulators can use CPMs to find the best providers and to identify providers with poor results. In addition, care teams can use these measures to find opportunities to improve outcomes using continuous quality improvement techniques. The measured hematocrit (Hct) and the delivered dose of dialysis calculated as the urea reduction ratio (URR) have been used as CPMs, and since 1993 there has been evidence that these measures have improved continuously.5 Should the serum albumin level join these CPMs?

Unlike the trend of improvement for Hct and URR, the serum albumin level has not changed in a national sampling of dialysis patients.5 With all the work by dedicated renal nutritionists across the country, why is it that the serum albumin values remain marginal in dialysis patients and have not improved in many years? This may be particularly true for peritoneal dialysis (PD) patients. A recent analysis of PD patients showed that serum albumin values do not correlate with the dose of dialysis and are not strongly correlated with alternative estimates of nutritional status.6 In one study, the serum albumin level did not correlate with any index of body composition, and the investigators conclude that the serum albumin level is not a useful marker of malnutrition in stable patients on PD.7 Hypoalbuminemia may be partly dependent on subclinical overhydration.8 In another study, there was no relationship between the delivered dose of dialysis (expressed as Kt/V urea) or normalized protein catabolic rate (nPCR) and serum albumin concentration. The investigators conclude that hypoalbuminemia in PD patients may be a consequence of transperitoneal albumin losses and the acute phase response.9 Thus, the serum albumin level is determined by many factors, and the relative role of nutrition in determining the serum albumin level may be different in different patients.

There is a complex relationship between residual renal function and serum albumin level. Residual renal function, even at low levels, is linked to decreased mortality and better nutritional status.10 Yet, patients starting hemodialysis show an increase in serum albumin level in the first 6 months of treatment, as endogenous renal function declines.11 As residual renal function declines, progressive retention of protein contributes to increased serum albumin levels. Thus, to some extent, improving albumin levels represents decreased excretion rather than improved nutrition.

Several lines of evidence suggest that inflammation may reduce the serum albumin concentration in hemodialysis patients. Changes in the concentration of long-lived acute phase proteins predict the future concentration of serum albumin.12 Changing from conventional to ultrapure dialysis fluid reduces markers of inflammation such as interleukin 6 and C-reactive protein and increases the serum albumin concentration.13 Patients who die during the first year of hemodialysis care have lower mean serum albumin levels and lower rates of increase in the serum albumin level than do those who survive.11 Recent study results suggest that albumin concentration in hemodialysis patients is dependent on both nutrition and inflammation: nutrition affects albumin synthesis, and inflammation increases albumin catabolism.14

There are 2 questions of practical importance to nutritionists, other dialysis care providers, payers, and regulators: 1. Do dietary counseling and dietary supplements change serum albumin levels? 2. Do dietary counseling and dietary supplements improve survival or other patient outcomes? Several small studies do suggest that serum albumin may be increased in some patients by dietary intervention.15, 16 Continuous quality improvement activity may improve serum albumin concentration.17 High-dose dialysis18 and prolonged, frequent, slow dialysis19 have been associated with increased serum albumin level. Nonetheless, a nationwide review of serum albumin levels over many years shows no clear evidence of improvement, despite the efforts of nutritionists in virtually every dialysis facility.5 How do we reconcile these findings? It may well be that the determinants of the serum albumin level are so complex that it is difficult to show a clear effect of any single manipulation, such as improved diet, in a heterogenous population of dialysis patients with varying degrees of inflammation, multisystem disease, and many dialysis treatment modalities.

Steinman20 reviewed the evidence and notes that it is still unknown whether the serum albumin level can be effectively increased in the chronic dialysis patient. Furthermore, there is no credible evidence that increasing the serum albumin level can alter long-term morbidity and mortality. Thus, it seems that nutritionists may only rarely be able to alter serum albumin levels and to improve patient outcomes.

Should a hard-working nutritionist be judged by whether the serum albumin level increases? The evidence says no: there is no clear relationship between efforts to improve nutrition and serum albumin level or survival in dialysis patients. As the relationship between inflammation, serum albumin level, nutrition, and survival is examined, we look forward to discovering effective ways to improve outcomes. Until then, the serum albumin level should not be used as a clinical performance measure.