Re-evaluating the volume-outcome relationship in hemodialysis patients.

Objectives: We sought to determine whether dialysis patient mortality rates are associated with differences in dialysis facility size, and whether this relationship differs among higher risk diabetic and lower-risk non-diabetic patients.
Methods: Using 186,554 adult end-stage renal disease patients initiating hemodialysis at standalone facilities in the United States between 1996 and 1999, we evaluated relationships between dialysis facility size and survival to 5 years. We performed separate analyses for patients with and without diabetes as their primary cause of end-stage renal disease. Facility size was defined according to the number of hemodialysis patients at year's end (small<or=60, medium 61-120, and large>or=120).
Results: Increasing facility size was associated with a reduced risk of mortality at 4 years for both diabetic (HR=0.983 per 10 unit increase, 95% CI=0.967, 0.999) and non-diabetic patients (HR 0.977 per 10 unit increase, 95% CI=0.963, 0.992) dialyzing in small facilities, and for diabetic patients (HR 0.989 per 10 unit increase, 95% CI=0.980, 0.998) dialyzing in medium size facilities.
Conclusions: Smaller facility size is associated with increasing long-term mortality for in-center hemodialysis patients. This relationship appears to be more pronounced among higher-risk diabetic vs. lower-risk non-diabetic patients.