Survival differences among older dialysis patients in the southeast

Older end-stage renal disease (ESRD) patients treated by chronic dialysis have higher mortality in the United States than in many other countries. While increasing age, white race, male sex, and/or diabetes are considered risk factors for survival, few studies of older dialysis patients have simultaneously considered multiple predictor variables and their interactions. Using information contained in the 1982 to 1986 ESRD Network 20 database for Georgia and South Carolina, we studied hospitalizations and survival of 1,354 blacks and 965 whites who were age 60 years or older when they began dialysis therapy. Survival time was modeled using the Cox life-table regression method. Older blacks' median age at dialysis initiation was 67.4, compared with 68.7 for older whites (P = 0.001). Blacks were more likely than whites (P < 0.001) to have hypertension-related or diabetes-related ESRD. White patients experienced approximately 25% more hospitalization when adjustment was made for patient-days at risk. Separate multivariate survival models were required for patients with diabetes-related versus non-diabetes-related ESRD. Among diabetics, mortality was higher among whites and among patients who were older when they began dialysis. Among patients with non-diabetes-related ESRD, mortality was higher among patients who were older when they began dialysis, but the age effect was much stronger for white males. Our hospitalization and mortality data support the view that unmeasured severity (or frailty) differences characterize white as compared with black dialysis patients. Among non-diabetes-related ESRD patients, the age effect on survival was more severe in white males than in blacks or in white females. The high mortality we observed among older dialysis patients in Georgia and South Carolina warrants further study; the data may in part reflect patients' lower socioeconomic status compared with age, race, and sex-matched controls.