Five-Year Survival for End-Stage Renal Disease Patients in the United States, Europe, and Japan, 1982 to 1987

We compared the 5-year survival for new end-stage renal disease (ESRD) patients accepted for renal replacement therapy (RRT) between 1982 and 1987 in the United States (n = 150,862), Europe (European Dialysis and Transplant Association [EDTA]) (n = 124,796), and Japan (n = 66,244). Given these large samples that approach a census in each of the three regions, all results are statistically significant. Our analysis showed that the US patients were older and more likely to be diabetic than the ptients in either EDTA or Japan. After correction for patient differences in age composition and the percent diabetic, Japan had the highest survival, followed by EDTA, and then the US. Overall, the US 5-year survival was 40%. When comparison is done by age, only the youngest patients in the US (< 15 years) have longer survival than their counterparts in Europe and Japan. For ages greater than 14 years, the survival differences between the US and EDTA and between the US and Japan grow larger with higher patient age. The comparisons of mortality by diagnosis showed that the differences between the US and EDTA and between the US and Japan were least for diabetes. For non-diabetic patients, the age adjusted relative risk (RR) of mortality for the US compared with EDTA was 1.22, ie, 22% higher in the US; for the US compared with Japan, the RR was 1.40. In contrast, the RR for diabetic patients in the US compared with EDTA was 1.07, and 1.23 for the US compared with Japan. Adjusting for age and diabetes, the RR for the US compared with EDTA was 1.15, and 1.33 for the US compared with Japan. Since the data on which the analysis is based come from three very different societies with possibly different acceptance criteria for RRT and three different data systems, caution in interpretation is advised. The observed large mortality differences for non-diabetic patients in the middle-age group suggest that acceptance criteria do not explain all the differences in observed mortality. Additional comparisons to the general populations in the three areas suggest that underlying mortality differences do not explain these differences in ESRD survival.