SURVIVAL ANALYSIS OF DIALYSIS PATIENTS IN OKINAWA, JAPAN (1971-1990)

We analyzed longitudinal data obtained from the initiation of chronic dialysis in Okinawa, Japan. A total of 1,982 patients (824 females and 1, 158 males) were registered in the Okinawa Dialysis Study (OKIDS) up to the end of 1990. The number of patients dying, undergoing renal transplantation, or being transferred was 605 (30.5%), 75 (3.8%), and 23 (1.2%), respectively. The mean acceptance rate per million population increased from 19.7 in 1971 to 1975 to 157.4 in 1986 to 1990. The percentage of diabetic patients and the annual gross mortality rate were, respectively 0% and 0.52 (1971 to 1975), 7.3% and 0. 12 (1976 to 1980), 14.4% and 0.06 (1981 to 1985), and 24.6% and 0.07 (1986 to 1990). Cox proportional hazard analysis was used to determine the relative risk (RR) for sex, primary renal disease, age at entry, and the year of starting dialysis. The RR for males was 1.09 (1.00 for females) and the 95% confidence interval (CI) was 0.93 to 1.28. The RR for diabetics was 1.88 (95% CI; 1.55 to 2.28) when that for nondiabetics was set at 1.00. The RR (95% CI) for starting dialysis in 1976 to 1980, 1981 to 1985, and 1986 to 1990 was 0.65 (0.59 to 0.72), 0.43 (0.35 to 0.52), and 0.28 (0.20 to 0.38), respectively, when the RR in 1971 to 1975 was taken as 1.00. During the last two decades, the survival of chronic dialysis patients in Okinawa has continued to improve despite the large increase in acceptance rate, the older age of the new patients, and the increase in diabetic patients. Better dialysis techniques and patient care may have contributed to these results.