Received June 7, 2006
Accepted on September 28, 2006

ORIGINAL ARTICLES

Exploring Secular Trends in the Likelihood of Receiving Treatment for End-Stage Renal Disease

Chi-yuan Hsu ^*1, Alan S. Go ^*, Charles E. McCulloch , Jeanne Darbinian , and Carlos Iribarren

Departments of *Medicine and Epidemiology and Biostatistics, University of California, San Francisco, San Francisco, and Division of Research, Kaiser Permanente of Northern California, Oakland, California

¹ To whom correspondence should be addressed. E-mail: hsuchi{at}medicine.ucsf.edu.

	Abstract

There is a limited understanding of the forces that drive the steady rise in the number of patients who receive treatment for ESRD. It was hypothesized that this is not simply due to increasing prevalence of chronic kidney disease (CKD) or changes in renal failure risk factors in the population from which ESRD cases develop. A noncurrent cohort study was conducted to quantify the change over time (per year) in the likelihood of receiving ESRD therapy in a cohort of 320,252 individuals who volunteered for health check-ups between 1964 and 1985. Initiation of ESRD treatment was ascertained using the US Renal Data System registry through 2000. A total of 1471 cases of ESRD were observed during 8,347,955 person-years of observation, with ESRD cases developing between 1973 and 2000. In unadjusted Cox proportional hazards analysis, individuals who were examined later in time had an 8% per year higher risk for progressing to receive treatment for ESRD (relative risk 1.08; 95% confidence interval 1.05 to 1.11). This temporal trend in risk for future ESRD associated with year of cohort entry (baseline examination) was not explained by increases over time in the prevalence of CKD or risk factors for renal failure. After adjustment for age, gender, race, diabetes, BP, body mass index, education level, smoking status, history of myocardial infarction, serum cholesterol, proteinuria, hematuria, and serum creatinine level, there remained an 8% per year increase in risk (relative risk 1.08; 95% confidence interval 1.06 to 1.11). Among individuals who were examined from the 1960s through the 1980s, those who were examined later were more likely to receive treatment for ESRD. This trend was not accounted for by increasing prevalence of baseline CKD or risk factors for renal failure. These findings should spur further research into other forces that drive the rise in treated ESRD.