Impact of years of dialysis therapy on mortality risk and the characteristics of longer term dialysis survivors

Am J Kidney Dis. 2002 Mar;39(3):533-8. doi: 10.1053/ajkd.2002.31403.


With improving survival and a decreasing probability of receiving a transplant, patients with end-stage renal disease (ESRD) are more likely to remain on hemodialysis therapy for more years than in the past. This study evaluates the effect of years on dialysis (vintage) on relative risk (RR) for death with and without adjustment for comorbidities and treatment factors. It also compares characteristics of patients on hemodialysis therapy for 7 years or longer with those on hemodialysis therapy for 1 to 7 years. Data were combined from two special US Renal Data System studies, the Case Mix Adequacy Study and Waves 1, 3, and 4 of the Dialysis Mortality and Morbidity Study. Excluding the first year of dialysis, 12,687 patients were studied during a 2-year follow-up, censoring at transplantation or loss to follow-up. Unadjusted analysis (vintage 1 to < 2 years as referent) showed that the risk for death remained nearly the same until the end of year 7 of dialysis therapy, after which the risk decreased significantly. However, with adjustment for demographics, comorbidities, and treatment factors, vintage was significantly associated with increased mortality risk during years 2 to less than 8 (RR = 1.12 to 1.30; P < 0.05). Vintage was independently associated with increased adjusted mortality among patients with and without diabetes until approximately 6 to less than 8 years of dialysis therapy. Patients on dialysis therapy for 7 years or longer were significantly (P < 0.05) more likely to be women, younger, and have lower phosphorus levels, higher hematocrits, and higher delivered dialysis doses. We conclude that adjusted mortality risk does not decrease with years on dialysis therapy, and modifiable factors deserve greater attention to improve survival among patients with ESRD with and without diabetes treated by hemodialysis.

Publication types

  • Research Support, Non-U.S. Gov't

MeSH terms

  • Diabetes Complications
  • Female
  • Humans
  • Kidney Failure, Chronic / complications
  • Kidney Failure, Chronic / mortality*
  • Kidney Failure, Chronic / therapy*
  • Logistic Models
  • Male
  • Middle Aged
  • Proportional Hazards Models
  • Renal Dialysis*
  • Time Factors
  • United States / epidemiology