Predictors of survival on renal replacement therapy

Background and aims. Patients with end-stage renal disease and on chronic renal replacement therapy are at increased risk of death compared to a population of the same age without end-stage renal disease. Despite some improvement in prognosis of end-stage renal disease patients during recent decades, the expected outcome remains dismal. Several factors are associated with impaired survival of patients with end-stage renal disease: especially high age, low serum albumin, chronic inflammation, and comorbidities such as diabetes and heart failure. However, a major portion of factors behind impaired survival have been insufficiently identified, and their prognostic weight is largely unknown. We therefore targeted for further exploration and measurement of factors potentially associated with survival of patients on chronic renal replacement therapy. Study patients and methods. In the four studies of this thesis, we investigated the survival of cohorts of adult patients in Finland after the start of chronic renal replacement therapy. These cohorts were in Study I, 1,604 patients with type 1 diabetes and 1,556 with glomerulonephritis who started chronic RRT during 1980–2005; in Study II, 319 patients during one year (1998) preceding start of chronic renal replacement therapy and thereafter; in Study III, all 4,463 patients who started chronic renal replacement therapy in 2000–2009; in Study IV, all 6,103 patients who entered chronic dialysis in 2000–2012. Data on patient cohorts came from the Finnish Registry for Kidney Diseases, a database including comprehensive information on Finnish patients on chronic renal replacement therapy since 1964. In Study III, data also came from the Finnish Kidney Transplant Registry. The statistical methods mainly employed in our survival analyses were Kaplan-Meier curves, the log rank test, and Cox proportional hazards regression and binary logistic regression for multivariable modeling. Results. In Study I, we showed that survival of patients with type 1 diabetes receiving chronic renal replacement therapy has improved significantly and consistently over the years and in all age-groups. Patients entering chronic renal replacement therapy in 2000–2005 had a 77% lower risk of death than those entering in 1980–1984. Said another way, median survival time of patients with type 1 diabetes on chronic renal replacement therapy has increased from 3.6 years to more than eight. In Study II, we detected a significantly higher age-adjusted mortality in those on chronic renal replacement therapy whose decline in estimated glomerular filtration rate during the predialysis phase had been the fastest. Their mortality risk was 73% higher in the patient tertile of fastest decliners compared to the slowest. This association, however, proved not to be causal, but instead jointly caused by many confounding factors, especially age, end-stage renal disease diagnoses type 1 diabetes and amyloidosis, and the comorbidities myocardial infarction and cancer. In Study III, results of our primary adjusted analyses (intention-to-treat) indicated no significant difference in risk of death between patients on chronic renal replacement therapy who started with peritoneal dialysis compared to those who started with hemodialysis. Without adjustment, the relative risk of death of peritoneal dialysis patients was 0.65 (95% CI 0.58-0.73, p