Observed versus expected mortality in a network of dialysis units in Colombia

Abstract

Background: mortality in dialysis is one of the main indicators of clinical management and is influenced by various socio-demographic and clinical factors. Objectives: to calculate the observed versus expected mortality in the dialysis units of the RTS network in Colombia.

Methods: a historical cohort of patients older than 18 years, prevalent on dialysis between January 1 and December 31, 2012, of 51 renal units of the Renal Therapy Services (RTS) network. The standardized mortality ratio (SMR) was calculated following the methodology proposed by the University of Michigan Center for Costs and Renal Epidemiology (UM-KECC); a Cox proportional hazards survival model was implemented in two stages, the first estimated the parameters associated with the explanatory variables and the second standardized the results.

Results: 9798 patients were evaluated, 4125 (42.1%) were women. The mean age was 59 years (SD = 15.6). There were 1253 death events (12.7%). The model gave a value of 1067 expected deaths, with an estimated SMR value of 1.17 (95% CI: 1.11-1.24). SMR was greater for diabetic patients 1.28 (95% CI: 1.19-1.38) and women (1.36 (95% CI: 1.25-1.48)) and it varied significantly depending on the region of the country (1.11 to 2.0)

Conclusions: important differences in SMR according to sex, presence of diabetes and by regions of the country were found. Further studies are required to better understand the influence of these and other variables on the mortality phenomenon in dialysis in our context