Impact of the clinical conditions at dialysis initiation on mortality in incident haemodialysis patients: a national cohort study in Taiwan

Background. Glomerular filtration rate (GFR) and co-morbidity at dialysis initiation in relation to mortality in end-stage renal disease is still controversial. We studied factors potentially related to the mortality in incident haemodialysis (HD) patients. Methods. A national database included 23 551 incident HD patients from July 2001 to December 2004. Kaplan–Meier and Cox regression analyses were performed to assess the association between GFR estimated by the four-variable Modified Diet in Renal Disease equation and all-cause mortality. Analyses were performed from Day 91 after the start of dialysis. Patients were classified into five groups (quintiles) based on estimated glomerular filtration rate (eGFR) at the start of dialysis. Results. The median eGFR at dialysis initiation was low (4.7 mL/min/1.73 m2), as was the mortality in the first year of dialysis [13.2/100 patient-year, 95% confidence interval (95% CI) = 12.8–13.7]. There was an inverse association between lower eGFR and higher survival rate. The Cox regression model revealed an increase in mortality risk in Q5 (hazard ratio [HR] = 2.44, 95% CI = 2.11–2.81), Q4 (HR = 1.66, 95% CI = 1.43–1.93), Q3 (HR = 1.21, 95% CI = 1.04–1.41) and Q2 (HR = 1.18, 95% CI = 1.01–1.37) compared with the reference group of Q1 after adjusting for year of application, primary diseases (chronic glomerulonephritis, diabetic nephropathy, hypertension, chronic tubulointerstitial nephritis and others), demographics (age, sex), presence of co-morbidity (diabetes mellitus, hypertension, congestive heart failure, ischaemic heart diseases, cerebrovascular diseases, malignancies, liver cirrhosis, tuberculosis, other diseases and free of reported of co-morbidities) and haematocrit. Propensity score analysis also showed a higher eGFR to be associated with increased mortality risks. Adjustment for all covariates explained a high percentage of excess risk of mortality in the groups with low eGFR, but less risk in the groups with higher eGFR. Conclusions. Lower eGFR at dialysis initiation is associated with lower mortality. Conditions at dialysis initiation explained excess 1-year mortality risk differently in patients who began dialysis at different levels of eGFR. Other factors likely contribute to the mortality of patients initiating dialysis at higher eGFR levels, and further study is needed.