Analysis of risk factors for mortality of incident patients commencing dialysis in East Yorkshire, UK

Aim: To examine various potential factors that may predict early and overall mortality. Design and Methods: We carried out an observational prospective study of a cohort of incident patients starting dialysis in a UK centre. Univariate analysis of factors and co-morbidities potentially affecting survival on dialysis were analysed to potential predictors. Factors affecting 1 year mortality were analysed using the t-test, the Mann-Whitney U-test or the chi-square test as appropriate. Mortality over the 5-year follow-up period was analysed using the Kaplan-Meier method. Results: Ninety-four patients [predominantly Caucasian (98%)], of mean age 63 years (15.6) (56% > 65 years) with a slight male preponderance were studied. Vascular disease (39%) and sepsis (33%) accounted for most of the deaths and a significant proportion of mortality was seen in the first year (56%). Patients with early mortality were older (68 vs. 61 years, P = 0.05) with lower haemoglobin (8.4 vs. 9.4 g/dl, P = 0.01) at the start of dialysis, commenced dialysis with a lower eGFR (5.4 vs. 6.5 ml/min/1.73 m(2), P = 0.06) and had more peripheral vascular disease (PVD) (39% mortality in patients with PVD vs. 18.5% in those without PVD, P = 0.04). Diabetes mellitus, high calcium phosphate product, older age and presence of vascular co-morbidities including ischaemic heart disease and peripheral vascular disease were associated with overall mortality over the 5-year follow-up period. Summary: In this study, elevated calcium phosphate product and diabetes mellitus in addition to the presence of vascular disease were associated with poor survival. Patients with low haemoglobin and lower first pre-dialysis eGFR suffered higher early mortality. These potentially modifiable factors that could be identified in the pre-dialysis stage provide a valuable opportunity for intervention.