Factors Predicting Mortality in Elderly Patients on Dialysis

Abstract

Elderly patients are increasingly enrolled in dialysis programs and present a series of special characteristics due to their high morbidity and mortality. Objective:To describe the characteristics of incident dialysis patients aged >75 years, their comorbidities and their admissions, with a view to determining the factors that influence their course and mortality. Patients and Methods: The study included all patients aged >75 years who started dialysis in our center since January 2000. The mean duration of follow-up was 3.3 ± 2.2 years. Data were collected on incident comorbidity, admissions and their causes. A total of 139 patients were included, with a mean age of 78.6 ± 2.6 years (67.6% males, 33.8% diabetic and 7.9% on peritoneal dialysis). Three groups were established according to age: 75–79, 80–85 and >85 years. The most frequent comorbidities were chronic obstructive pulmonary disease (25.9%), ischemic heart disease (25.2%), heart failure (25.9%), neoplasms (23.7%), peripheral vascular disease (23.7%), cerebrovascular disease (18.7%) and arterial hypertension (81%). The Charlson index was calculated, not adjusted for age, and comorbidity tertiles were established. Results: During follow-up, the patients presented 0.82 ± 0.99 admissions/patient/year, with a duration of 12.1 ± 20.6 days/patient/year. The main causes of admission were infection (33%), vascular access problems (27%) and peripheral vascular events (14%). A total of 61 patients died (44%), and 4 underwent kidney transplantation (2.9%). The mean duration of follow-up of the transplanted patients was 3.6 ± 1.8 years. The main causes of death were infection (32%), cardiovascular problems (28.3%) and neoplastic disease (11.3%). The global survival rate was 90, 82 and 53% after 1, 2 and 5 years, respectively. No significant differences in survival or annual admission rate were found in relation to age group and dialysis technique. In contrast, the annual admission rate and days of admission were directly correlated to the Charlson index (p = 0.009 and p = 0.032, respectively). No significant differences in the Charlson index were found between the patients on hemodialysis and those subjected to peritoneal dialysis. In the univariate model, the factors associated to mortality were diabetes, chronic obstructive pulmonary disease, heart failure and the Charlson index. In the multivariate model, only the Charlson index remained as an independent predictive factor (p = 0.006). Conclusions: Unlike the general population, age does not influence mortality or admissions in elderly patients on dialysis. Incident comorbidity is the factor exerting the greatest influence upon mortality and admissions. Advanced age in itself should not be regarded as an excluding factor for starting dialysis.