Risk factors for mortality in long term hemodialysis

We studied a number of variables suspected of affecting the mortality rate during long term hemodialysis to identify those that were risk factors and to measure their relative weight We used the Cox model of proportional regressive risk to establish the contribution of each variable to the final result - survival or death. We analysed the longitudinal data obtained from 1992 to 1997 from all hemodialysis patients in three hospital dialysis units. Over the 5 year period we studied 236 patients (140 male/96 female, average age 56.5 +/- 15 years). The variables studied were age, sex, primary renal disease, comorbid conditions, KT/V, PCR, body mass index, arm muscle circumference, serum creatinine, total protein, albumin, prealbumin, C3, chelesterol and triglycerides, blood lymphocyte count and dietary intake. During the study 65 (27.5%) patients died. In those who died, age was higher comorbid conditions more common and KT/V, PCR, creatinine, arm muscle circumference, serum albumin and prealbumin, lymphocyte and caloric intake lower than in the survivors. The relative risk (RR) was: age 1,03, sex (male) 1.52, comorbid conditions 1.18 and plasma albumin 0.18. During the course of dialysis KT/V and PCR were increased but the biochemical and nutritional parameters hardly changed; once depressed they showed little recovery. The survival of patients starting dialysis depends an their condition when they start (age comorbid conditions). The should be given adequate dialysis and their nutritional state should be corrected. This action must be early and sustained since once nutritional parameters have worsened through inadequate dialysis or deficient intake of nutrients, they are difficult to restore.