Hospital admissions in elderly patients on chronic hemodialysis

The purpose of this study was to determine whether number of hospital admissions per patient per year (n/[pt-yr]) and hospital days per patient per year (d/[pt-yr]) differ between elderly and younger patients on chronic hemodialysis (HD). In a retrospective cohort analysis of incident HD patients in one dialysis unit over 15 years, we compared 166 HD patients older than 70 years (77.1 +/- A 4.7 yrs) at the onset of HD (group A) and 216 patients younger than 70 years both at onset (57.1 +/- A 7.6 yrs) and at the end of the HD period (group B). Eighty (48.2%) of group A and 141 (65.3%) patients of group B had diabetes mellitus. No differences were noted in the overall hospitalization rate, presented as mean, {95% Confidence interval} (group A 2.40 {2.04-2.75}, group B 2.03 {1.89-2.16} n[pt-yr]) and days/[pt-year] (group A 33.6 {25.3-41.8}, group B 24.1 {18.9-29.23}). Group A had higher number of hospitalization days (P = 0.012) for surgery or trauma and higher rate (P = 0.045) and days (P = 0.041) of hospitalization for miscellaneous causes, primarily pulmonary disease, or malignancy. Among diabetic patients, group A had only a greater number of hospital days for cardiac disease (P = 0.050). Among patients without diabetes, group A had a higher number for hospital days for surgery or trauma (P = 0.027). All other univariate comparisons were not significant. Multiple linear regression identified comorbidity, quantified by the Charlson index, Caucasian race and poor compliance with the HD schedule as predictors of admission rate and days per year for vascular access issues and comorbidity, poor compliance, and advanced age at onset of HD as predictors of admission for causes other than vascular access related. Hospitalizations, which affect quality of life, differ little between elderly and younger patients on HD. Therefore, hospitalizations do not constitute an argument for restricting access to HD to elderly patients.