Simultaneous analysis of morbidity and mortality factors in chronic hemodialysis patients

Abstract

The Cox stepwise logistic regression model was applied to analyze 22 factors potentially affecting morbidity and mortality (MM) in a cohort of 104 patients on chronic hemodialysis (CHD). Two groups of predictor variables were considered: patients' characteristics at the start of the study, and treatment-related factors recorded throughout the observation period. End points were either failure (death or admission to a hospital) or success. Patients were followed for 24 months. Thirty-nine patients were hospitalized and seven died in the interval. The two leading causes of failure were cardiovascular and infectious complications. Variables significantly associated with the result were: cardiac status (score greater than 2, beta = 1.16), mean predialysis blood pressure (greater than 115 mm Hg, beta = 0.94), total dialysis dose (greater than 0.90, beta = -0.59) and age (greater than 55 years, beta = 0.51). The probability of failure was 0.13 for patients who presented the four variables in the lowest risk class. This increased to a maximum of 0.60 with one risk factor, to 0.91 with two risk factors, and to 0.99 with three or more risk factors. We conclude that, given the conditions for this study, two treatment-related variables of CHD (mean predialysis blood pressure and total dialysis dose) are MM factors even when simultaneously analyzed with other well-established predictors (cardiac status and age). Mean arterial pressure (MAP) is the most important CHD treatment-related MM predictor.