Distinguishing Among Causes of Death for Patients with Kidney Failure on Hemodialysis.

Abstract

Patients treated with maintenance hemodialysis (HD) are at high risk of death from a variety of causes. To identify markers (i.e., risk phenotypes) that distinguish among causes of death, we used dialysis electronic health record data for a cohort of adults treated with maintenance in-center HD who died between 2003-2016. Patients were linked to the United States Renal Data System (USRDS) Files. We classified USRDS-reported causes of death into five categories: Sudden Cardiac Death (SCD), non-SCD Cardiovascular Death, Infection, Others, and Unknown. A sub-cohort was linked to the National Death Index (NDI) with similar categories defined. We used ensemble classification trees to discriminate among causes using demographics, vital signs, laboratory measures, health service utilization, and comorbidity claims from 30 days prior to death. The area under the receiver operating curves (AUCs) were all between 0.59-0.70, suggesting minimal ability to distinguish among causes using clinical risk markers. We then created nested case-control populations for each cause of death and used ridge logistic regression to evaluate clinical risk markers that associate with distinct causes. Model coefficients were similar and highly correlated across different cause of death models (i.e., 0.87 - 0.94). This suggests that most clinical risk markers are shared across causes without distinct risk phenotypes. We conclude that different causes of death may share similar clinical risk markers in the setting of kidney failure or that the causes of death attributed on USRDS or NDI forms are not precise.