Integration of clinical and imaging data to predict death in hemodialysis patients

Abstract

In a prior publication, we demonstrated that a model integrating clinical and simple imaging data predicted the presence and severity of coronary artery calcification in prevalent hemodialysis patients. Herein we report the ability of the same model to predict all-cause death. We assessed all-cause mortality in 141 consecutive maintenance hemodialysis patients from two dialysis centers followed for a median of 79 months from enrollment. Patients were risk stratified according to a simple cardiovascular calcification index (CCI) that included patient's age, dialysis vintage, calcification of the cardiac valves, and abdominal aorta. The mean patients' age was 55 ± 14 years. Abdominal aorta calcification was present in 57% of the patients, and 44% and 38% had aortic and mitral valve calcification, respectively. During follow-up, 75 deaths (93 deaths per 1000 person-years) were recorded. The CCI was linearly associated with risk of death, such that the unadjusted hazard risk (HR) increased by 12% for each point increase in CCI (P < 0.001). Further adjustments for age, sex, study center, diabetes mellitus, history of cardiovascular disease, hypertension, congestive heart failure, left ventricular hypertrophy, systolic, and diastolic blood pressure did not substantially change the strength of this association (HR 1.10; 95%CI: 1.00-1.21; P = 0.03). The CCI is a simple clinical model that can be used to risk stratify maintenance hemodialysis patients.