Abstract
Background: Chronic kidney disease (CKD) elevates the risk of cardiovascular and overall mortality among maintenance hemodialysis (MHD) patients. Although initially designed to predict stroke in cases of atrial fibrillation, the CHA2DS2-VASc score demonstrates predictive utility, spanning several cardiovascular conditions. This study seeks to evaluate if the CHA2DS2-VASc score is effective in predicting cardiovascular and all-cause mortality for MHD patients. Methods: The data are part of the "Prospective Study of the Prognosis of Patients on Chronic Hemodialysis" (PROHEMO) developed in Salvador, BA, Brazil. We grouped patients according to CHA2DS2-VASc score ≤2 (group 1) and >2 (group 2). Cox regression was used to estimate the hazard ratio of death (HR): unadjusted; and adjusted for hemoglobin, creatinine, albumin, phosphorus, PTH, liver disease, neoplasia/cancer, months of hemodialysis. Additionally, the distribution of each variable in the CHA2DS2-VASc score and its association with mortality were evaluated. Based on the observed associations and the distribution of age (with only 1.3% of patients aged >75 years) and hypertension (only 4.6% normotensive, leading to an imprecise association estimate), a modified CHA2DS2-VASc score was created. Results: A total of 237 patients on hemodialysis (51.57 ± 12.46, 57% male) were included. Mean age was 51.6 ± 12.5 years. A total of 55 deaths, 21 from cardiovascular causes. Compared with CHA2DS2-VASc score <2, the unadjusted hazard of death (Model 1) for score ≥2 was twofold for all-cause mortality (HR=2.05; 95% CI: 1.20, 3.49) and more than three times higher for cardiovascular deaths (HR=3.53; 95% CI: 1.46, 8.54). These HRs did not change substantially with adjustment for covariates. In the most comprehensively adjusted Cox model, the HR for all-cause mortality was 2.43 (95% CI: 1.38, 4.23) and for cardiovascular mortality was 3.52 (95% CI: 1.40, 8.84). These results were similar to those observe for the modified version of CHA2DS2-VASc score.