A Multimorbidity-Based, Risk-Stratified Reanalysis of the Atrial Fibrillation Follow-Up Investigation of Rhythm Management (AFFIRM) Trial.

Abstract

Multimorbidity is common in patients with cardiovascular disease. Clinical trials in cardiovascular medicine mostly enroll patients who are younger, healthier, and more affluent than average patients with the condition of interest. These trials rarely account for patient-level multimorbidity in a systematic fashion. Further, treatment effect heterogeneity is usually tested across subgroups of patients based on the presence or absence of individual variables, not on the basis of summative risk scores that account for multimorbidity. Thus, the impact of multimorbidity on treatment effects is poorly understood. In this study, we performed a multimorbidity-based risk-stratified reanalysis of the AFFIRM (Atrial Fibrillation Follow-Up Investigation of Rhythm Management) trial. Our objectives were to describe the distribution of multimorbidity using a modified version of the Charlson Comorbidity Index (mCCI), scale 0-14, and to assess its impact on the original primary endpoint of all-cause mortality. The majority of patients in the AFFIRM trial had an mCCI score of ≤4 (55.5%), and there was no statistically significant difference in the risk of death for rate versus rhythm control in these patients (7.9 vs. 8.8%; p=0.44). However, for patients with an mCCI ≥5 (44.5%), there was a strong trend toward a reduction in death with rate control that nearly reached statistical significance despite being underpowered (24.5 vs. 28.3%; p=0.07). This proof-of-concept study supports the idea that clinical trials in cardiovascular medicine should systematically assess for multimorbidity and investigate its potential impact on treatment effects.