Original ResearchEffects of Rivaroxaban Versus Warfarin on Hospitalization Days and Other Health Care Resource Utilization in Patients With Nonvalvular Atrial Fibrillation: An Observational Study From a Cohort of Matched Users

PurposeCompared with warfarin, the new target-specific oral anticoagulant agents may have advantages, such as shorter hospital length of stay, in patients with nonvalvular atrial fibrillation (NVAF). The objective of the present study was to assess, among patients with NVAF, the effects of rivaroxaban versus warfarin on the number of hospitalization days and other health care resource utilization in a cohort of rivaroxaban users and matched warfarin users.MethodsData from health care claims dated from May 2011 to December 2012 from the Humana database were analyzed. Adult patients newly initiated on treatment with rivaroxaban or warfarin, with ≥2 diagnoses of AF (ICD-9-CM code 427.31), and without valvular AF were identified. Based on propensity score methods, warfarin recipients were matched 1:1 to rivaroxaban recipients. The end of the observation period was defined as the end of data availability, the end of insurance coverage, death, the date of a switch to another anticoagulant agent, or day 14 of treatment nonpersistence. The total number of hospitalization days and other health care resource utilization parameters (numbers of hospitalizations, emergency department [ED] visits, and outpatient visits) were evaluated using the method by Lin et al.FindingsMatches for all rivaroxaban recipients were found, and the characteristics of the matched groups (n = 2253 per group) were well balanced. The mean age of both cohorts was 74 years; 46% were female. The estimated mean total numbers of hospitalization days were significantly less in rivaroxaban users compared with those in warfarin users (all-cause, 2.71 vs 3.87 days [P = 0.032]; AF-related, 2.11 vs 3.02 days [P = 0.014]). The numbers of outpatient visits were also significantly less (all-cause, 25.26 vs 35.79 visits [P < 0.001]; AF-related, 5.48 vs 9.06 visits [P < 0.001]). Rivaroxaban users had a lesser estimated mean number of all-cause hospitalizations compared with warfarin users (0.55 vs 0.73; P = 0.084), and a significantly lesser estimated mean number of AF-related hospitalizations (0.40 vs 0.57; P = 0.022). The difference in the estimated mean numbers of all-cause ED visits was not statistically significant between the rivaroxaban and warfarin users.ImplicationsIn this study conducted in clinical practice, the estimated mean numbers of hospitalization days, outpatient visits, and AF-related hospitalizations associated with rivaroxaban were significantly less than were those associated with warfarin in these patients with NVAF. The corresponding estimated difference in all-cause ED visits was not statistically significant.