Baseline factors associated with congestive heart failure in patients receiving etoricoxib or diclofenac: multivariate analysis of the MEDAL program

Abstract

Non-steroidal anti-inflammatory drugs have been associated with increased risk of congestive heart failure (CHF). We aimed to assess the impact of treatment with etoricoxib or diclofenac on risk of CHF relative to baseline risk factors. We performed a multivariate analysis of 34 701 patients with arthritis receiving etoricoxib 60 or 90 mg, or diclofenac 150 mg, daily for a mean of 18 months, to assess the incidence of confirmed, adjudicated CHF events resulting in emergency room visit or hospitalization. Analyses were performed using a Cox proportional hazard model to evaluate the hazard ratio (HR) between the levels of each risk marker for the incidence of CHF. Significant risk markers included history of CHF (HR: 6.69, 95% CI 3.59-12.47; P <0.0001), age > or = 65 years (2.56, 1.65-3.98; P <0.0001), and history of hypertension (1.83, 1.16-2.89; P = 0.0094) or diabetes (1.83, 1.15-2.94; P = 0.0116). Etoricoxib vs. diclofenac was a significant risk factor only when pooling the etoricoxib 90 mg cohorts (1.88; 1.13-3.10; P = 0.0143). Etoricoxib 60 mg did not significantly increase risk vs. diclofenac. History of CHF was highly associated with risk for CHF hospitalization. Hypertension, diabetes, and older age also increased risk modestly. There appeared to be a dose-related increase in CHF with etoricoxib compared with diclofenac, which reached statistical significance when the etoricoxib 90 mg groups (osteoarthritis and rheumatoid arthritis) were pooled.