Outcomes in patients hospitalized for heart failure and chronic obstructive pulmonary disease: differences in clinical profile and treatment between 2002 and 2009.

Aims: Heart failure (HF) and chronic obstructive pulmonary disease (COPD) frequently co-exist, and each is a major public health issue. In a large cohort of hospitalized HF patients, we evaluated: (i) the impact of COPD on clinical outcomes; (ii) whether outcomes and treatments changed from 2002 to 2009; and (iii) the relationship between outcomes and treatments focusing on beta-blockers (BBs) and bronchodilators (BDs).

Methods And Results: From linkable Lombardy administrative health databases, we selected individuals with a discharge diagnosis of HF with or without concomitant COPD (HF yesCOPD and HF noCOPD) in 2002 and 2009. Patients were followed up for 4 years. Outcomes were total mortality, first readmission for HF, and their combination. Unadjusted and adjusted Cox proportional models and competing risk analyses were used. We identified 11 274 patients with HF noCOPD and 2837 with HF yesCOPD. HF yesCOPD patients in 2002 and 2009 had a 20% higher risk of the outcomes. From 2002 to 2009, BB and BD prescriptions increased significantly. In HF noCOPD patients, risks for mortality [adjusted hazard ratio (HR) 0.91, 95% confidence interval (CI) 0.86-0.97], first HF readmission (HR 0.79, 95% CI 0.74-0.85), and the combined endpoint (HR 0.88, 95% CI 0.84-0.92) declined (all P < 0.003) while in HF yesCOPD only the risk for first HF readmission dropped (HR 0.86, 95% CI 0.76-0.97; P = 0.018). BBs were associated with significantly lower mortality in both groups, but with a higher risk for first HF readmission in HF noCOPD. Outcomes did not significantly differ in HF yesCOPD treated or not with BDs.

Conclusions: The prognosis of patients hospitalized for HF, either with or without COPD, seemed to improve between 2002 and 2009, with possibly better survival of those on BBs. Because of residual confounding in observational studies, a randomized controlled trial should be considered to confirm these results.