Mortality associated with cardiovascular drugs in patients with chronic obstructive pulmonary disease and right-sided heart failure: A danish nationwide registry-based study

Publikation: Bidrag til tidsskriftTidsskriftartikelForskningfagfællebedømt

Background: The optimal medical treatment in patients with chronic obstructive pulmonary disease (COPD)and right-sided heart failure (RHF)is unknown. We aimed to estimate the risks of all-cause mortality associated with the current clinical use of various cardiovascular drugs in this patient-group. Methods: We followed all patients with registered COPD and RHF (defined as a diagnosis of pulmonary hypertension plus use of loop-diuretics)for the risk of all-cause mortality (Jan 1, 1995 to Dec 31, 2015)using the Danish nationwide administrative registries. The association between mortality and claimed prescriptions for cardiovascular drugs was assessed by multivariable Cox regression models. Results: 5991 patients (mean age 74 ± standard deviation 10 years, 51% women)were included. Of these, 1440 (24%)used beta-blockers, 2149 (36%)renin-angiotensin system inhibitors [RASi], 1340 (22%)oral anticoagulants, 1376 (23%)calcium channel blockers, 1194 (20%)statins, 1824 (30%)spironolactone, and 2099 (35%)low-dose aspirin. During an average follow-up of 2.2 years (±standard deviation 2.8, min-max 0–19.6 years), 5071 (85%)died, corresponding to a mortality rate of 38 per 100 person-years (95% confidence interval 37–39). Compared to no use, beta-blockers were associated with adjusted hazards ratio 0.90 (95% confidence interval 0.84–0.98), RASi 0.92 (0.86–0.98), calcium channel blockers 0.86 (0.80–0.92), spironolactone 1.17 (1.10–1.24), statins 0.85 (0.78–0.92), oral anticoagulants 0.87 (0.79–0.95), and aspirin 0.99 (0.93–1.05). Propensity-score matched analyses and inverse-probability-weighted models yielded similar results. Conclusion: Several cardiovascular drugs may be associated with lowered mortality in COPD and RHF. Given the grave prognosis, randomized clinical trials are warranted to test this hypothesis.

OriginalsprogEngelsk
TidsskriftEuropean Journal of Internal Medicine
Vol/bind63
Sider (fra-til)56-61
ISSN0953-6205
DOI
StatusUdgivet - 2019

ID: 239959180