Improved Outcomes in Heart Failure Treated With High-Dose ACE Inhibitors and ARBs: A Population-Based Study

Elevated doses of angiotensin II–converting enzyme (ACE) inhibitors and angiotensin receptor blockers (ARBs) have similarly reduced morbidity and mortality in congestive heart failure (CHF) trials.^1,2 However, despite the recommendations of consensus CHF guidelines to achieve elevated target doses of ACE inhibitors or ARBs,^3,4 patients often receive doses that are lower than those used in large clinical trials, possibly owing to adverse effects.^2,5,6 We conducted a population-based retrospective cohort study to estimate the effect, in real-world clinical practice, of different doses of ACE inhibitors and ARBs on all-cause mortality and CHF readmission in patients with a first CHF admission.

Data on all patients 65 years or older who were admitted for a first CHF diagnosis in the province of Quebec, Canada, between January 1, 1998, and March 31, 2007, were obtained from the hospital discharge summary database of Quebec and the provincial physician and drug claims databases. All patients had CHF recorded as the primary diagnosis (International Classification of Diseases, Ninth Revision, code 428). Patients were included in the cohort if they had a first CHF-related hospital admission, were discharged alive, and filled prescriptions for an ACE inhibitor (captopril, cilazapril, enalapril, fosinopril, lisinopril, perindopril, quinapril, ramipril, and/or trandolapril) or ARB (candesartan, eprosartan, irbesartan, losartan, telmisartan, and/or valsartan) after discharge. Follow-up for each patient was from the time of first prescription filled for any ACE inhibitor or ARB (time 0) to the time of death, CHF readmission, end of the study period, or switch to a different drug class. Three study groups were formed according to the initial dose of ACE inhibitor or ARB prescribed: low-, medium-, and high-dose groups.

In addition to descriptive statistics on baseline characteristics, multivariate Cox proportional hazards models were used to assess the associations between dose and all-cause mortality or the composite end point of all-cause mortality or CHF readmission. The medium-dose groups for ACE inhibitors and ARBs were used as the reference, and adjusted hazard ratios (HRs) and 95% CIs were estimated for the low- and high-dose groups, with adjustments made for all baseline characteristics and potential confounders, including cardiovascular and lung diseases, chronic kidney and liver conditions, diabetes mellitus, dementia, malignancy, concurrent use of other cardioprotective medications, in-hospital therapeutic procedures, length of hospital stay, specialty of the treating physician, year of hospitalization, and time to the first ACE inhibitor or ARB prescription. Sensitivity analyses were also performed among dose groups, which were adjusted with propensity score analysis to further control for selection bias and potential residual confounding.⁷

The study included 43 405 individuals (mean [SD] age, 79.5 [7.5] years; 45% men). After discharge, 73% filled a prescription for an ACE inhibitor, 27% for an ARB. Of these, 29% received a low-dose prescription of either drug. Compared with patients in the low- and medium-dose groups, those receiving a high dose of either drug were more likely to have hypertension (52% vs 40%) and diabetes mellitus (45% vs 35%), whereas, patients receiving low doses of either drug were more likely to have renal disease (31% vs 21%). The rest of the baseline characteristics were similar among the 3 groups.

After adjusting for all baseline variables and potential confounders, we found that low-dose users of either drug class had significantly higher mortality and CHF readmissions than patients who filled prescriptions for medium or high doses (Figure). Treatment with high-dose ACE inhibitors significantly decreased mortality and the composite end point, while high-dose ARB treatment improved mortality and the composite end point only when compared with middle- and low-dose treatment combined. Sensitivity analyses between dose groups were adjusted with propensity scores, and Cox regression models were performed using the propensity score–matched cohorts. In these analyses, treatment with high-dose ACE inhibitors and ARBs reduced mortality and CHF readmissions more than the medium- and low-dose treatments combined (Figure).

To our knowledge, this is the first study to estimate the effect of different doses of ACE inhibitors and ARBs on all-cause mortality and CHF readmission in patients 65 years or older with a first CHF admission. Unlike clinical trials, our study included a representative sample of unselected patients with CHF and reflects real-world clinical practice. We demonstrated that, of over 43 000 patients with CHF, approximately one-third were prescribed low doses of ACE inhibitors or ARBs. Low-dose users had significantly greater all-cause mortality and CHF readmissions. Both ACE inhibitors and ARBs decreased mortality and the composite end point in a dose-dependent manner, with high-dose users having the best outcome. However, ACE inhibitors were more effective than ARBs at reducing the composite end point. Our results demonstrate that target doses of ACE inhibitors or ARBs are reached in only one-third of patients with CHF. Physicians should strive, whenever possible, to treat patients with CHF with high doses of ACE inhibitors or ARBs to improve outcomes.

Correspondence: Dr Daskalopoulou, Division of General Internal Medicine, Department of Medicine, McGill University Health Centre, Montreal General Hospital, 1650 Cedar Ave, B2.101.4, Montreal, QC H3G 1A4, Canada (stella.daskalopoulou@mcgill.ca).

Published Online: July 2, 2012. doi:10.1001 /archinternmed.2012.2514

Author Contributions:Study concept and design: Daskalopoulou. Acquisition of data: Pilote and Daskalopoulou. Analysis and interpretation of data: Egiziano, Behlouli, and Daskalopoulou. Drafting of the manuscript: Egiziano and Daskalopoulou. Critical revision of the manuscript for important intellectual content: Egiziano, Pilote, Behlouli, and Daskalopoulou. Statistical analysis: Pilote, Behlouli, and Daskalopoulou. Obtained funding: Pilote. Administrative, technical, and material support: Daskalopoulou. Study supervision: Daskalopoulou.

Funding/Support: This study was funded in part by grant MOP-84304 from the Canadian Institutes of Health Research. Drs Pilote and Daskalopoulou are supported by the Fonds de la recherche en santé du Québec.