Adverse effects of ACE inhibitors in patients with chronic heart failure and/or ventricular dysfunction : meta-analysis of randomised clinical trials.

BACKGROUND

The evidence-based benefit/risk evaluation of therapeutic interventions in randomised clinical trials should include both the assessment of the benefits and of the adverse outcomes. There is ample evidence that ACE inhibitors improve the symptoms and prognosis of chronic heart failure (CHF) and ventricular dysfunction. However, there is little systematic information on the tolerability and adverse effects associated with their use in these conditions.

OBJECTIVE

To estimate the adverse events related to ACE inhibitor use in the treatment of CHF and ventricular dysfunction.

DESIGN AND METHODS

Description of adverse events in reports of randomised clinical trials of ACE inhibitors in CHF or ventricular dysfunction was examined, and a meta-analysis was performed. Trials were included if they were placebo- or standard treatment-controlled, and if the treatment duration was at least 8 weeks. Relative risks and their 95% CIs were estimated with a random effects model.

RESULTS

Only 22 (43%) of 51 original reports contained information on the number of withdrawals and their causes. Missing information from the remaining 29 trials was obtained from the authors. The weighted mean duration of treatment was 100.2 weeks. After excluding administrative reasons, heart failure, myocardial infarction and hypertension, the withdrawal rates attributed to adverse events were 13.8% and 9.4% for the ACE inhibitor and control groups, respectively (RR = 1.54 [95% CI 1.30-1.83]; weighted difference = 3.1 per 100 treated patients [95% CI 1.8-4.4]). Cough, hypotension, renal dysfunction, dizziness, hyperkalaemia, and impotence were all significantly more prevalent among patients treated with ACE inhibitors than among those in the control groups.

CONCLUSIONS

Among patients with CHF or ventricular dysfunction enrolled in randomised clinical trials, treatment with an ACE inhibitor for an average of 2 years leads to an additional 3% of treatment withdrawals. In a significant proportion of the reports on these randomised clinical trials, information on adverse events leading to treatment withdrawal was inadequate. Proper evidence-based evaluation of the benefit/risk of therapeutic interventions needs a more systematic approach to reporting of adverse events experiences recorded in clinical trials.