Are patients in heart failure trials representative of primary care populations? A systematic review

Background Guidelines recommend drug treatment for patients with heart failure with a reduced ejection fraction (HFrEF), however the evidence for benefit in patients with mild disease, such as most in primary care, is uncertain. Importantly drugs commonly used in heart failure account for one in seven of emergency admissions for adverse drug reactions. Aim To determine to what extent patients included in studies of heart failure treatment with beta blockers, ACE inhibitors and aldosterone antagonists were representative of a typical primary care population with HFrEF in England. Design and Setting Systematic review of RCTs of drug treatment in patients with HFrEF. Method MEDLINE, MEDLINE In-process, EMBASE, and CENTRAL were searched from inception to March 2015. We compared the characteristics of the patient's NYHA classification with a primary care reference population with HFrEF. Results 30 studies were included. Two had incomplete data. None had a 'close match' (<10% deviation from reference study) for NYHA class I disease, 5/28were a close match for NYHA class II, 5/28 for NYHA class III, and 18/28 for NYHA class IV. In general, pre-existing cardiovascular conditions, risk factors and comorbidities were representative of the reference population. Conclusion Patients recruited to studies typically had more severe heart failure than the reference primary care population. When evidence from sicker patients is generalised to less sick people, there is increased uncertainty about benefit and also a risk of harm from overtreatment. More evidence is needed on the effectiveness of treatment of heart failure in asymptomatic patients with NYHA class 1.