Initiating angiotensin converting enzyme inhibitors in mild to moderate heart failure in general practice: randomised, placebo controlled trial

BMJ 1998; 317 doi: (Published 14 November 1998) Cite this as: BMJ 1998;317:1352
  1. Murray Lough, general practitioner (Administrator{at},
  2. John Cleland, British Heart Foundation senior research fellowb,
  3. John Langan, general practitionerc,
  4. Alan Cowley, consultant cardiologistd,
  5. Alan Wade, general practitionere
  1. aAirdrie Health Centre, Monkscourt Avenue, Airdrie ML6 0JU
  2. bClinical Research Initiative in Heart Failure, University of Glasgow, Glasgow G12 8QQ
  3. cBailleston Health Centre, Glasgow G69 7AD
  4. dQueen's Medical Centre, Nottingham NG7 2UH
  5. eClydebank Health Centre, Glasgow G81 2TQ
  1. Correspondence to: Dr Lough
  • Accepted 28 April 1998

Less than 50% of patients with heart failure in community practice receive an angiotensin converting enzyme inhibitor and then only usually at the instigation of or when prescribed by a hospital doctor.1-3 Fear of side effects seems to be a barrier to starting treatment with angiotensin converting enzyme inhibitors, 3 4 reflecting the lack of substantial studies to show the safety of giving them in primary care.

Methods and results

General practitioners in 47 practices in the United Kingdom recruited patients with mild to moderate heart failure who had been receiving chronic diuretic treatment. Exclusioncriteria were age >80 years, frusemide dose >100 mg/day, systolic arterial pressure <100 mm Hg, serum creatinine concentration >250 £mol/l, and sodium concentration <135 mmol/l.

Plasma concentrations of N-terminal atrial natriuretic peptide were measured from blood samples taken before randomisation at a core laboratory.5 Plasma concentrations >2.8 ng/ml indicated important cardiac dysfunction.6

Patients were randomised, double blind, to receive quinapril 5 mg or placebo. Blood pressure was monitored for 3 hours. Quinapril or matching placebo was subsequently titrated over 3 days to 20 mg/day. After 1 week patients were reassessed and then, without breaking blinded-treatment, all received 5 mg of quinapril. Patients were again monitored, titrated to 20 mg of open label quinapril, and reviewed after a further week.

The original intention was to recruit 1000 patients giving a 95% probability of observing any adverse event with a frequency <0.34%. The power of the study was reduced because only 178 patients were randomised, 96 of whom received placebo initially. The study was terminated because of slow recruitment. Plasma concentrations of NT-ANP are shown in the figure.


N-terminal atrial natriuretic peptide (NT-ANP) versus age in study population. Lower line (2.8 ng/ml) indicates optimal value for distinguishing between patients with and without major left ventricular systolic dysfunction (ejection fraction ≤ 30%) in a population survey6; upper line (4.8 ng/ml) is median in study. Assays in this study and from the population survey were performed in the same laboratory

No serious adverse events occurred within 24 hours of starting quinapril. Blood pressure fell to a nadir of 133/78 mm Hg in patients receiving quinapril and 138/82 mm Hg in those receiving placebo at 2 hours post dose. Eleven patients (13.4%) randomised to receive quinapril and 5 (5.2%) receiving placebo had an asymptomatic fall in systolic blood pressure >20 mm Hg or to <90 mm Hg (all predefined). After one week serum creatinine concentration did not differ from baseline (112 £mol/l (interquartile range 98-122 £mol/l) with quinapril and 110 £mol/l (92-120 £mol/l) with placebo).


This is the first placebo controlled clinical trial reporting the safety of starting angiotensin converting enzyme inhibitors in primary care for selected patients with heart failure. Although the study was stopped prematurely because of slow recruitment (which may reflect continuing safety concerns), a frequency of serious adverse events of >2% (one in 50 initiations) was excluded. Furthermore, monitoring blood pressure after the first dose of quinapril seems unnecessary in appropriately selected patients.

The certainty of the diagnosis of heart failure in this study is of concern but the aim was to reflect the usual clinical setting. Access to echocardiography is limited and only the minority of patients undergo echocardiography.2 Although diuretics that lower plasma concentrations of atrial natriuretic peptide were used,6 76% of patients had concentrations of N-terminal atrial natriuretic peptide that seem diagnostic of important ventricular dysfunction in epidemiological studies.7 This implies that general practitioners identified patients with cardiac dysfunction reasonably accurately in this study, although precise identification of the cause of dysfunction still requires echocardiography.


  • Funding All drugs for the study were supplied by Warner-Lambert Pharmaceuticals.

  • Conflict of interest None.


View Abstract