The importance of validating the diagnosis of coronary heart disease when measuring secondary prevention: a cross-sectional study in general practice

Pharmacoepidemiol Drug Saf. 2002 Jun;11(4):311-7. doi: 10.1002/pds.709.


Purpose: To compare levels of recorded risk factors and drug treatment between patients with validated and non-validated diagnoses of coronary heart disease (CHD) in Northern Ireland.

Methods: Patients with a nitrate prescription in the previous year or a CHD Read code were identified from computer records of 25 practices, stratified by partnership size and area board. Computer and paper records of a random sample of 10% of these were searched for specified criteria to validate the diagnosis of CHD. The diagnosis was considered valid if the patient was found to have one or more positive investigations for CHD. Records of blood pressure, cholesterol, blood sugar, body mass index and drugs prescribed were taken into account.

Results: The combined practice population was 151,071; 7338 (4.86%) were identified by the computer search as meeting the defined entry criteria for CHD. Among the 10% random sample the diagnosis of CHD could not be validated for 36.5% (265/727). Significantly more patients with a validated than non-validated diagnosis had recorded cholesterol levels below 5.0 mmol/l (55.8 vs. 34.5%, p < 0.001) and were prescribed aspirin (75.3 vs. 40.8%, p < 0.001), beta-blockers (51.5 vs. 28.3%, p < 0.001), angiotensin-converting-enzyme inhibitors (29.2 vs. 15.5%, p < 0.001) and lipid-lowering drugs (50.9 vs. 23.0%, p < 0.001). A recent nitrate prescription had a higher predictive value for validated CHD than a Read code for CHD alone (71.2 vs. 53.1%, p < 0.001). No other significant differences were found between the two groups regarding the extent or levels of recorded risk factors.

Conclusions: Patients with a validated diagnosis of CHD appear to be better managed than those whose diagnosis has not been confirmed. Validation of diagnosis has important implications for assessing the provision of secondary prevention and for clinical governance.

Publication types

  • Research Support, Non-U.S. Gov't

MeSH terms

  • Adolescent
  • Adult
  • Aged
  • Blood Glucose
  • Body Mass Index
  • Child
  • Child, Preschool
  • Cholesterol / blood
  • Coronary Disease / diagnosis*
  • Coronary Disease / drug therapy
  • Coronary Disease / epidemiology
  • Diagnostic Errors
  • Family Practice*
  • Humans
  • Infant
  • Infant, Newborn
  • Medical Records Systems, Computerized
  • Middle Aged
  • Northern Ireland / epidemiology
  • Prevalence
  • Preventive Health Services
  • Reproducibility of Results
  • Risk Factors


  • Blood Glucose
  • Cholesterol