Related Experiment Videos
Validation of general practitioner-diagnosed COPD in the UK General Practice Research Database
J B Soriano1, W C Maier, G Visick
1Worldwide Epidemiology, GlaxoSmithKline Research and Development, Greenford, Middlesex, UK. jbs42397@gsk.com
European Journal of Epidemiology
|January 18, 2003
Summary
A computer algorithm accurately identifies chronic obstructive pulmonary disease (COPD) and its severity in the UK General Practice Research Database, proving useful for screening and differentiating COPD from asthma.
Area of Science:
- Epidemiology
- Health Informatics
Background:
- Large automated databases offer insights into respiratory disease natural history.
- Validating definitions within these databases is crucial for accurate research.
Purpose of the Study:
- To assess a computer algorithm's accuracy in identifying chronic obstructive pulmonary disease (COPD) and its severity using the UK General Practice Research Database (GPRD).
- To evaluate the algorithm's utility in distinguishing COPD from asthma.
Main Methods:
- A GPRD algorithm identified COPD patients by diagnostic codes and drug use for severity grading.
- 225 COPD patients and 75 asthma patients were randomly selected.
- General practitioners (GPs) confirmed diagnoses and severity via questionnaires; agreement was measured using the kappa index.
Main Results:
- The GPRD algorithm showed moderate agreement with GP records for COPD diagnosis (kappa=0.52) and severity (kappa=0.54).
- Agreement improved with increasing COPD severity (kappa up to 0.68).
- Consistent agreement was observed across various demographic and clinical subgroups.
Conclusions:
- GPRD algorithms serve as effective screening tools for COPD in the UK.
- These algorithms satisfactorily differentiate COPD from asthma, especially in moderate to severe cases.