Do people self-reporting information about chronic respiratory disease have corroborative evidence in their general practice medical records? A study of intermethod reliability.

AIMS: To use intermethod reliability to compare self-reported data about chronic respiratory disease and health service utilisation with data contained in general practice medical records.
METHODS: Self-reported postal questionnaire information from a small cohort of an age-sex stratified sample of 2318 patients was compared with information contained in their medical records. The agreement between the two sources of information was assessed.
RESULTS: The case notes of 115/135 individuals from eight general practices were examined. For self-reported chest injury or operation (kappa, kappa=-0.03), or chronic bronchitis (kappa=0.10), agreement was poor. Agreement for self-reported pleurisy (kappa=0.32), hay fever or rhinitis (kappa=0.40), or eczema or dermatitis (kappa=0.30) was fair; for chronic obstructive pulmonary disease (COPD) or emphysema (kappa=0.56), or heart trouble (kappa=0.54), agreement was moderate; for asthma (kappa=0.78) or pneumonia (kappa=0.62), agreement was good; and for pulmonary tuberculosis (kappa=0.88), agreement was very good. The strength of agreement for information about health service utilisation for respiratory problems ranged from moderate to very good and was good for smoking status.
CONCLUSIONS: Although based on small numbers, our results suggest good or very good agreement between self-reported data and general practice medical records for the absence or presence of some respiratory conditions and some types of respiratory-related health care utilisation. Depending on the research question being examined self-reported information may be appropriate.