Identification of asthmatic children using prescription data and diagnosis.

OBJECTIVE: The aim of the study was to develop and validate a method for identifying asthmatic children between 6 and 14 years of age based on prescription data on anti-asthmatic drugs and diagnostic data.
METHODS: A register-based study of 125,907 Danish children aged 6-14 years identified 9695 children who had redeemed at least one anti-asthmatic drug prescription in 2002. The asthma diagnosis in these children was validated by discharge information and by a questionnaire completed by general practitioners. Models based on combinations of different types of drugs were tested to find the best model that would include as many children as possible with a validated diagnosis and exclude as many false positives as possible. Different time windows were tested in terms of detecting the children and the observation period of refilling prescriptions.
RESULTS: The highest specificity of 0.86 [95% confidence interval (95% CI): 0.84-0.87] together with a sensitivity of 0.63 (95% CI: 0.62-0.65) were found in the model that included children who had redeemed a prescription for any anti-asthmatic drug - with the exception of prescriptions for beta2-agonists as liquid, one prescription only of inhaled beta2-agonist or an inhaled steroid - during a 12-month period. Lengthening the observation time by 6 months did not significantly improve the specificity (0.87; 95% CI: 0.85-0.88), but it did result in a statistically significantly lower sensitivity (0.59; 95% CI: 0.58-0.60).
CONCLUSION: Register-based data on redeemed prescriptions can be utilised to identify asthmatic school children. This method will be useful in health services research and in the proactive care of asthmatic children.