Assessment of a simple scoring system applied to a screening questionnaire for asthma in children aged 5-15 yrs

The aim of the present study was to validate a simple scoring system using a parent-completed screening questionnaire to identify children aged 5-15 yrs who may have asthma. A stratified random sample of 157 children of 1,808 whose parents had answered a postal respiratory questionnaire underwent detailed clinical evaluation. The results were reviewed by three independent paediatricians whose opinions were combined to reach, for each child, decisions regarding three standards: 1) "possible asthma" defined as >50% likelihood of having asthma; 2) "possible asthma" defined as meriting a clinical trial of asthma medication; and 3) "probable asthma" defined as >90% likelihood of having asthma. The combined decisions were compared to three sets of questionnaire scores, in order to determine the positive predictive value, sensitivity and specificity of each set in identifying children with probable/possible asthma. The three sets of chosen questionnaire scores all had positive predictive values of 79-96% for predicting possible asthma, using either the combined expert opinion ">50% likelihood of asthma" or that of "warrants a trial of treatment" as the definition. This suggests that a low proportion of false positives would be obtained were this scoring system to be used for a screening programme. The combined decision >90% chance of asthma could be used as a means of estimating prevalence of asthma in the survey. When used for this, the prevalence of asthma in the surveyed population was 18.8% (95% confidence interval 13.1-26.3). In conclusion, the present scoring system, based on a simple respiratory questionnaire, provides a valid method of identifying children likely to have asthma, and who, if unknown to the medical services, would benefit from clinical review.