Childhood asthma: can computers aid detection in general practice?

Background: Childhood asthma remains underdiagnosed in general practice. Computers with a patient interface have the potential to screen children for asthma in a time-efficient manner.

Aim: To develop a concise, validated self-report measure that calculates an 'asthma score' that predicts likelihood of asthma and its severity in childhood.

Design Of Study: Computerised questionnaire survey in general practitioners' (GPs') waiting rooms, followed by a written questionnaire and either bronchial challenge or skin allergy testing at the regional teaching hospital.

Setting: Children between 18 months and 18 years old accompanied by a parent or guardian in five group practices in Newcastle in New South Wales, Australia.

Method: The responses from both the computerised questionnaire and the written questionnaire were compared with physician assessment of asthma, based on an existing validated questionnaire and clinical tests.

Results: Six items were identified to be independently and significantly associated (at P < 0.05) with the presence of asthma and its severity: parent or self-reported asthma, previous diagnosis, wheeze in the past year, physical activity affected by symptoms, night cough in the past year, and visits to a GP in the past year. From the regression model a linear score was derived that indicates whether a child is likely to have asthma and its likely severity.

Conclusions: The asthma score is a valid indicator of asthma and its severity in children in general practice.