Background: The relationship between symptoms and objectively defined asthma is unclear. Also, the predictive value of symptoms to determine response to inhaled steroid (ICS) treatment has not previously been assessed. Aims: To establish the discriminatory value of symptoms for (1) diagnosis of asthma based on positive methacholine challenge (PC 20 ) or significant bronchodilator reversibility and (2) prediction of ICS response in a cohort of treatment naive patients presenting to primary care. Methods: 74 patients with symptoms suggestive of asthma were questioned regarding 6 distinct symptoms before spirometry, PC 20 , and FE NO at baseline, 4 and 12 weeks. ICS were started after the baseline visit. Asthma and ICS response were defined as previously(1). Odds ratios (OR) and sensitivity, specificity and predictive values for each symptom to diagnose asthma and predict ICS response were calculated. Results: 32/74 patients had investigations consistent with asthma and 27/67 responded to 4 weeks of ICS treatment. None of the symptoms alone showed significant diagnostic value for asthma, although cough and wheeze reached borderline significance. Cough predicted response to ICS (OR 10.6 95% CI 3.0-37.3, p Conclusions: In patients presenting with symptoms suggestive of asthma, symptoms are not predictive of objectively defined asthma but cough is sensitive for predicting ICS treatment response. References: 1) Martin MJ, Wilson E, Gerrard-Tarpey W, Meakin G, Hearson G, McKeever TM, Shaw DE and Harrison TW. The utility of exhaled nitric oxide in patients with suspected asthma. Thorax. In press.
