Impact of spirometry on GPs' diagnostic differentiation and decision-making

Background: Spirometry is increasingly implemented in general practice, while the ability of general practitioners (GPs) to interpret flow-volume curves (F-V curves) has been questioned. Furthermore, the role of spirometry in the GPs decision-making process has barely been studied. Aim: To compare the achievements of trained GPs in spirometric diagnosis with an expert consensus panel (1) and to assess the influence of spirometry on the GPs decision-making (2). Method: Twelve cases including a wide range of F-V curves were interpreted by 39 GPs as well as the expert panel. Diagnostic test characteristics were calculated using multi-level analysis and summarised by diagnostic odds ratios (DOR). Differences in decision-making indicators were expressed as odds ratios and 95% confidence intervals. Results: Normal F-V curves (DOR 65.0) and obstructive F-V curves (DOR 48.9) were reasonably well diagnosed, white rare and mixed pathological patterns achieved considerably lower scores (DOR 3.8). Intermediate scores were obtained in the recognition of incorrect test manoeuvres (DOR 24.4). Spirometry influenced the GPs decision-making in reducing the number of alternative diagnoses (OR 0.266 [0.200, 0.353]), but also increased referral rates (7.26 [4.71, 11.2]) and the use of diagnostic prednisolone courses (4.55 [3.12, 6.64]) substantially. Conclusion: Trained GPs were able to differentiate between normal and obstructive disease patterns, while F-V curves suggestive of rare and mixed pathology were often missed. Spirometry seems to influence the decision-making process of the GP; whether this represents an initial or a more sustained effect remains to be evaluated in studies of daily primary care practice. (C) 2004 Elsevier Ltd. All rights reserved.