Background: Evidence-based guidelines recommend serial measurements of peak expiratory flow (PEF) on days at and away from work as the first step in the objective confirmation of occupational asthma. The aim of this study was to improve the diagnostic value of computer-based PEF analysis by using the program Oasys-2 to calculate a score from the area between the curves (ABC) of PEF on days at and away from work.
Methods: Mean 2-hourly PEFs were plotted separately for workdays and rest days for 109 workers with occupational asthma and 117 control asthmatics. A score based on the ABC was computed from records containing >or= 4 day shifts, >or= 4 rest days, and >or= 6 readings per day. Patients were randomly classified into two data sets (analysis and test sets). Receiver operator characteristic (ROC) curve analysis determined a cutoff point from set 1 that best identified those with occupational asthma, which was then tested in set 2.
Results: Logistic regression analysis showed that all ABC PEF scores were significant predictors of occupational asthma, with the best being ABC per hour from waking (odds ratio, 11.9 per 10 L/h/min; 95% confidence interval, 10.8 to 13.1). ROC curve analysis showed that a difference of 15 L/min/h provided a high specificity without compromising sensitivity in diagnosing occupational asthma. Analysis of data set 2 confirmed a specificity of 100% and sensitivity of 72%.
Conclusion: The ABC PEF score is sensitive and specific for the diagnosis of occupational asthma and can be calculated from a shorter PEF surveillance than is needed for the current Oasys-2 work effect index.