Evaluation of internists' spirometric interpretations.

BACKGROUND: Correct interpretation of screening spirometry results is essential in making accurate clinical diagnoses and directing subsequent pulmonary evaluation. The general internist is largely responsible for interpreting screening spirometric tests at community hospitals. However, reports of new guidelines for screening spirometry are infrequently published in the general internal medicine literature. This can lead to incorrect interpretations. We sought to evaluate whether spirometric interpretations by a group of practicing general internists differed from those of two board-certified pulmonologists using guidelines published by the American Thoracic Society (ATS).
METHODS: As part of a Continuous Quality Improvement project, all available screening spirometric tests over a 3-month period at two area community hospitals were reviewed. Only those performed on individuals age 18 or older were included in the analysis. Comparison was made between the interpretations of staff internists and those of two pulmonologists, who were blinded to the results of all other interpretations. We analyzed 110 screening spirometric tests from 84 males and 26 females. The patients ranged in age from 18 to 77 (mean 41 +/- 13 years of age).
RESULTS: There was 97% concordance between the two pulmonologists' interpretations. In three cases, interpretations of only one pulmonologist agreed with those of the internists. The internists and both pulmonologists agreed in 73 cases. The majority of spirometric results in this subgroup were normal (n = 54). Both pulmonologists disagreed with internists' nomenclature in five cases. There was complete disagreement between the pulmonologists and the internists in the other 29 cases. Using the pulmonologists' interpretations as the "gold standard," the sensitivity (the internists' ability to correctly identify abnormal spirometric results) was 58.8% (95% confidence interval [CI] 42.2%, 73.3%), the specificity was 81.8% (95% CI 70.0%, 89.8%), the positive predictive value was 66.7% (95% CI 49.0%, 80.9%), and the negative predictive value was 76.1% (95% CI 64.3%, 85.0%). The most common inaccurate interpretations made by internists were "small airways disease" when spirometric results were normal (n = 8); "normal" when a restrictive pattern was present (n = 6), and "normal" when an abnormal flow-volume loop suggesting possible upper airway obstruction was present (n = 5).
CONCLUSIONS: The spirometric interpretations of a group of general internists differed significantly from those of two board-certified pulmonologists using published guidelines in approximately one third of cases. This may be because subspecialty guidelines are infrequently published in the general internal medicine literature. We believe that wider dissemination of these interpretative guidelines and ongoing physician education would improve general internists' ability to identify patients who require further pulmonary evaluation.