Ventilation/perfusion lung scan probability category distributions in university and community hospitals

The criteria used in the Prospective Investigation of Pulmonary Embolism Diagnosis (PIOPED) study for the interpretation of ventilation/perfusion scans are widely used and the probability of pulmonary embolism is determined from these criteria. The prevalence of pulmonary embolism in the PIOPED study was 33%. To investigate the similarity of patient populations who have ventilation/perfusion scans at one of the medical centers that participated in the PIOPED study and a small community hospital, the authors evaluated the probability category distributions of lung scans at the two institutions. They retrospectively interpreted 54 and 49 ventilation/perfusion lung scans selected from January, 1991, to June, 1992, at Duke University Medical Center and at Central Carolina Hospital, respectively. Studies were interpreted according to the PIOPED criteria. The percentage of studies assigned to each category at Duke University Medical Center and Central Carolina Hospital were 17% and 27% normal or very low probability, 31% and 59% low probability, 39% and 10% intermediate probability, and 13% and 4% high probability, respectively. The different distribution of probability categories between university and community hospitals suggests that the prevalence of disease may also be different. The post-test probability of pulmonary embolism is related to the prevalence of disease and the sensitivity and specificity of the ventilation/perfusion scan. Because these variables may differ in community hospital settings, the post-test probability of pulmonary embolism as determined by data from the PIOPED study should only be used in institutions with similar populations. Clinical management based upon the results of the PIOPED study may not be applicable to patients who have ventilation/perfusion scans performed in a community hospital.