Goss, Brian C.

Purpose: We previously analyzed the 2006 appropriateness criteria for cardiac computed tomography (CT) and found 42% of our patients were not classifiable, suggesting the criteria were difficult to apply clinically. These criteria were revised in 2010. The purpose of our study is to compare the new 2010 appropriate use criteria to the original 2006 criteria to determine if the 2010 criteria were more applicable. Materials and Methods: In a previous study we applied the 2006 appropriateness criteria to 251 patients that underwent cardiac CT from June 1 to December 31, 2007. Each patient was assigned an indication from the 2006 appropriate use criteria by two observers, classifying patients as appropriate, inappropriate, uncertain, or nonclassifiable. A third observer settled disagreements. This process was applied to the same 251 patients using the 2010 criteria. The data from the previous study using the 2006 criteria was compared to the data obtained using the 2010 criteria to determine changes in the number of nonclassifiable, appropriate, inappropriate, and uncertain exams. These results were then analyzed using Bowker's test of symmetry. Observer agreement was calculated using the kappa statistic. Results: 115 patients (46%) were not classifiable using the 2006 criteria. When the 2010 criteria were applied, the number of not classifiable patients decreased to 40 (16%), with the plurality, 41 (16%), of the not classifiable patients moving to inappropriate. In 2006, 69 (27%) were classified as appropriate, 25 (10%) were uncertain, 42 (17%) inappropriate. In 2010, the appropriate increased to 84 (33%), uncertain increased to 40 (16%), and inappropriate increased to 87 (35%). Bowker's test of symmetry demonstrated a statistically significant change with a p<0.001. Agreement between the two observers assigning patients to classification groups also improved with a kappa of 0.55 using the 2010 criteria compared to kappa of 0.31 in 2006. Conclusion: The number of not classifiable patients decreased using the 2010 criteria and the interobserver agreement improved, suggesting they are more applicable and complete than the 2006 criteria. This decrease in not classifiable patients came primarily from an increase in the number of inappropriate indications.

Speech recognition software can increase the frequency of errors in radiology reports, which may affect patient care. We retrieved 213,977 speech recognition software–generated reports from 147 different radiologists and proofread them for errors. Errors were classified as “material” if they were believed to alter interpretation of the report. “Immaterial” errors were subclassified as intrusion/omission or spelling errors. The proportion of errors and error type were compared among individual radiologists, imaging subspecialty, and time periods. In all, 20,759 reports (9.7%) contained errors, of which 3992 (1.9%) were material errors. Among immaterial errors, spelling errors were more common than intrusion/omission errors ( p < .001). Proportion of errors and fraction of material errors varied significantly among radiologists and between imaging subspecialties ( p < .001). Errors were more common in cross-sectional reports, reports reinterpreting results of outside examinations, and procedural studies (all p < .001). Error rate decreased over time ( p < .001), which suggests that a quality control program with regular feedback may reduce errors.