Second Opinion Interpretations by Specialty Radiologists at a Pediatric Hospital: Rate of Disagreement and Clinical Implications

Abstract

The objective of our study was to identify whether a substantive difference exists between the imaging interpretations of radiologists at outside referring institutions and those of radiologists at a tertiary care children's hospital and whether such reinterpretation affects the clinical management of pediatric patients. This retrospective chart review examined the diagnostic imaging reports of all pediatric patients referred to a tertiary care freestanding children's hospital over a 17-month period (January 1, 2009-May 31, 2010); 773 examinations met the inclusion criteria. The original and second interpretations were compared. A fellowship-trained pediatric radiologist and neuroradiologist categorized each case using the content of the two radiology reports as agreement versus minor or major disagreement, and the results were analyzed for statistical significance. A cohort of cases in which a final diagnosis could be confirmed was also analyzed to evaluate the accuracy of both interpretations. Disagreements were found in 323 of 773 reports (41.8%): 168 (21.7%) were major and 155 (20.0%), minor. Neurologic studies were most frequently requested for reinterpretation, 427 (55.2%), most commonly in the setting of trauma, 286 (67.0%). Among the 427 neuroimaging studies, major and minor disagreements occurred in 54 (12.6%) and 91 (21.3%) cases, respectively. Major disagreements most frequently observed were about the presence of fracture and hemorrhage. Among 305 body imaging cases, major and minor disagreements occurred in 99 (32.6%) and 57 (18.7%) cases, respectively. The most common setting for nontraumatic body imaging was concern for appendicitis (168/305 [55.1%]); this indication for imaging was responsible for 40.3% of major disagreements in nontraumatic abdominal imaging. Reinterpretation was rarely requested for radiographic studies (41/773 [5.3%]), which had major and minor disagreement rates of 36.6% and 17.1%, respectively. In the cohort of cases analyzed for final diagnosis, the second interpretation was more accurate than the original in 90.2% of cases with a p value of less than 0.0001. Our findings suggest that discrepancy rates for second interpretations in studies of pediatric patients transferred to tertiary care pediatric institutions are substantial. Although the original and second interpretations in the majority of cases were in agreement, major discrepancies were prevalent--12.6% and 32.6% of neuroimaging and body studies, respectively--and the second interpretations were significantly correlated with the final diagnosis. These results indicate that interpretations by subspecialty radiologists at a point-of-care facility provide important clinical information about the pediatric patient and should be recognized by payers as integral to optimal care.