Purpose or Case Report: RADPEER has been criticized for lack of clinical implications, yet few studies have compared RADPEER to alternative peer-review systems or correlated peer review scores with changes in patient management. We aimed to implement a pediatric subspecialty over-read program and compare RADPEER to a Clinical Outcomes Scale, which graded impact on patient outcomes.
Methods & Materials: Consecutive neuroradiology CTs dictated after 5pm and on weekends by general pediatric radiologists were selected and were over-read by pediatric neuroradiologists the following day. Each case was assigned a RADPEER score (1=Concur, 2=discrepancy/not ordinarily expected to be made, 3=discrepancy/should be made most of the time, 4=discrepancy/should be made almost every time). Outcomes were determined by chart review and patient course over the subsequent 24-hour period, and assigned a Clinical Outcomes Scale score (1=No error, 2=incidental error without clinical detriment, 3=error requiring further investigation or observation, 4=error requiring medical or surgical intervention).
Results: A total of 508 cases were reviewed. Applying RADPEER resulted in 469 (92.3%) level 1, 24 (4.7%) level 2, 7 (1.4%) level 3, and 8 (1.6%) level 4 scores. Using the Clinical Outcomes Scale, we found 471 (92.7%) level 1, 25 (4.9%) level 2, 4 (0.8%) level 3, and 8 (1.6%) level 4 cases. Missed findings were responsible for the majority of level 3 and 4 cases for both scoring systems. 6/8 level 4 cases identified by the Outcomes Scale were also labeled level 4 by RADPEER and one additional case was labeled level 3 instead of level 4 by RADPEER. 3/4 level 3 cases identified by the Outcomes Scale were also labeled as level 3 by RADPEER. One level 4 and one level 3 discrepancy based on patient outcome was identified as level 2 by RADPEER.
Conclusions: While not as specific, RADPEER demonstrated similar sensitivity as the Clinical Outcomes Scale in identifying discrepancies that lead to changes in clinical management. The optional RADPEER likelihood of clinical significance sub-score is not widely used but may have increased the specificity of RADPEER with regards to impact on patient care. Our outcomes data validates the use of the RADPEER scoring system and suggests direct clinical implication of the score for pediatric neuroradiology CT studies.