Prospective Blinded Study of The Validity of Radiation Oncology Chart Rounds as an Effective Quality Assurance Process

Abstract

Radiation oncology is one of few fields in medicine to incorporate a formal peer-review process, though no studies exist that validate the effectiveness of weekly physician chart rounds (C-R) as a quality assurance measure. We conducted a prospective blinded study to evaluate and quantify the efficacy of physician peer review during C-R as a method of detecting treatment planning errors. Sham plans were generated for select cases incorporating deliberate errors that were anticipated to be identifiable at weekly peer-reviewed C-R. An error rate of 7.3% (12/164 total cases) was introduced randomly into C-R over four weeks at an NCI designated Comprehensive Cancer Center. Sham plans were shown by the chief therapist in lieu of actual plans during clinical case presentations. All physicians were unaware of the study except the first and senior authors, chief physicist, and chief therapist. Actual plans chosen for sham replacement underwent peer review using a specialized QA process to ensure clinical quality and correct plan delivery. An error was scored as detected if a physician identified the error incorporated into the sham plan at C-R. Initial study design was to have 60 cases with an anticipated error-detection rate of 70%. The study was stopped early after the first 25% of sham case presentations revealed an unexpectedly low detection rate. Sham cases included various disease sites, treatment intent, and techniques, while incorporating errors of target verification (laterality, spine level, intact versus postop) and dose. 16.7% (2/12) of the presented sham cases were detected. The 2 detected errors included a right lung SBRT case where a plan treating the contralateral lung was shown and a brain SRS plan depicting a lesion in the contralateral side of the brain. Physician attendance did not correlate with error detection rates as 7 physicians were present when both errors were detected; between 4 and 8 were present during the remaining sham presentations. The treating physician was present during 75% (9/12) of sham presentations. In both detections, the treating physician identified the error, not their peers. This is the first prospective study attempting to quantify and validate radiation oncology C-R as an effective component of the quality assurance process. In this study, the error detection rate was substantially lower than anticipated. All errors were identified by the treating physician rather than by their peers. Subsequent study to evaluate the impact of a planned intervention to improve the integrity of C-R as a peer review process is underway.Abstract 2445; TableDetails of Sham CasesDisease SiteIncorporated ErrorDetection RateBreastLaterality0/4ProstateIntact vs Postop0/4Lung SBRTLaterality1/2SpineVertebral level0/1Brain SRSLaterality1/1 Open table in a new tab