Quantifying Value in an Iterative Disease-Site Specific Peer Review System

Abstract

To assess the greatest impact of peer review on treatment planning and plan modification in two disease-site specific peer review groups to improve quality assurance workflow. We implemented an iterative peer review checklist tool and treatment planning registry to document recommendations and time spent in breast and thoracic disease-site specific peer review. Plan review included discussion of the clinical scenario, radiation dose/fractionation, prescription, organs at risk and target volumes, planning metrics including dose volume histogram and patient setup. Pre-treatment (Pre-tx) review of volumes and plans occurred prior to treatment start for definitive plans of ≤five fractions or using protons, and cases were re-reviewed in subsequent sessions to complete checklist items as needed. Reviews were graded as follows: discussion (no required changes), minor recommendations (changes recommended at attending discretion) and major recommendations (changes required). Students t-test and chi-squared analysis were used to identify statistical significance. We identified 723 peer reviews over a 3-month period: 275 thoracic reviews of 195 unique patients (32% reviewed ≥twice), and 448 breast reviews of 273 unique patients (48% reviewed ≥twice, most commonly for a boost plan). On average, pre-tx review doubled the amount of time spent per case, (4:47min versus 2:03min, p<0.01) and resulted in significantly more discussion (29.3% vs 9.3%, p<0.01), minor changes recommended (13.2% versus 4.8%, p<0.01) and plan revision (19.2% versus 2.4%, p<0.01). Thoracic pre-tx reviews were twice as long as breast (5:43min versus 3:00min, p<0.01), and had a greater proportion of cases prompting discussion (26.7% vs 13.2%, p<0.01) and minor changes recommended (12.8% vs 6.3%, p<0.01). No time difference was found in regular review (2:19min vs 1:58 min, p = 0.09). The most frequent discussion/recommendation topics also varied by group with planning metrics in thoracic cases (14.9%) and setup imaging including PORT films and electron setup (5.1%) in breast (Table). Iterative plan review with pre-review for selected cases encouraged educational discussion and plan revision, with notable variability by disease site. Thoracic cases required longer review time and had more frequent plan change recommendations likely reflecting differences in case complexity. In contrast, setup imaging in breast cases, which are more commonly 3D or electron-based, more frequently warranted discussion/revision and should be considered within the peer review paradigm.