Prospective Peer Review Research Articles

SBRT for Pancreatic Cancer: A Radiosurgery Society Case-Based Practical Guidelines to Challenging Cases

The use of radiation therapy (RT) for pancreatic cancer continues to be controversial, despite recent technical advances. Improvements in systemic control have created an evolving role for RT and the need for improved local tumor control, but currently, no standardized approach exists. Advances in stereotactic body RT, motion management, real-time image guidance, and adaptive therapy have renewed hopes of improved outcomes in this devastating disease with one of the lowest survival rates. This case-based guide provides a practical framework for delivering stereotactic body RT for locally advanced pancreatic cancer. In conjunction with multidisciplinary care, an intradisciplinary approach should guide treatment of the high-risk cases outlined within these guidelines for prospective peer review and treatment safety discussions.

Factors Associated with Stereotactic Radiation Plan Revision in Prospective Peer Review.

Stereotactic radiotherapy (SR) is highly effective but has risks for serious toxicity. We sought to identify risk factors for SR plans requiring revision in a comprehensive, prospective peer review program conducted across a network of affiliated radiation centers. We hypothesized that increased physician experience and SR case volume would be associated with lower rates of SR plan revision. Weekly peer review rounds were conducted to review SR cases for image fusion, contours, plan, and dose constraints, all prior to start of radiation. Cases recommended for revision were recorded and tracked prospectively. Factors potentially associated with case revision including body site, SR type, physician experience, and physician case load were assessed for significance using univariate and multivariable logistic regression. From March 2019 to January 2023, 1,015 SR cases were prospectively reviewed, including 312 brain stereotactic radiosurgery (SRS), 190 multi-fraction brain SRS (fSRS), and 513 stereotactic body radiotherapy (SBRT). Revision was recommended in 177 cases (17%). The yearly revision rate was 21% in 2019, 16% in 2020, 17% in 2021, and 18% in 2022. There were 13 individual treating physicians with a median of 5 years' experience (range: 2-18 years), measured at the time of each SR case review. Physicians were categorized as junior (< = 2 years of experience), mid-career (3-9 years), or senior (> = 10 years). The physician's SR case volume in the preceding 3 months (median 25 cases) was dichotomized as low volume (< = 25) or high volume (>25). Logistic regression results are shown in Table 1. Statistical significance was determined by p <0.05. SR plans utilizing SRS (as opposed to SBRT) and treated by physicians with high case volume had lower revision rates. Junior attendings had higher revision rates. These data imply a high value to peer review for junior attendings and for those with low SR case volume. Adequate case volume may be a critical factor for safe delivery of SR, analogous to surgical literature. Annual revision rates in the program remained stable over time, demonstrating the ongoing importance of an effective peer review program for SR.

Leveraging a Real-Time Machine Learning Platform to Personalize Quality Assurance at a Large Multi-Site Oncology Practice

Peer review is a critical component of safe and effective oncology care and provides care teams with an opportunity to improve patient outcomes with predictive analytics. We describe a paradigm for incorporating predictive analytics into daily prospective peer review at a large multisite oncology center. Previously, we deployed a department-wide proprietary oncology analytics platform connected to the Electronic Medical Record System (outpatient and inpatient), Oncology Information System, Tumor Registry, Treatment Planning System and Radiology PACS. Variables extracted and calculated by the analytics platform on a nightly basis include patient demographics, past medical history, detailed pathology data, recent laboratory data, vitals, prior systemic cancer therapies, prior surgical treatments, prior radiation treatments, recent hospitalization, vital status, and current treatment details. Using this platform, we trained machine learning models for critical patient events, such as unplanned hospitalization, mortality, occurrence of grade 3 or higher toxicities, and treatment interruption. Importantly, these models leverage so-called interpretable machine learning techniques, allowing us to interrogate the models and obtain explanations for why a patient is predicted to be at higher or lower risk of an event. All such models were previously evaluated for performance on standalone testing sets. In this abstract, we propose a paradigm to include predictive analytics in our daily prospective peer review process by creating a list of patients at high risk of critical events (hospitalization, mortality, treatment interruption, and serious adverse events) and reviewing those lists in conjunction with our daily peer review conference. We also propose a process to review any gaps in clinical care (e.g., abnormal labs, decreases in patient weight, or missed progress visits) to recommend suitable interventions to the care team to reduce patient risk. A total of 64,286 radiation treatment cases were included in our analytics platform and used to train and validate our predictive models. Moving forward, we aim to peer review 20 patient charts per day using this paradigm. While numerous studies in the field of radiation oncology have focused on the benefits of machine learning when applied to contouring and personalized radiation dose, few have focused on how machine learning can improve supportive care. In this abstract, we propose a framework to introduce machine learning into quality assurance, with the goal of improving patient outcomes.

Analysis of Retrospective Versus Prospective Peer Review in a Multisite Academic Radiation Department

Our multisite academic radiation department reviewed our experience with transitioning from weekly primarily retrospective to daily primarily prospective peer review to improve plan quality and decrease the rate of plan revisions after treatment start. This study was an institutional review board-approved prospective comparison of radiation treatment plan review outcomes of plans reviewed weekly (majority within 1 week after treatment start) versus plans reviewed daily (majority before treatment start, except brachytherapy, frame-based radiosurgery, and some emergent plans). Deviations were based on peer comments and considered major if plan revisions were recommended before the next fraction and minor if modifications were suggested but not required. Categorical variables were compared using χ2 distribution tests of independence; means were compared using independent t tests. In all, 798 patients with 1124 plans were reviewed: 611 plans weekly and 513 plans daily. Overall, 76 deviations (6.8%) were noted. Rates of any deviation were increased in the daily era (8.6% vs 5.2%; P = .026), with higher rates of major deviations in the daily era (4.1% vs 1.6%; P = .012). Median working days between initial simulation and treatment was the same across eras (8 days). Deviations led to a plan revision at a higher rate in the daily era (84.1% vs 31.3%; P < .001). Daily prospective peer review is feasible in a multisite academic setting. Daily peer review with emphasis on prospective plan evaluation increased constructive plan feedback, plan revisions, and plan revisions being implemented before treatment start.

A Decade of Prospective Peer Review: Impact on Safety Culture and Lessons Learned in a Multicenter Radiation Medicine Department

Quality assurance (QA) is critical to the success of radiation therapy (RT) for patients with cancer and affects clinical outcomes. We report longitudinal findings of a prospective peer review evaluation system implemented at a major academic health system as part of RT QA during a 10-year period. All cases treated within our department undergo prospective multidisciplinary peer review and are assigned a grade (A, B, or C). "A" cases require no changes, "B" cases require minor modification, and "C" cases require major modification before treatment planning. The z-ratio test for the significance of the difference between the 5-year baseline (2012-2016) and follow-up (2017-2021) period was used to compare grades between the 2 periods. A 2-tailed P value <.05 was considered significant. Of the 20,069 cases, 15,659 (78%) were curative and were analyzed. The fraction of A cases decreased from 74.8% (baseline) to 64.5% (follow-up), whereas B cases increased from 19.4% to 35.4% and C cases decreased from 5.8% to 0.1%. Of the 9 treatment locations, the main hospital site had a higher percentage of A grades relative to community locations in the baseline (78.6% vs 67.8%; P < .002) and follow-up (66.9% vs 62.3%; P < .002) periods. There was a decrease in the percentage of A cases from the baseline to the follow-up period regardless of plan type (complex vs intermediate vs simple). There was a decrease in the percentage of A cases among specialists from baseline to follow-up (78.2% to 67.7%; P < .002) and among generalists from baseline to follow-up (69.7% to 61.7%; P < .002). Our 10-year experience in contour peer review identifiedincreasedopportunities in improving treatment plan quality over time. The drop in A scores and rise in B scores suggests increased scrutiny and findings-based improvements over time, whereas the drop in C scores indicates amelioration of "major failures" addressed in the startup years. Peer review rounds upstream of treatment planning provide valuable RT QA and should be considered by other departments to enhance the quality and consistency of RT.

The impact of neuroradiology collaboration in head and neck cancer radiotherapy peer review.

To quantify the impact of neuroradiologist presence on head and neck cancer (HNC) radiotherapy peer review (PR) changes. Prospective data were collected from HNC radiotherapy PR meetings; major, minor, and organ at risk (OAR) changes recorded. Differences in changes made with a neuroradiologist present were determined. χ2 tests of statistical significance were performed. Multivariate logistic regression identified potential predictors of changes. Prospective PR was performed in 125/160 (78%) patients undergoing radical (chemo)radiotherapy for HNC between October 2018 and September 2019. Full PR documentation was available for 120/160 meetings (75%), with a neuroradiologist present in 53/120 (44%). Overall, 51/120 (42.5%) had changes made to target volumes or OARs. When a neuroradiologist was present, 29/53 (55%) of plans had changes made, compared to 22/67 (33%) in their absence. On multivariate analysis, neuroradiologist presence significantly influenced any changes made during the PR meetings (OR 2.59; 95% CI 1.05-6.43; p = 0.039). Neuroradiologist presence at PR meetings significantly influences changes made to HNC contouring, likely improving consistency and enhancing quality assurance. This is the first published UK series demonstrating that a collaborative approach between radiology and oncology in PR meetings is significant in leading to contour changes for HNC.

Peer Review of Head and Neck Cancer Planning Target Volumes in Radiation Oncology

<h3>Purpose/Objective(s)</h3> Radiation treatment plans undergo peer review during chart rounds, but changes to treatment volumes would require replanning. Our group implemented weekly head and neck cancer "volume rounds" to peer review all target volumes for head and neck cancer prior to radiation therapy (RT) planning and chart rounds. <h3>Materials/Methods</h3> We analyzed modifications made to planning target volumes (PTVs) at volume rounds for consecutive non-proton head and neck cancer cases from May 2020 to May 2021. Nine head and neck radiation oncologists participated in weekly volume rounds during this time. Recommendations were categorized as no changes, minor changes, major changes, additional workup (e.g., biopsy or imaging), and consultation or tumor board discussion needed prior to the start of RT. Minor changes to PTVs generally did not require a second review prior to treatment planning while major changes did. <h3>Results</h3> PTVs for 511 cases involving 432 patients underwent peer review and 298 (58.3%) of these cases did not require any modifications prior to treatment planning. Minor and major changes were recommended in 75 (14.7%) and 86 (16.8%) cases, respectively. Additionally, 45 (8.8%) cases were recommended to have further workup and 23 (4.5%) required further consultation with other non-radiation oncology services. Of the 45 cases that were recommended for further workup, 40 underwent biopsy or imaging. Positive findings on imaging or biopsy occurred in 13 patients leading to a significant change in management, including 4 patients that underwent further surgery after positive findings prior to the start of RT. <h3>Conclusion</h3> Prospective peer review during head and neck cancer volume rounds led to frequent minor and major alterations to PTVs. Significant changes in the overall treatment plan, such as additional surgery prior to start of RT, occurred in a minority of patients.

Prospective Peer Review for Hypo-Fractionated Treatments During the COVID-19 Pandemic

Purpose/Objective(s)One of the measures taken in our department to minimize the risk of infection during the advent of the COVID-19 pandemic was to reduce the number of treatment fractions for appropriate disease sites. In this work we sought to assess peer-review grades for such hypofractionated cases during daily pre-dosimetric treatment planning multidisciplinary reviews as part of our ongoing quality control program.Materials/MethodsIn our multi-center radiation medicine department, prospective video-conferenced reviews are conducted daily on all cases requiring a new treatment plan by a multidisciplinary team using an in-house electronic whiteboard linked with our oncology-information systems. Each case is allocated a grade of A, B or C on the treatment directive and contours to respectively proceed with, suggest minor revisions, or to require major revisions prior to dosimetric treatment planning. The database was queried for the years 2017 through 2021 to assess trends of grades assigned for hypofractionated plans for comparison with those for conventional and SBRT plans for the same anatomic sites, for both pre-COVID (2017-2019) and COVID (2020-2021) years.ResultsOf the 7179 plans generated during this period for 11 anatomic sites, 980 were hypofractionated, 4823 were conventionally fractionated, and 1376 were SBRT plans with overall grade breakdown as follows: A: 69.9%, B: 30.0%, C: 0.1%. The fraction of B grades for conventional, hypofractionated and SBRT cases respectively were: Bladder (51.4%, 34.3%, -), Bone (10.5%, 13.0%, 35.7%), Brain (29.3%, 28.1%, 30.2%), Breast (20.9%, 19.8%, 100.0%), Endometrium (45.6%, 0.0%, -), Lung (30.1%, 26.7%, 34.3%), Pancreas (44.0%, 16.7%, 41.6%), Pelvis (29.0%, 100.0%, -), Prostate (33.5%, 24.5%, 36.4%), Rectum (40.9%, 21.9%, -), Skin (34.9%, 33.3%, -). For most sites, hypofractionated cases had lower fractions of B-grades and higher fractions of A grades compared with conventional or SBRT cases. The differences overall were not statistically significant (Mann-Whitney non-parametric test, p-value>0.05). Differences in grade fractions were minor in COVID versus non-COVID years. The only statistically significant differences noted in time trends were between hypofractionated cases versus SBRT cases for the bone sites (p-value 0.04).ConclusionReducing the fractionation regimens for appropriate anatomic sites during the COVID-19 pandemic yielded no major differences in prospective peer-review assessments compared with conventional or SBRT assessments. Peer review enabled continued resilience during the pandemic ensuring continuity of the quality of care.

Improving Radiation Quality Through the Implementation of Prospective Peer Review

<h3>Purpose/Objective(s)</h3> Quality assurance (QA) is critical to the successful delivery of radiation therapy (RT) for cancer patients and also impacts outcomes. The present study reports the longitudinal results of a prospective peer review evaluation system (PES) implemented at a major academic health system as an integral part of RT QA. <h3>Materials/Methods</h3> All patient cases treated with RT by one of 32 physicians at 14 treatment locations within our radiation oncology department undergo prospective peer review in daily rounds. Each case is assigned a grade (A, B, C) based on the following criteria: clinical suitability for RT, prescription, contours, and treatment plan, where applicable. "A" cases require no changes, "B's" require minor modification, and "C's" require major modification with re-review prior to treatment planning. The present study is a longitudinal analysis of grade trends over time by hospital, plan type, and physician. The Wilcoxon-signed-ranked test was used to compare grades during the baseline PES period (2013-2016) to the follow-up period (2017-2021). A p-value below 0.05 is considered significant. <h3>Results</h3> A total of 20,069 cases for 14,300 patients were reviewed from September 2013 to November 2021. Of the total cases, 15,659 (78%) were curative and were included in the present analysis; 5,868 from the baseline period and 9,791 in the follow-up period. The fraction of A cases decreased from 74.8% (baseline) to 64.5% (follow-up), while the fraction of B's increased from 19.4% (baseline) to 35.4% (follow-up), and the fraction of C's decreased from 5.8% (baseline) to 0.1% (follow-up). The main hospital site contributed 54.8% of all cases and 48.1% of all B grades. IMRT plans accounted for 55.4% of all plans and 68.2% of all B scores, while 3D-conformal plans accounted for 24.4% of cases and 15.4% of all B scores. When normalized by total department A/B/C scores, there were no statistically significant changes in percent A/B/C scores over the time period studied (p-value> 0.05). However, for the 13 physicians with cases throughout the study period, there was a statistically significant increase in the fraction of B scores (p value = 0.01) and decrease in the fraction of C scores (p value < 0.01) between the two time periods when individual physician grades were normalized to their individual caseload. <h3>Conclusion</h3> PES improves the identification of quality-related factors that can impact downstream radiation treatment plan quality. The diminishing fraction of C grades over time indicates amelioration of ‘major failures,' which were addressed in the startup years of PES. The proportional increase in B's, at the cost of fewer putative A's, suggests increased scrutiny of cases, as well as findings-based improvements in contouring and planning, over time. Peer review rounds upstream of treatment planning provides valuable RT QA and should be considered by other departments to enhance the quality and consistency of RT.

Cardiac stereotactic ablative radiotherapy for control of refractory ventricular tachycardia: initial UK multicentre experience

BackgroundOptions for patients with ventricular tachycardia (VT) refractory to antiarrhythmic drugs and/or catheter ablation remain limited. Stereotactic radiotherapy has been described as a novel treatment option.MethodsSeven patients with recurrent refractory...

Evaluation of a Prospective Radiation Oncology Departmental Peer Review Process using Standardized Simulation Directives

<h3>Purpose/Objective(s)</h3> The primary objective of this study is to evaluate the utility and value of an institutional, multi-disciplinary peer review process prior to radiotherapy simulation. <h3>Materials/Methods</h3> Over a period of 3 months and through an iterative process, a standardized simulation requisition directive (SSRD) was developed, piloted, modified, and subsequently implemented for all patients treated with external beam radiotherapy at a single tertiary care institution from January to December 2020. The SSRDs were reviewed at a daily multi-disciplinary peer review conference ("morning huddle"); modifications in the simulation process consequential to the review were prospectively recorded in a quality database. Descriptive statistics were used to identify characteristics associated with modifications. Independent sample t test and Chi-square test were used to compare continuous and categorical variables. <h3>Results</h3> A total of 1500 consecutive SSRDs were prospectively reviewed and met inclusion criteria for this study. The median time between the physician SSRD order and date of simulation was 4 days (Range: 0-97 days), and the majority of patients were treated with curative intent (n = 872, 58%). A total of 369 modifications on 269 (17.9%) simulation directives were recorded and parsed into 17 categories. The most common modifications resulted from omission of pregnancy testing orders (n = 92, 24.9%), immobilization device changes (n = 88, 23.8%), changes in the radiotherapy care path (n = 56, 15.1%), and arm positioning (n = 43, 11.6%). Modifications were less likely to occur if the directives were entered within 1 week of simulation (15.6% vs. 21.7%, <i>P</i> = 0.0028). Significant differences were also observed across tumor sites (<i>P</i> = 0.0091) with the highest modification rates observed for stomach, esophagus, and pelvis sites (40%, 30%, and 26.9% modified, respectively). A significant change in department workflow and clinic visits occurred in March 2020 as a result of COVID-19, with transition to virtual platforms. An increased rate of simulation directive modifications was also observed for patients simulated after these changes were implemented (April – December 2020 19.3% vs. Jan – March 2020 13.5%, <i>P</i> = 0.013). No differences in modification rates were observed by modality, i.e., photon or proton therapy (<i>P</i> = 0.20). Overall, with this prospective peer review process, only 14 patients (0.9%) needed re-simulation during the entire study period. <h3>Conclusion</h3> Prospective peer review prior to simulation in radiotherapy identifies actionable change in approximately 18% of procedures, and results in an extremely low, < 1% rate of re-simulation. SSRDs ordered > 1 week before from simulation and gastrointestinal and pelvic sites were at higher risk of requiring modifications during peer review. As departmental processes transition to virtual meeting platforms, more thorough attention is needed to identify patients at higher risk of simulation modifications.

A prospective multidisciplinary review of radiotherapy processes during the telemedicine era.

282 Background: The COVID-19 pandemic resulted in the use of telemedicine for evaluation and management visits in radiation oncology departments to decrease in-person interactions. The primary objective of this study is to evaluate the utility of telemedicine for patient consultation and its effect on radiotherapy simulation and treatment processes. Methods: A standardized simulation requisition directive was used for all consults (telemedicine and in-person) undergoing simulation for external beam radiation therapy at a single tertiary care institution from January to December 2020. These directives were reviewed at daily multi-disciplinary peer review meetings; modifications occurring as a result of this review were prospectively recorded in a departmental quality database. Descriptive statistics were used to identify characteristics associated with consultation type. Mann Whitney and Chi-square tests were used to compare continuous and categorical variables. Results: 1500 consecutive patients were reviewed in this prospective peer review process; 444 (29.6%) had telemedicine visits preceding simulation and 1056 (70.4%) had in-person consults. The median time between physician simulation order and date of simulation was 5 days (IQR: 2-11 days) for telemedicine visits and 4 days (IQR: 1-8 days) for in-person consults ( p&lt;0.05). Significant differences were observed in telemedicine usage across months ( p&lt;0.05) with the highest percentage in July, September, and August (50%, 45.9% and 45%, respectively). As a result of prospective multidisciplinary peer review, 397 modifications in 290 simulations were recorded in total; 101/444 (22.7%) telemedicine simulations had modifications compared to 189/1056 (17.9%) following in-person consultation ( p&lt;0.05). The most common modifications for telemedicine visits resulted from immobilization device changes (n=32, 23.5%), arm positioning (n=19, 14.0%), and changes in the radiotherapy care path (n=17, 12.5%). For telemedicine consults, the median radiotherapy fraction dose was 2.66 Gy (2-4 Gy) and median fraction number was 16 (5-28). There was no difference for fractionation preference between telemedicine and in person consults ( p=0.084). Seven (1.6%) telemedicine visits and 7 (0.7%) in-person consults needed re-simulation during the entire study period ( p=0.136). Conclusions: Telemedicine is a powerful tool with the potential to revolutionize the radiation oncology daily practice. In the initial learning phase, it appears that there is a higher frequency of simulation modifications for patients evaluated by telemedicine. Therefore, as departmental processes incorporate telemedicine in the future, thorough attention is needed to encourage review of common modifications as well as identify patients at high risk of error at time of simulation who may also benefit from in-person evaluation prior to simulation.

Analysis of Virtual Versus In-Person Prospective Peer Review Workflow in a Multisite Academic Radiation Oncology Department

PurposeIn radiation oncology, peer review is a process where subjective treatment planning decisions are assessed by those independent of the prescribing physician. Before March 2020, all peer review sessions occurred in person; however due to the COVID-19 pandemic, the peer-review workflow was transitioned from in-person to virtual. We sought to assess any differences between virtual versus in-person prospective peer review. Methods and MaterialsPatients scheduled to receive nonemergent nonprocedural radiation therapy (RT) were presented daily at prospective peer-review before the start of RT administration. Planning software was used, with critical evaluation of several variables including treatment intent, contour definition, treatment target coverage, and risk to critical structures. A deviation was defined as any suggested plan revision. ResultsIn the study, 274 treatment plans evaluated in-person in 2017 to 2018 were compared with 195 plans evaluated virtually in 2021. There were significant differences in palliative intent (36% vs 22%; P = .002), but not in total time between simulation and the start of treatment (9.2 vs 10.0 days; P = .10). Overall deviations (8.0% in-person vs 2.6% virtual; P = .015) were significantly reduced in virtual peer review. ConclusionsProspective daily peer review of radiation oncology treatment plans can be performed virtually with similar timeliness of patient care compared with in-person peer review. A decrease in deviation rate in the virtual peer review setting will need to be further investigated to determine whether virtual workflow can be considered a standard of care.

Prospective Peer Review for Patients Undergoing Stereotactic Body Radiation Therapy and Intracranial Radiosurgery in a Multi-Site Community-Setting is Both Feasible and Effective

Prospective Peer Review for Patients Undergoing Stereotactic Body Radiation Therapy and Intracranial Radiosurgery in a Multi-Site Community-Setting is Both Feasible and Effective

Development of a Comprehensive, Contour-Based, Peer Review Workflow at a Community Proton Center

PurposeQuality assurance and continuing quality improvement are integral parts of any radiation oncology practice. With increasingly conformal radiation treatments, it has become critical to focus on every slice of the target contour to ensure adequate tumor coverage and optimal normal tissue sparing. Proton therapy centers open internationally with increasing frequency, and radiation oncologists with varying degrees of subspecialization apply proton therapy in daily practice. Precise treatment with proton therapy allows us to limit toxicity but requires in-depth knowledge of the unique properties of proton beam delivery. To address this need at our proton therapy center, we developed a comprehensive peer review program to help improve the quality of care that we were providing for our patients.Materials and MethodsWe implemented a policy of comprehensive peer review for all patients treated at our community proton facility starting in January 2013. Peer review begins at the time of referral with prospective cases being reviewed for appropriateness for proton therapy at daily rounds. There is then biweekly review of target contouring and treatment plans.ResultsDuring a 6-month period from June 2013 to November 2013, a total of 223 new patients were treated. Documentation of peer review at chart rounds was completed for 222 of the 223 patients (99.6%). An average of 10.7 cases were reviewed in each biweekly chart rounds session, with a total of 560 case presentations. The average time required for contour review was 145 seconds (±71 seconds) and plan review was 120 seconds (±64 seconds). Modifications were suggested for 21 patients (7.9%) during contour review and for 19 patients (6.4%) during treatment plan review. An average of 4 physicians were present at each session.ConclusionsWe demonstrated that the implementation of a comprehensive, prospective peer review program is feasible in the community setting. This article can serve as a framework for future quality assurance programs.

Utility of Artificial Intelligence Tool as a Prospective Radiology Peer Reviewer — Detection of Unreported Intracranial Hemorrhage

Misdiagnosis of intracranial hemorrhage (ICH) can adversely impact patient outcomes. The increasing workload on the radiologists may increase the chance of error and compromise the quality of care provided by the radiologists. We used an FDA approved artificial intelligence (AI) solution based on a convolutional neural network to assess the prevalence of ICH in scans, which were reported as negative for ICH. We retrospectively applied the AI solution to all consecutive noncontrast computed tomography (CT) head scans performed at eight imaging sites affiliated to our institution. In the 6565 noncontrast CT head scans, which met the inclusion criteria, 5585 scans were reported to have no ICH ("negative-by-report" cases). We applied AI solution to these "negative-by-report" cases. AI solution suggested there were ICH in 28 of these scans ("negative-by-report" and "positive-by-AI solution"). After consensus review by three neuroradiologists, 16 of these scans were found to have ICH, which was not reported (missed diagnosis by radiologists), with a false-negative rate of radiologists for ICH detection at 1.6%. Most commonly missed ICH was overlying the cerebral convexity and in the parafalcine regions. Our study demonstrates that an AI solution can help radiologists to diagnose ICH and thus decrease the error rate. AI solution can serve as a prospective peer review tool for non-contrast head CT scans to identify ICH and thus minimize false negatives.

Analysing breast dose in female lymphoma patients who received radiotherapy: a retrospective audit

Keywords: Lymphoma, Breast cancer, Involved field, Involved node, 3DCRT Introduction: Radiotherapy improves progression-free survival in lymphoma, though it increases the risk of secondary malignancies post-treatment1,2. Second primary breast cancers contribute most to the overall absolute excess risk of secondary malignancy in Hodgkin’s lymphoma patients, with radiotherapy thought to play a major role in carcinogenesis2. However, the relationship between doses received and risk is unclear, and referral to screening programmes is mainly based upon data from out-dated mantle treatments1-5. The aim of this study therefore was to audit breast dose in female patients treatment with conformal radiotherapy approaches for mediastinal-involved lymphoma. Methodology: Patients were selected retrospectively from a locally held database6 of women aged ≤36 who received supradiaphragmatic radiotherapy for lymphoma between 2008-2012. Inclusion criteria: •Mediastinal involvement.•30Gy prescription dose.•AP-PA field arrangement. Exclusion criteria:•Neck only treatments•Plans with breast tissue pre-contoured•Plans with incomplete breast tissue. All breast contours were constructed to RTOG/EORTC guidelines and reviewed by a radiologist7,8. Radiotherapy dose statistics were retrieved from the departmental planning software. Results: 26 patients were analysed. Mean dose to both breasts was 2.3Gy, with some patients with advanced disease receiving at least 4Gy to >80% of a single breast. A mean dose for both breasts of 10.6Gy was observed. Mean Dose to D1% of both breasts combined was 24.5Gy, despite most plans having no breast tissue within the PTV. Discussion and conclusion: Mean dose, V4Gy, V20Gy, and D1% results were predominantly within the maximum and minimum range of mean results in the literature for conformal treatments9-18. Differences in results between patients, stemmed from variability in disease spread and disease location, along with variation in field parameters. Subsequently, advanced stage patients tended to receive significantly higher breast doses, and further audits may wish to examine the implications of this. This audit was limited by a lack of information regarding target volume outlining decision-making and staging. Future audits may also wish to examine the effect of prospective peer review of target volumes on breast dose. Numerical References

Prospective Peer Review in Radiation Therapy Treatment Planning: Long-Term Results From a Longitudinal Study

PurposeTo present the longitudinal results of a prospective peer review evaluation system (PES) before treatment planning. Methods and MaterialsAll cases undergoing radiation therapy (RT) at high-volume academic institutions were graded in daily prospective multidisciplinary contouring rounds (CRs). The clinical suitability for RT, prescription, contours, and written directives were peer reviewed, compared with departmental care pathways, and recorded in a prospective database. Grades were assigned as follows: A (score 4.0) = no deficiencies; B (3.0) = minor modifications of the planning target volume, organs at risk, written directives, or a prescription/care pathway mismatch; and C (2.0) = incomplete target volume or organ-at-risk contours, unsuitable use or inappropriate planned administration of RT, significant contour modifications, prescription changes, or laterality modifications. Information was pooled to determine pretreatment planning work performance by assigning a grade point average (GPA) for each physician as well as compositely. ResultsA total of 11,843 treatment plans from 7854 patients were reviewed using the PES from September 2013 to May 2018. Twenty-seven point nine percent of cases (n = 3303) required modifications before treatment planning commenced. The overall breakdown of grades was 72.1% As, 21.7% Bs, and 6.2% Cs. The median physician CR GPA was 3.60 (average 3.7) with a range of 3.0 to 3.9. Seventy-five percent of physicians demonstrated improvement of their CR GPA since inception of the program, and all physicians demonstrated a drop in the percentage of cases that were assigned a grade of C. ConclusionsThe PES can transparently quantify clinical performance in a single metric. The PES was impactful, with 75% of physicians demonstrating improvement in their CR GPA over time. In contrast to traditional chart rounds, this peer review was meaningful when done before planning commenced, a trend that was observed throughout the study period. Twenty-seven point nine percent of all cases required modification before starting treatment planning, and 6.2% of cases required significant remediation.

Radiation Oncology Peer Review in a Community Setting

Peer review is an important component of any radiation oncology continuous quality improvement program. While limited guidelines exist, there is no consensus about how peer review should be performed, and large variations exist among different institutions. The purpose of this report is to retrospectively describe our experience with peer review at a busy community radiation oncology clinic. Here we report the results of our program and its impact on the treatment process. In 2012 our clinic implemented a prospective peer review policy with a goal of reviewing all cases before treatment begins. As soon as treatment planning is completed, the patient is added to the peer review list for the upcoming weekly session. Peer review sessions are held weekly and attended by physicians, physicists and dosimetrists. Our targets of peer review include patient history, diagnosis, prescription, dosimetric constraints, and plan quality. Starting in 2015, every peer review session is tracked including recommended changes. Our clinical workflow is designed so that every new patient is peer reviewed. Between April 2015 and February 2019, a total of 2581 patients were peer reviewed. Out of those, 1317 were prospective reviews (51%). Those that were reviewed after the treatment had already begun, were reviewed within the first week of treatment. Only 30 plans (1.2 %) were changed based on peer review recommendations. Out of those that were changed, 25 (1%) were changed before the start of treatment. The changes to plans were made to reduce dose to organs at risk (OAR) (15), change prescription (5), change setup or rescan for anatomical changes (5), change treatment volume (4), and change OAR volume (1). Changes that were made to the 5 patients (0.2%) already on treatment were to reduce dose to OARs (bowel (2), brachial plexus, and contralateral breast) and change prescription dose (1). While there is no national standard for peer review, it is evident that prospective peer review is preferable. Since 49% of our patients were not prospectively reviewed, peer review at our institution should be revised in the future. Plans reviewed before treatment were 5 times more likely to be changed by peer review than those reviewed after the start of treatment (1% vs. 0.2%). This may reflect a subconscious reluctance to change plans already underway, which could be a barrier to improving plans with the peer review process. Rather than reviewing in a group setting, it would be ideal to individually assign review that is embedded in the clinical flow, assuring prospective review for all patients prior to final physician approval. Peer review in a group setting also may be less candid, due to interpersonal concerns about publicly disagreeing with colleagues. One of the biggest obstacles to prospective peer review is the limited time between simulation and treatment, and adding a required prospective peer review step could slow down the clinical flow and delay patients’ treatment.

Assessing variability in surgical decision making among attending neurosurgeons at an academic center.

Although it is known that intersurgeon variability in offering elective surgery can have major consequences for patient morbidity and healthcare spending, data addressing variability within neurosurgery are scarce. The authors performed a prospective peer review study of randomly selected neurosurgery cases in order to assess the extent of consensus regarding the decision to offer elective surgery among attending neurosurgeons across one large academic institution. All consecutive patients who had undergone standard inpatient surgical interventions of 1 of 4 types (craniotomy for tumor [CFT], nonacute redo CFT, first-time spine surgery with/without instrumentation, and nonacute redo spine surgery with/without instrumentation) during the period 2015-2017 were retrospectively enrolled (n = 9156 patient surgeries, n = 80 randomly selected individual cases, n = 20 index cases of each type randomly selected for review). The selected cases were scored by attending neurosurgeons using a need for surgery (NFS) score based on clinical data (patient demographics, preoperative notes, radiology reports, and operative notes; n = 616 independent case reviews). Attending neurosurgeon reviewers were blinded as to performing provider and surgical outcome. Aggregate NFS scores across various categories were measured. The authors employed a repeated-measures mixed ANOVA model with autoregressive variance structure to compute omnibus statistical tests across the various surgery types. Interrater reliability (IRR) was measured using Cohen's kappa based on binary NFS scores. Overall, the authors found that most of the neurosurgical procedures studied were rated as "indicated" by blinded attending neurosurgeons (mean NFS = 88.3, all p values < 0.001) with greater agreement among neurosurgeon raters than expected by chance (IRR = 81.78%, p = 0.016). Redo surgery had lower NFS scores and IRR scores than first-time surgery, both for craniotomy and spine surgery (ANOVA, all p values < 0.01). Spine surgeries with fusion had lower NFS scores than spine surgeries without fusion procedures (p < 0.01). There was general agreement among neurosurgeons in terms of indication for surgery; however, revision surgery of all types and spine surgery with fusion procedures had the lowest amount of decision consensus. These results should guide efforts aimed at reducing unnecessary variability in surgical practice with the goal of effective allocation of healthcare resources to advance the value paradigm in neurosurgery.