Purpose or Case Report: To diagnose intestinal disease or guide an injection treatment, radiation from a fluoroscopy exam is often nontrivial. In addition to regulatory and accreditation requirements, it is clinically important to establish a mechanism to review and improve the use of fluoroscopy, particularly for pediatric patients who are more sensitive to radiation than adults. In this study, we aimed to implement a fluoroscopy dose monitor program, assess overall fluoroscopy performance, and evaluate radiologist performance, in a pediatric hospital system.
Methods & Materials: In April 2020, five under table fluoroscopy units were configured to send exam dose reports to PACS and dose reports were then retrieved by commercial software for analysis. Reference levels of accumulative air kerma (AK) and fluoroscopy time were determined as 95% of system-wise data to trigger alerts and were re-calculated in a quarterly interval. Quarterly dose reports were generated and reviewed by the Fluoroscopy Dose Review Committee which consists of radiologist fluoroscopy director, lead technologists, and medical physicists. In quarterly meeting, the overall alerts, high dose and long procedure events, performance of each radiologist, and fluoroscopy protocols were reviewed. Performance improvement recommendations from the committee were provided for the radiologists and technologists team during staff meetings; necessary protocol optimization was carried out by medical physicists.
Results: For a 2-year span, the percentages of exams triggering alerts were gradually decreased for both accumulative reference dose (from 5.26% to 3.64%) and fluoroscopy time (from 5.41% to 3.08%). For 3 most performed protocols, dose reference levels of video swallow study (VFSS) and voiding cystourethrogram (VCUG) was stabilized ~3.6mGy and 0.75mGy, respectively, while that of upper gastrointestinal (UGI) substantially reduced from 12.5mGy to 2.3mGy. The overall individual radiologist performance was also slightly improved over 2 years. The most common reasons of high dose alerts over 2 years were large patient habitus, complicated medical conditions, and uncooperative patients. The AK values of VFSS, VCUG and UGI, are linearly correlated with patient weight (linear regression, p<0.0001), with a slope of 0.005, 0.02, 0.07 mGy/kg, respectively.
Conclusions: We demonstrated successful implementation of a pediatric fluoroscopy dose monitor program and observed improvement of clinical performance via quarterly review with radiologist and technologist education.