Leveraging Dose Tracking Software to Monitor CT Repeat Rates: A Multi-Protocol Analysis
Abstract
Purpose
To utilize data from a dose tracking software to retrospectively track repeat rates for frequently used CT exams
Methods
Data was exported from a dose tracking system (DTS) (Imalogix) for head without contrast, abdomen pelvis with contrast, and chest without contrast exams performed on 15 scanners over 4 years. Additionally, coronary CT angiography data were utilized for a wide-beam scanner over 2 years. 174,605 exams were included in the analysis done with Python. Due to acquisition-level protocol mapping in the DTS, only relevant acquisitions to the protocol category in multiple-series exams were assumed to remain in the data. To account for true repeats, CTDIvol and scan-length thresholds were used to exclude incorrectly mapped scout, bolus-tracking, and other acquisitions. The number of acquisitions (NOA) was assigned corresponding to each exam. The data were grouped by protocol name, and the most frequent appearing NOA in each group was deemed the “normal NOA”. Any exam with a NOA greater than the normal NOA was deemed a repeat. Data was analyzed monthly. Repeated exams were carefully reviewed in PACS to validate the method.
Results
Monthly repeat rates ranged from 1.9-3.8% (Head WO), 2.9-6.4% (Abdomen Pelvis W), 0.5-5.8% (Chest WO), and 0-3% (CCTA). Investigation into individual exams showed repeats were often due to patient motion. The discovery of unjustified repeats due to technologist errors led to one-on-one retraining. The repeat rate for the chest protocol decreased, while the rate for head, abdomen pelvis, and CCTA protocols remained steady.
Conclusion
Utilizing data from a DTS provides a unique opportunity for institutions to track CT repeat rates. Tracking CT repeat rates may lead to insights into factors outside of the protocol parameters that lead to excess patient dose. In alignment with ALARA, retraining efforts are in progress to try to reduce unjustified repeats.