By systematically analyzing the root causes of radiation-dose outliers—those with volumetric or dose-length product values higher than the 99th percentile and less than the first percentile in a large cohort—University of Toronto researchers were able to pinpoint corrective actions for CT scanning at a 463-bed teaching hospital.

Yingming Amy Chen, MD, and colleagues flagged and extracted the outliers from 34,615 CTs across three common exam types at St. Michael’s Hospital, Toronto, over a four-year period ending in December 2013, according to a study posted online March 4 in JACR.

In the study, the total outlier count was 529 (approximately 1.5 percent), and outlier CT exams by number were noncontrast head (n = 397), pulmonary angiographic (52) and renal colic (80).

The researchers’ analysis revealed the most frequent root causes behind the high-dose outliers were repeat exams due to patient motion (31 percent), modified protocols mislabeled as “routine” (18 percent), higher dose exams for larger patients (7 percent), repeat exams due to technical artifacts (5 percent) and repeat exams due to suboptimal contrast timing (5 percent).  

Root causes for low-outlier exams included low-dose protocols (29 percent) and aborted exams (2 percent).

“Outlier root-cause analysis helps an institution prioritize dose-optimization strategies that would have the most meaningful impact on patient dose,” the authors offer as one of their study’s key takeaways. “In our institution, lack of standardization and the use of weight-based protocols [for larger patients] were identified as the areas of dose excess needing the most improvement.”

JACR subscribers will find the study report a detailed case study demonstrating how to use radiation dose index monitoring (RDIM) software to identify, investigate and categorize high- and low-dose outliers for common CT exams.