Comparison of the Effectiveness of Single-Component and Multicomponent Interventions for Reducing Radiation Doses in Patients Undergoing Computed Tomography: A Randomized Clinical Trial

Rebecca Smith-Bindman, Philip Chu, Yifei Wang, Robert Chung, Naomi Lopez-Solano, Andrew J. Einstein, Leif Solberg, Luisa F. Cervantes, Thomas Yellen-Nelson, William Boswell, Bradley N. Delman, Phuong Anh Duong, Allen R. Goode, Nima Kasraie, Ryan K. Lee, Rebecca Neill, Anokh Pahwa, Pavlina Pike, Jodi Roehm, Sebastian SchinderaJay Starkey, Saravanabavaan Suntharalingam, Cécile R.L.P.N. Jeukens, Diana L. Miglioretti

Research output: Contribution to journalArticlepeer-review

12 Scopus citations

Abstract

Importance: Computed tomography (CT) radiation doses vary across institutions and are often higher than needed. Objective: To assess the effectiveness of 2 interventions to reduce radiation doses in patients undergoing CT. Design, Setting, and Participants: This randomized clinical trial included 864080 adults older than 18 years who underwent CT of the abdomen, chest, combined abdomen and chest, or head at 100 facilities in 6 countries from November 1, 2015, to September 21, 2017. Data analysis was performed from October 4, 2017, to December 14, 2018. Interventions: Imaging facilities received audit feedback alone comparing radiation-dose metrics with those of other facilities followed by the multicomponent intervention, including audit feedback with targeted suggestions, a 7-week quality improvement collaborative, and best-practice sharing. Facilities were randomly allocated to the time crossing from usual care to the intervention. Main Outcomes and Measures: Primary outcomes were the proportion of high-dose CT scans and mean effective dose at the facility level. Secondary outcomes were organ doses. Outcomes after interventions were compared with those before interventions using hierarchical generalized linear models adjusting for temporal trends and patient characteristics. Results: Across 100 facilities, 864080 adults underwent 1156657 CT scans. The multicomponent intervention significantly reduced proportions of high-dose CT scans, measured using effective dose. Absolute changes in proportions of high-dose scans were 1.1% to 7.9%, with percentage reductions in the proportion of high-dose scans of 4% to 30% (abdomen: odds ratio [OR], 0.82; 95% CI, 0.77-0.88; P <.001; chest: OR, 0.92; 95% CI, 0.86-0.99; P =.03; combined abdomen and chest: OR, 0.49; 95% CI, 0.41-0.59; P <.001; and head: OR, 0.71; 95% CI, 0.66-0.76; P <.001). Reductions in the proportions of high-dose scans were greater when measured using organ doses. The absolute reduction in the proportion of high-dose scans was 6.0% to 17.2%, reflecting 23% to 58% reductions in the proportions of high-dose scans across anatomical areas. Mean effective doses were significantly reduced after multicomponent intervention for abdomen (6% reduction, P <.001), chest (4%, P <.001), and chest and abdomen (14%, P <.001) CT scans. Larger reductions in mean organ doses were 8% to 43% across anatomical areas. Audit feedback alone reduced the proportions of high-dose scans and mean dose, but reductions in observed dose were smaller. Radiologist's satisfaction with CT image quality was unchanged and high during all periods. Conclusions and Relevance: For imaging facilities, detailed feedback on CT radiation dose combined with actionable suggestions and quality improvement education significantly reduced doses, particularly organ doses. Effects of audit feedback alone were modest. Trial Registration: ClinicalTrials.gov Identifier: NCT03000751.

Original languageEnglish (US)
Pages (from-to)666-675
Number of pages10
JournalJAMA Internal Medicine
Volume180
Issue number5
DOIs
StatePublished - May 2020

ASJC Scopus subject areas

  • Internal Medicine

Fingerprint

Dive into the research topics of 'Comparison of the Effectiveness of Single-Component and Multicomponent Interventions for Reducing Radiation Doses in Patients Undergoing Computed Tomography: A Randomized Clinical Trial'. Together they form a unique fingerprint.

Cite this