Redundant Imaging in Transient Ischemic Attack: Evidence From the Nationwide Emergency Department Sample

Vincent M Timpone; Alexandria Jensen; Sharon N Poisson; Margaret Reid; Matthew Salzberg; Premal S Trivedi

doi:10.1016/j.jacr.2021.07.003

Redundant Imaging in Transient Ischemic Attack: Evidence From the Nationwide Emergency Department Sample

J Am Coll Radiol. 2021 Nov;18(11):1525-1531. doi: 10.1016/j.jacr.2021.07.003. Epub 2021 Jul 28.

Authors

Vincent M Timpone¹, Alexandria Jensen², Sharon N Poisson³, Margaret Reid⁴, Matthew Salzberg⁵, Premal S Trivedi⁶

Affiliations

¹ Director, Stroke and Vascular Imaging and Co-Director Neuroradiology Spine Intervention Service, Department of Radiology, University of Colorado Hospital, Aurora, Colorado. Electronic address: vincent.timpone@cuanschutz.edu.
² Department of Biostatistics & Informatics, Colorado School of Public Health, Aurora, Colorado.
³ Director, Vascular and Stroke Research Fellowship, Department of Neurology, University of Colorado Hospital, Aurora, Colorado.
⁴ Department of Health Systems, Management & Policy, Colorado School of Public Health, Aurora, Colorado.
⁵ Assistant Medical Director, Department of Emergency Medicine, University of Colorado Hospital, Aurora, Colorado.
⁶ Director, Health Services Research, Department of Radiology, University of Colorado Hospital, Aurora, CO.

PMID: 34329612
DOI: 10.1016/j.jacr.2021.07.003

Abstract

Purpose: Increasing emergency department (ED) compliance with transient ischemic attack (TIA) imaging guidelines has previously been demonstrated, along with a substantial rise in imaging utilization over the past decade. The purpose of this study was to characterize the most commonly used combinations of imaging studies during ED workup of TIA and to quantify prevalence of redundant imaging (RI).

Methods: TIA discharges from EDs in the United States from 2006 to 2017 were identified in the Nationwide Emergency Department Sample. Brain and neurovascular imaging obtained during the encounter was identified using Current Procedural Terminology codes. RI was defined as an ED encounter with any duplicate cross-sectional brain, brain-vascular, or neck-vascular imaging. Patient demographics and hospital characteristics were incorporated into a multivariable logistic regression analysis to identify significant associations with RI.

Results: There were 184,870 discharges with TIA from EDs in 2017. RI (brain) was observed in 55,513 (30%) of encounters. RI (brain-vascular) and RI (neck-vascular) imaging was identified in 5,149 (2.8%) and 1,325 (0.7%) of encounters, respectively. Decreased odds of obtaining RI was observed in Medicaid patients (odds ratio [OR]: 0.72, 95% confidence interval [CI]: 0.64-0.81), non-trauma centers (OR: 0.49, 95% CI: 0.26-0.93), rural hospital locations (OR: 0.18, 95% CI: 0.11-0.29), and weekend encounters (OR: 0.9, 95% CI: 0.85-0.96). Trend analysis from 2006 to 2017 demonstrated a rise in RI (brain) from 2.3% of encounters in 2006 to 30% of encounters in 2017. RI for patients discharged from EDs with TIA in 2017 resulted in additional charges of approximately US$8,670,832.

Conclusion: Increased imaging utilization for TIA workup across EDs in the United States is associated with rising use of redundant imaging. We identify imaging practices that could be targeted to mitigate health care expenditures while adhering to TIA imaging guidelines.

Keywords: Appropriateness Criteria; NEDS; imaging utilization; transient ischemic attack.

MeSH terms

Cross-Sectional Studies
Emergency Service, Hospital
Humans
Ischemic Attack, Transient* / diagnostic imaging
Ischemic Attack, Transient* / epidemiology
Odds Ratio
Patient Discharge
Retrospective Studies
Stroke*
United States / epidemiology