Abstract 4140366: Diagnostic Stewardship to Improve Appropriate Use of Inpatient Echocardiography

Background: Overuse of cardiovascular imaging, including echocardiography, is a form of low-value care contributing to waste in the US healthcare system. Historically, appropriate use criteria (AUC)-based educational programs and feedback reports have improved transthoracic echocardiography (TTE) ordering practices, but prior efforts have shown only modest and short-lived impact and have focused on the outpatient setting. Contemporary capacity and financial constraints highlight the importance of leveraging AUC to promote more efficient utilization of echocardiography resources. Methods: We conducted a prospective controlled cohort study testing an education-based diagnostic stewardship intervention to reduce rarely appropriate TTEs on inpatient medical services. AUC categorize TTE indications into three groups: appropriate (A), may be appropriate (mA), and rarely appropriate (rA). Modeled after successful antimicrobial stewardship programs, the intervention involved screening TTE orders from control and intervention units for appropriateness in real-time. For medical units in the intervention group, the diagnostic steward contacted ordering clinicians of those with mA and rA orders. The steward clarified order indications and made recommendations regarding the utility of the TTE. The ordering clinician chose to continue or discontinue the order. The primary outcome was the odds of rA TTEs in the intervention vs control groups. Results: Over 3 months, 150 TTEs were screened in the control group and 160 were screened in the intervention group. After initial screening, 8.8% (13) and 10.6% (17) were rA in the control and intervention groups, respectively. The stewardship intervention was employed for 17 rA TTEs and 13 mA TTEs. After the stewardship intervention, 17 of the rA TTEs remained classified as rA and 1 of the mA TTEs was reclassified as rA. Of the 18 rA orders, 16 were discontinued and 2 were completed (88% reduction). The stewardship intervention was associated with a significant reduction in completed rA TTEs when compared to the control group (OR = 0.15; 95% CI 0.03, 0.67; p = 0.005; Table). Conclusion: A stewardship intervention resulted in a significant decrease in completed rA inpatient TTEs among screened orders. In the era of capacity constraints, such a reduction represents a solution to improve clinical operations through targeted resource allocation. Future work to partially automate this initiative is important for scalability.