Abstract 15537: An Electronic Health Record Intervention to Optimize Utilization of Transthoracic Echocardiography: Opportunity for Cost Savings in the Era of Value Based-care

Introduction: In the current value-based era, targeting diagnostic resources and minimizing unnecessary testing is of paramount importance. Transthoracic echocardiography (TTE) is a common and costly test, and available Appropriate Use Criteria (AUC) guide optimal utilization. Limited TTE (L-TTE) shortens sonographer time, lowers cost and may be ideal for repeat TTEs (R-TTE) with a focused indication. However, many clinicians are unfamiliar with the AUC and opportunities for L-TTE. We prospectively tested an Electronic Health Record (EHR)-based intervention aimed at optimizing TTE utilization in a large academic medical center. Methods: TTE utilization at the University of Chicago Medicine was assessed over a 6-month period and complete TTE (C-TTE), L-TTE and R-TTE (TTE repeated within 6 months) were recorded. An EHR-based intervention was then implemented and TTE utilization was assessed over the ensuing 8 weeks. The intervention included presenting new descriptive L-TTE options (i.e. “Limited TTE: EF or Effusion Only”) when any “echo” was searched in the EHR order panel, an alert to prior TTEs (i.e. date & LVEF) and a link to AUC-based guidance for TTE ordering. Educational materials were also distributed to frequent TTE ordering providers. Results: Among 9121 TTEs (53% inpatient) pre-intervention , 11% (n=1002) were L-TTEs and 25% (n=2320) were R-TTEs. There were more L-TTEs and R-TTEs in pre-intervention inpatients compared to outpatients (L-TTE 14% vs 7%, p<0.0001, R-TTE 33% vs 17%, p<0.0001). Post-intervention (2879 TTEs, 53% inpatient), R-TTEs significantly decreased (22.6% vs 25.4%, p=0.0019) and L-TTEs significantly increased (14% vs 11%, p<0.0001) compared to pre-intervention, with inpatient TTEs most impacted (R-TTE 28% vs 33%, p=0.0016, L-TTE 19% vs 14%, p<0.001). The intervention’s greatest impact was to markedly increase L-TTEs among inpatient R-TTEs (44% vs 35%, p=0.0002). Conclusions: Despite AUC discouraging frequent repeat TTEs, R-TTEs are common in an academic medical center and utilization of L-TTE is rare. An EHR-based intervention with prior TTE alerts and descriptive L-TTE options increases L-TTEs and reduces R-TTEs. Further study is warranted to describe the full clinical and financial impact of this intervention.