Low-value transthoracic echocardiography, healthcare utilization, and clinical outcomes in patients with coronary artery disease

Background — The relationship between ordering frequency of rarely appropriate transthoracic echocardiograms on healthcare utilization and patient outcomes in coronary artery disease (CAD) is not known. Our objective was to investigate practice patterns of cardiologists who order a high frequency of low-value transthoracic echocardiograms in patients with CAD and whether practice behavior influences patient outcomes.

Methods and Results — A retrospective cohort of outpatient CAD patients was accrued by identifying patients with at least 1 visit to 1 of 35 Ontario-based cardiologists in the EchoWISELY randomized clinical trial (Will Inappropriate Scenarios for Echocardiography Lessen Significantly) control group. The main outcomes of interest were patient-level receipt of diagnostic tests, physician visits, medication prescriptions, and clinical outcomes at 1 year. Our cohort consisted of 3966 patients with CAD (mean [SD] age, 67.8 [12.0] years; 72% men), with an outpatient visit to 1 of 35 eligible cardiologists, stratified into 3 ordering tertiles. Patients of cardiologists in the top ordering tertile of rarely appropriate transthoracic echocardiograms had significantly lower odds of receiving the following services at 1 year compared with patients in the low ordering group: cholesterol assessment (odds ratio [OR], 0.77 [95% CI, 0.65–0.91]); hemoglobin A1c assessment (OR, 0.79 [95% CI, 0.66–0.94]); β-blocker prescription (OR, 0.70 [95% CI, 0.55–0.90]); and aldosterone receptor antagonist prescription (OR, 0.46 [95% CI, 0.22–0.98]). Patients of high ordering cardiologists had greater odds of all-cause mortality at 1 year (OR, 1.54 [95% CI, 1.04–2.28]), although all other outcomes were similar.

Conclusions — Patients with CAD seen by cardiologist who ordered a high rate of rarely appropriate transthoracic echocardiograms were less likely to receive potentially high-value screening tests and evidence-based medications than low ordering cardiologists.