Association Between Physicians' Appropriate Use of Echocardiography and Subsequent Healthcare Use and Outcomes in Patients With Heart Failure.

Background There is little understanding of whether a physician's tendency to order an inappropriate cardiac service is associated with the use of other cardiac services and clinical outcomes in their patients with heart failure (HF). Methods and Results We conducted a secondary analysis of 35 Ontario-based cardiologists who participated in the control arm of the Echo WISELY (Will Inappropriate Scenarios for Echocardiography Lessen Significantly) trial. Transthoracic echocardiograms, ordered during the trial, were classified as rarely appropriate (rA), appropriate, or maybe appropriate on the basis of the 2011 appropriate use criteria. Cardiologists were grouped into tertiles of rA transthoracic echocardiogram ordering frequency: low ordering (bottom tertile), n=11; moderate ordering, n=12; or high ordering (top tertile), n=12. The main outcomes were measures of cardiac service use, including cardiology-related physician visits, tests, and medications. Among 1677 patients with heart failure and an outpatient visit to 1 of 35 cardiologists, we found no significant association between rA transthoracic echocardiogram ordering frequency (by tertile) and cardiac testing use, although patients of cardiologists in the high ordering group had fewer physician visits, on average, than patients seen by low ordering cardiologists. In addition, patients of cardiologists in the highest rA ordering tertile had significantly lower odds of receiving potentially effective interventions, such as β blockers (odds ratio, 0.62; 95% CI, 0.43-0.89), than the low ordering group. Conclusions Although patients of cardiologists who frequently order rA transthoracic echocardiograms do not appear more (or less) likely to have subsequent cardiac tests, these patients have fewer follow-up visits and lower odds of receiving evidence-based medications. Clinical Trial Registration URL: http://www.clinicaltrials.gov. Unique identifier: NCT02038101.