Bhatia, R.S., et al, JAMA Intern Med 177(9):1326, September 1, 2017
BACKGROUND: Both the USPSTF and the Choosing Wisely campaign recommend against routine ECG screening in low-risk patients. A routine ECG at the time of an annual physical in this population is considered to be an example of low-value care.
METHODS: The authors, coordinated at Women’s College Hospital in Toronto, examined the frequency of ECGs after annual health examinations in low-risk adults seen in primary care. Scrutiny of provincial databases identified 3,629,859 patients who had an annual exam in 2010-2015, excluding those with a history of cardiac disease or high-risk criteria. The primary outcomes were receipt of an ECG within 30 days after the annual exam and downstream cardiac care (cardiac testing or consultations) within 90 days.
RESULTS: Just over one-fifth of the patients (21.5%) had an ECG following the annual exam. Rates of ECG ordering varied widely among regions (from 0.7% to 24.4%), among the 679 primary care practices (1.8% to 76.1% of patients), and among the 8036 primary care physicians (1.1% to 94.9%). Receipt of an ECG was significantly more likely for older patients with certain comorbidities (cancer, rheumatologic disease) and less likely for rural residents. Physician traits associated with ECG ordering included male sex, medical school in an international program, and practicing for 30 years or more; in fact, practice-level variation explained 22% of the variation in ECG use. Patients having (versus not having) ECGs had significantly higher rates of cardiac consultations (odds ratio [OR] 5.38; 95% CI 5.24- 5.52) and cardiac tests (transthoracic echocardiogram, OR 7.1; stress test, OR 6.5; and nuclear stress test, OR 4.2). Despite this, one-year follow-up was consistent with low rates of cardiac morbidity and mortality in both groups.
CONCLUSIONS: Routine performance of an ECG appears to be relatively common in low-risk patients and increases the likelihood of unnecessary downstream testing. 37 references ( firstname.lastname@example.org – no reprints)