Summary
A medRxiv preprint found that appropriate-use guidance reduced nuclear stress-test ordering for some emergency observation-unit patients. Echocardiogram use increased, leaving combined use of the two tests essentially unchanged.
A programme promoting appropriate-use guidelines reduced nuclear stress-test ordering for some patients in an emergency-department observation unit, but it did not reduce the combined use of nuclear stress tests and echocardiograms, according to a medRxiv preprint posted on September 17, 2026.
The before-and-after analysis compared adults admitted to the observation unit during 2024, before implementation, with those admitted during 2025, the first full year after the programme began. Eligible patients had chest pain, dyspnoea or shortness of breath, dizziness, or syncope.
What changed after the guidance was introduced
The programme was developed with the hospital’s cardiology department and promoted guidelines for using echocardiograms and stress tests with nuclear components. An echocardiogram uses ultrasound to assess the heart, while a nuclear stress test uses a radioactive tracer to evaluate blood flow and heart function during stress.
Nuclear stress-test use fell by 9.3 percentage points among patients presenting with dyspnoea or shortness of breath, a statistically significant change with p = 0.0009. Across all four complaint groups, nuclear stress-test use fell by 2.4 percentage points, with p = 0.0095.
Echocardiogram use moved in the opposite direction. Ordering increased by 3.7 percentage points for patients with chest pain, with p = 0.005, and by 2.0 percentage points across all eligible patients, with p = 0.0454.
Taken together, the two types of testing were used for 85.2% of eligible patients in 2024 and 84.8% in 2025. That difference was not statistically significant, with p = 0.5768. The programme therefore changed the balance between the tests more clearly than it changed the overall proportion of patients receiving either test.
Why the result matters
The authors describe unnecessary cardiac imaging as a concern because it can increase healthcare costs, limit access to testing resources and expose patients to avoidable radiation from nuclear studies. The reduction in nuclear stress-test use is consequently relevant to diagnostic stewardship: selecting tests according to clinical indications rather than ordering them routinely for broad symptom categories.
The findings also show why changing overall test use can be difficult. A reduction in one test may be accompanied by greater use of another modality. In this setting, the guidance was associated with fewer nuclear stress tests but more echocardiograms, leaving combined use broadly stable.
The authors attribute the reduction in nuclear testing partly to involvement from both emergency medicine and cardiology, use of guidelines adopted by professional organisations and wider efforts to promote evidence-based care. They suggest that the lack of a comparable reduction in echocardiograms may reflect slow adoption of updated medical knowledge, limited involvement of house staff and advanced-practice clinicians in developing the guidance, and insufficient reinforcement through multiple communication channels and departmental leadership.
This was a retrospective comparison of test-ordering patterns rather than a randomised trial. The preprint reports utilisation changes, not whether the programme affected diagnostic accuracy, missed conditions, patient outcomes or total costs. It is also a medRxiv preprint and has not yet undergone peer review.