Inpatient Stress Test Modality and Downstream Clinical Pathways in Troponin-Negative Patients: A Real-World Cohort...

Background Troponin-negative patients presenting with suspected acute coronary syndrome are frequently admitted for inpatient stress testing. The relative yield of different stress test modalities and their influence on downstream clinical decision-making in this population remain uncertain. Methods We conducted a retrospective cohort study of 769 consecutive troponin-negative adult patients (≥18 years) undergoing inpatient stress testing at St Luke's, a community teaching hospital in the Midwest, USA. Stress test modalities were categorized as nuclear myocardial perfusion imaging (MPI), stress echocardiography, or exercise electrocardiography (ECG)-only. The primary outcome was stress test positivity defined as the presence of inducible ischemia as interpreted by the reading cardiologist. This included reversible perfusion defects on nuclear MPI, inducible regional wall motion abnormalities on stress echocardiography, or ischemic ECG changes during exercise testing. Secondary outcomes included referral to invasive coronary angiography (ICA), detection of obstructive coronary artery disease (CAD), and subsequent revascularization. Multivariable Poisson regression models with robust standard errors were used to evaluate the association between stress test modality and outcomes after adjustment for baseline characteristics. Results Nuclear MPI was the most commonly conducted modality (n=616, 80.1%), followed by stress echocardiography (n=144, 18.7%) and exercise ECG-only testing (n=9, 1.2%). Comparisons were performed between nuclear MPI and stress echocardiography, as the ECG-only subgroup was small. Stress test positivity rates were similar between nuclear MPI and stress echocardiography (n=60, 9.7% vs n=12, 8.3%, p=0.80). No positive tests were observed in the ECG-only group, though this subgroup was very small. Referral to ICA was primarily driven by stress test results rather than modality. Among patients with positive stress tests, ICA was performed in 80% (48/60) of nuclear MPI and 75% (9/12) of stress echocardiography cases (p=0.71), whereas ICA was infrequently performed following negative tests (5.8% (32/556) vs 2.3% (3/132), respectively; p=0.23). Overall, ICA utilization did not differ significantly between nuclear MPI (n=80, 13.0%) and stress echocardiography (n=12, 8.3%), p=0.21. Among patients undergoing ICA, rates of obstructive CAD (37.5% (30/80) vs 41.7% (5/12), p=0.76) and revascularization (32.5% (26/80) vs 41.7% (5/12), p=0.53) were also similar between nuclear MPI and stress echocardiography groups, respectively. In adjusted analyses, stress test modality was not independently associated with stress test positivity (adjusted risk ratio 0.97, 95% CI 0.51-1.85; p=0.94) or referral to ICA (adjusted risk ratio 0.61, 95% CI 0.32-1.14; p=0.12). Conclusions In this single-center retrospective cohort, both unadjusted and adjusted analyses demonstrated similar stress test yield and downstream referral to ICA between nuclear MPI and stress echocardiography. These findings suggest that, in this setting, stress test modality may have limited influence on clinical decision-making following negative troponin evaluation. Despite consistent findings after adjustment, residual confounding cannot be excluded; therefore, these results should be interpreted as observational associations rather than evidence of causal effects.