Background: The evaluation of hospitalized patients with chest pain and non-diagnostic electrocardiogram is problematic and the optimal cost-effective strategy for their management controversial.

Objectives: To determine the utility of myocardial perfusion imaging with thallium-201 for predicting outcome of hospitalized patients with chest pain and a normal or non-diagnostic ECG.

Methods: On pain cessation, 109 hospitalized patients, age 61+14 years (mean+SD), with chest pain and non-diagnostic ECG underwent stress myocardial perfusion SPECT imaging with thallium-201. Costs related to their management were calculated. The occurrence of non-fatal myocardial infarction or cardiac death was recorded at 12+5 months follow-up.

Results: A normal SPECT was found in 84 patients (77%). During one year follow-up, only 1 (1.2%) compared to 7 (28%) cardiac events (6 myocardial infarctions, 1 cardiac death) occurred in patients with normal versus abnormal scans respectively (P < 0.0001). Negative predictive value and accuracy of the method were 99% and 83% respectively. Multivariate regression analysis identified an abnormal SPECT as the only independent predictor of adverse cardiac event (P = 0.0016). Total cost from admission until discharge was 11,193 vs. 31,079 shekels (P < 0.0001) for normal and abnormal scan. Considering its high negative predictive value, shortening the hospital stay from admission until scan performance to 2 days would result in considerably reduced management costs (from NIS 11,193 to 7,243) per patient.

Conclusion: Stress SPECT applied to hospitalized patients with chest pain and a normal or non-diagnostic ECG is safe, highly accurate and potentially cost effective in distinguishing between Iow and high risk patients.