Introduction: Of all stress tests done in low risk Emergency Department observation units (OU), a small, but significant number may be reported as positive or indeterminate. The objective of this study is to quantify the prevalence and costs associated with positive and indeterminate stress tests that result in negative cardiac catheterization.
Methods: Retrospective observational cohort study over 9 months. All patients undergoing the chest pain protocol who got cardiac stress testing in the OU were eligible for inclusion. Cost data were derived from an institutional activity-based cost system utilizing actual costs. Chart review was completed on all patients with positive and indeterminate stress tests and a randomly chosen sample of those with negative stress tests.
Results: Of the 1194 patients who met the inclusion criteria, 1084 (90.8%) had a negative stress test. Sixty-two (5.2%) had a positive stress test, and 48 (4.0%) had an indeterminate stress test. Of all 59 patients who underwent catheterization, 41 (69.5%) were negative cardiac catheterizations. The prevalence among all OU stress test patients of positive or indeterminate stress tests with subsequent negative cardiac catheterization was 41/1194 (3.4%; 95% CI 2.5%-4.6%). The prevalence of significant coronary artery disease at cardiac catheterization was 18/1194 (1.5%; 95% CI 1.0%-2.4%). Patients with a positive or indeterminate stress test who had a negative catheterization incurred increased OU costs ($1385 vs. $1,039, P = 0.012), total costs ($7298 vs. $1562, P < 0.001) and length of inpatient stay (1.83 days vs. 0.00 days) when compared with those who had a negative stress test.
Conclusion: The probability of going to the OU and having a positive or indeterminate stress test resulting in a subsequent negative catheterization was double the probability of having a stress test result in catheterization that detected significant coronary artery disease. These patients incurred 5 times the total cost when compared with those patients with negative stress testing. Further investigation is warranted to determine alternative risk stratification methods for these low risk chest pain patients with positive stress tests.