Safety and clinical outcomes of a chest pain protocol to reduce unnecessary testing in low-risk patients

Background / Introduction

There is excessive testing in low-risk patients with stable chest pain. This leads to increased cost of care as well as increased strain on cardiac laboratories. International guidelines [1] recommend the use of clinical decision pathways and structured risk assessment to reduce unnecessary testing in these patients.

Purpose

To evaluate the safety of a structured chest pain protocol that recommends deferring cardiac testing in low-risk patients presenting with stable chest pain.

Methods

From March 2022 to December 2023, a chest pain protocol (Figure 1) incorporating the locally validated PRECISE risk score [2] was utilized to guide management for 407 patients with chest pain (protocol group) in two tertiary hospitals in Singapore. Physicians using the protocol were encouraged to defer cardiac testing and explore other causes of chest pain in these low-risk patients. The primary safety outcomes were 1-year cardiovascular hospitalisation, myocardial infarction, and all-cause mortality. These were compared with a similar cohort receiving usual care in clinics where the protocol was not utilized (Usual care group). Rates of advanced cardiac testing (stress echocardiography, myocardial perfusion imaging, computed tomography coronary angiography) were reported as a secondary outcome.

Results

353 (86.7%) patients in the protocol group were low-risk defined as PRECISE pre-test-probability of obstructive coronary artery disease <15%. Usual-care group had 367 low-risk patients.

Protocol group had lower rates of advanced cardiac testing (stress echocardiography, myocardial perfusion imaging, CT coronary angiogram) (21.5% vs 61.9%, p < 0.001).

One-year outcomes are described in Table 1. At 1-year, cardiovascular hospitalisations occurred in 3 patients (0.85%) in the Protocol group and 10 patients (2.72%) in the usual care group, corresponding on an absolute risk difference of approximately 1.9% (p=0.058). In the protocol group, there were no myocardial infarction or deaths at 1 year. In the usual care group, there were 2 (0.54%) myocardial infarctions, and no deaths at 1 year. Using multivariable logistic regression adjusting for age, sex, hypertension, diabetes mellitus, hyperlipidaemia and smoking, use of the chest pain protocol was not associated with higher odds of the composite 1-year cardiovascular outcome of CV hospitalisation, myocardial infarction or death (adjusted OR 0.43, 95% CI 0.11 to 1.42; p = 0.189).

Conclusion

In low-risk patients with stable chest pain, reduction of cardiac testing using chest pain protocol was not associated with an increase in adverse outcomes at 1 year, supporting the safety of risk-based test deferral.