Aim Assessment of chest pain patients often requires extensive investigations, while existing risk-scores are designed for use in Emergency Departments (ED). We aim to develop a risk score for chest pain patients in pre-hospital settings and ambulances by utilising heart rate variability (HRV), a rapid, non-invasive reflection of cardiac stress states.

Method This study analysed 902 chest pain cases presenting to Singapore General Hospital (SGH) ED between 2010 and 2015. The data was divided into a model derivation set (616 cases, 32% meeting outcomes) and a validation set (286 cases, 28.7% meeting outcomes). HRV Prediction Model (HRV-PM) for 30 day Major Adverse Cardiac Events (MACE) was derived using backward-stepwise logistic regression. HEART, TIMI and GRACE scores were used as comparators for the HRV-PM.

Results In the derivation set, 66.9% of patients were male, with a mean (SD) age of 60.71 (12.97). HRV-PM encompasses 9 parameters: 6 HRV variables (triangular interpolation of normal-to-normal intervals, total power, very-low/low/high-frequency power, approximate entropy), and 3 ECG variables (ST-elevation, ST-depression, Q-wave). Validation of HRV-PM shows similar performance to HEART for 30 day MACE prediction with Area Under Receiver-Operating-Curve (AUC) of 0.737 (95% CI: 0.673 to 0.800) versus 0.739 (95% CI: 0.679 to 0.799). The addition of 1 set of troponin results in HRV-PM improves the AUC to 0.749 (95% CI: 0.686 to 0.813), thereby outperforming HEART.

Conclusion HRV-PM is a simple, non-invasive test demonstrating comparable performance with HEART without need of blood-tests. We hope to validate and apply HRV-PM for assessment of chest pain patients by EMS in pre-hospital settings including ambulances.

Conflict of interest None

Funding None