Enhancing predictive accuracy of the 13-item Acute Coronary Syndrome checklist: a novel approach to improving risk assessment and diagnosis

This study aimed to evaluate the discriminatory capacity of the 13-Item ACS checklist and improve the accuracy of ACS diagnosis through the application of weighted regression analysis. This predictive correlation study enrolled 300 patients admitted to Emergency Department between February 2021 and January 2022. The ACS checklist was administered upon initial triage, followed by patient tracking over a one-month hospitalisation period, capturing ACS diagnoses. Data analysis employed STATA 17 and MEDCALC 20.0.13 software. Findings indicated that patients with sweating and shortness of breath symptoms had a heightened likelihood of true ACS diagnosis by 14% and 11%, respectively, compared to those without ACS (p = 0.005 and 0.019). Conversely, palpitations were associated with a 20% decreased likelihood of authentic ACS diagnosis (p < 0.001). Integration of significant regression coefficients - palpitation severity (-21), sweating severity (13.7), and shortness of breath severity (11) demonstrated significant discriminatory enhancements in the checklists. The weighted 13-item ACS checklist surpassed the unweighted version's performance, yielding superior discriminatory power for ACS diagnosis (p < 0.001 and p = 0.089). The weighted checklist elevated the AUC score from 55% to 70%. Incorporating weighted factors - shortness of breath severity, sweating severity, and palpitations severity - into the checklist notably enhanced ACS identification. However, it's important to note that this tool, while showing promise, is not intended to serve as a standalone diagnostic tool for ACS. Instead, this tool has the potential to enhance risk assessment and aid in clinical decision-making.