Structured prehospital chest pain assessment and clinical diagnostic score for prehospital identification of ST-segment elevation myocardial infarction before an electrocardiogram

Abstract

<p>Background: In resource-limited settings where prehospital electrocardiogram (ECG) is not available, prehospital recognition of ST-segment elevation myocardial infarction (STEMI) remains challenging. This study aimed to evaluate the accuracy of a structured prehospital questionnaire and develop a clinical diagnostic score for prehospital STEMI identification. Methods: We analyzed 324 adult patients with chest pain transported by ambulance in the Hong Kong West Cluster from 1 February 2018 to 12 August 2018. Participants' symptoms were systematically assessed by the ambulance crew using the standardized 5-item prehospital chest pain (PHCP) questionnaire, which produced a composite score, and its diagnostic performance was evaluated. Univariate analysis followed by multivariable backward logistic regression was conducted to identify independent predictors of STEMI. A clinical diagnostic score, named the MANS score, was consequently developed based on the regression coefficients of these predictors and its discriminatory performance was compared with the PHCP score. Results: The PHCP score demonstrated suboptimal performance, with an area under the receiver operating characteristic curve (AUROC) of 0.63 (95% confidence interval [CI] 0.47–0.79). The MANS score (male sex, aspirin use, and nausea or vomiting and sweating) was derived from three independent predictors: male sex (1 point), aspirin use (−2 points), and the cooccurrence of nausea or vomiting and sweating (1 point). The score ranges from −2 to 2, with a higher score indicating a higher risk of STEMI. The MANS score achieved an AUROC of 0.85 (95% CI 0.79–0.91). Conclusions: While the PHCP questionnaire exhibited suboptimal diagnostic performance, the MANS score may help to identify STEMI in the absence of prehospital ECG. Further external validation is necessary to evaluate its generalizability.</p>