A Novel Clinical Score for Differential Diagnosis Between Acute Myocarditis and Acute Coronary Syndrome – The SAlzburg MYocarditis (SAMY) Score

Abstract

BackgroundAcute myocarditis and acute coronary syndrome (ACS) are important differential diagnoses in patients with new-onset chest pain. To date, no clinical score exists to support the differentiation between these two diseases. The aim of this study was to develop such a score to aid the physician in scenarios where discrimination between myocarditis and ACS appears difficult.Materials and MethodsPatients with ACS (n = 233) and acute myocarditis (n = 123) were retrospectively enrolled. Least absolute shrinkage and selection operator (LASSO) regression was conducted to identify parameters associated with the highest or least probability for acute myocarditis. Logistic regression was conducted using the identified parameters and score points for each level of the predictors were calculated. Cutoffs for the prediction of myocarditis were calculated. Validation was conducted in a separate cohort of 90 patients.ResultsA score for prediction of acute myocarditis was calculated using six parameters [age, previous infection, hyperlipidemia, hypertension, C-reactive protein (CRP), and leukocyte count]. Logistic regression analysis showed a significant association between total score points and the presence of myocarditis (B = 0.9078, p < 0.0001). Cutoff #1 for the prediction of myocarditis was calculated at ≥ 4 (Sens.: 90.3%, Spec.: 93.1%; 46.3% predicted probability for acute myocarditis), cutoff #2 was calculated at ≥ 7 (Sens.: 73.1%, Spec.: > 99.9%; 92.9% pred. prob.). Validation showed good discrimination [area under the curve (AUC) = 0.935] and calibration of the score.ConclusionOur clinical score showed good discrimination and calibration for differentiating patients with acute myocarditis and ACS. Thus, it could support the differential diagnosis between these two disease entities and could facilitate clinical decisions in affected patients.

Establishment and validation of a prediction model for myocarditis in Chinese children below 14 years old: a protocol for a retrospective cohort study

“…To identify reliable predictors and address collinearity among independent variables, we will employ the LASSO regression analysis 24. As the outcome event pertains to myocarditis development in children, we will use multiple logistic regression analyses18 25–29 to construct an early prediction model for myocarditis. This model estimates the probability of myocarditis by assessing the child’s condition during a visit.…”