Abstract 4138211: A Machine Learning-Based Novel Risk Score Model for Takotsubo Cardiomyopathy

Background: Takotsubo cardiomyopathy (TC) is defined by acute but typically reversible left ventricular systolic dysfunction, typically precipitated by profound emotional or physical stress. Due to the scarcity of a prognostic scoring system for TC, we aimed to develop a novel machine learning-based risk score model to predict in-hospital mortality in TC patients. Methods: The National Inpatient Sample (NIS) 2016-2020 database was used to identify all adult patients (≥18 years of age) with TC, using ICD-10 code I51.81. Relevant clinical characteristics and outcomes were extracted. The primary endpoint was in-hospital all-cause mortality. The dataset was randomly split into 3 subsets: training, validation, and testing, with a ratio of 0.7, 0.2, and 0.1, respectively. The risk score was generated using the Autoscore package in R software, a machine learning-based tool for automatic clinical score generation. The performance was evaluated using the area under the receiver-operative characteristics curve (AUC) with 95% confidence intervals (95%CI). Results: Among the 38,662 TC patients identified [age 67.1±14.1 years, female 32,089 (83%)], there were 2499 (6.4%) primary outcome events. Out of the top 20 covariates seen in the Parsimony plot, we used eight for our risk score (0–127): age, comorbidity burden, cardiac arrest, cardiogenic shock, race, acute kidney injury, hypertension, and cardiac arrhythmia (Figure 1A). The AUC for the derivation and validation cohorts was 0.809 (0.780–0.837) and 0.838 (0.820–0.856), respectively (Figures 1B and 1C). Conclusion: We developed and internally validated a machine learning based-risk score model to predict outcomes in TC patients, achieving a high AUC of 0.838 (0.820-0.856). Risk score models guide risk stratification and management, and large future studies are needed to externally evaluate our results.