A CLINICAL PREDICTION MODEL FOR SHORT-TERM PROGNOSIS IN PATIENTS WITH NON-ACUTE MYOCARDIAL INFARCTION-RELATED CARDIOGENIC SHOCK.

Abstract

Abstract: Background : While acute myocardial infarction (AMI) is widely recognized as the primary cause of cardiogenic shock (CS), non-AMI-related CS has been excluded from the majority of CS studies. Information on its prognostic factors remains largely understudied, and it is necessary to focus on these patients to identify the specific risk factors. In this study, we aimed to build and validate a predictive nomogram and risk classification system. Methods: 1298 patients and 548 patients with CS from the Medical Information Mart for Intensive Care IV and III databases were included in the study after excluding patients with AMI. Lasso and logistic regression analysis were used to identify statistically significant predictors, which were finally involved in the nomogram. The predictive performance of the nomogram was validated by calibration plots and was compared with other scoring systems by area under curve and decision curve analysis curves. Results: Age, heart rate, white blood cell count, albumin level, lactic acid level, GCS score, 24-h urine output, and vasopressor use were identified as the most critical factors for in-hospital death. Based on these results, a nomogram was established for predicting in-hospital mortality. The area under curve value of the nomogram was 0.806 in the training group and 0.814 and 0.730 in the internal and external validation sets, respectively, which were significantly higher than those of other commonly used intensive care unit scoring systems (Simplified Acute Physiology Score II, Acute Physiology Score III, and Sequential Organ Failure Assessment). In addition, the survival curve showed significant differences in the 30-day survival of the three risk subgroups divided by the nomogram. Conclusion: For non-AMI-associated CS, a predictive nomogram and risk classification system were developed and validated, and the nomogram demonstrated good performance in prognostic prediction and risk stratification.