ABSTRACT
Objective: Protocols to evaluate for myocardial infarction (MI) using high-sensitivity cardiac troponin (hs-cTn) have the potential to drive costs upward due to the added sensitivity. We performed an economic evaluation of an accelerated protocol (AP) to evaluate for MI using hs-cTn to identify changes in costs of treatment and length of stay compared with conventional testing. Methods: We performed a planned secondary economic analysis of a large, cluster randomized trial across nine emergency departments (EDs) from July 2020 to April 2021. Patients were included if they were 18 years or older with clinical suspicion for MI. In the AP, patients could be discharged without further testing at 0 h if they had a hs-cTnI < 4 ng/L and at 1 h if the initial value were 4 ng/L and the 1-h value ≤7 ng/L. Patients in the standard of care (SC) protocol used conventional cTn testing at 0 and 3 h. The primary outcome was the total cost of treatment, and the secondary outcome was ED length of stay. Results: Among 32,450 included patients, an AP had no significant differences in cost (+$89, CI: -$714, $893 hospital cost, +$362, CI: -$414, $1138 health system cost) or ED length of stay (+46, CI: -28, 120 min) compared with the SC protocol. In lower acuity, free-standing EDs, patients under the AP experienced shorter length of stay (-37 min, CI: -62, 12 min) and reduced health system cost (-$112, CI: -$250, $25). Conclusion: Overall, the implementation of AP using hs-cTn does not result in higher costs.
ABSTRACT
BACKGROUND AND OBJECTIVE: Clinical concern for acute coronary syndrome (ACS) is one of emergency medicine's most common patient encounters. This study aims to develop an ensemble learning-driven framework as a diagnostic support tool to prevent misdiagnosis. METHODS: We obtained extensive clinical electronic health data on patient encounters with clinical concerns for ACS from a large urban emergency department (ED) between January 2017 and August 2020. We applied an analytical framework equipped with many well-developed algorithms to improve the data quality by addressing missing values, dimensionality reduction, and data imbalance. We trained ensemble learning algorithms to classify patients with ACS or non-ACS etiologies of their symptoms. We used performance evaluation metrics such as accuracy, sensitivity, precision, F1-score, and the area under the receiver operating characteristic (AUROC) to measure the model's performance. RESULTS: The analysis included 31,228 patients, of whom 563 (1.8%) had ACS and 30,665 (98.2%) had alternative diagnoses. Eleven features, including systolic blood pressure, brain natriuretic peptide, chronic heart disease, coronary artery disease, creatinine, glucose, heart attack, heart rate, nephrotic syndrome, red cell distribution width, and troponin level, are reported as significantly contributing risk factors. The proposed framework successfully classifies these cohorts with sensitivity and AUROC as high as 86.3% and 93.3%. Our proposed model's accuracy, precision, specificity, Matthew's correlation coefficient, and F1-score were 85.7%, 86.3%, 93%, 80%, and 86.3%, respectively. CONCLUSION: Our proposed framework can identify early patients with ACS through further refinement and validation.
