First, Do No Harm: Predictive Analytics to Reduce In-Hospital Adverse Events
遵循设计科学范式，开发了一种名为SALT的新型院内不良事件预测模型，该模型整合了广义线性混合模型、多任务学习和随机时间序列过程，实证表明其预测性能优于现有技术，并能显著节约医疗成本。
Inadequate patient safety is a serious issue in current medical practice. Medical errors cause adverse events (AEs) for patients and lead to premature deaths, unintended complications, prolonged hospital stays, and higher medical costs. Although the importance of AE prediction and prevention is well recognized in the information systems literature, there is a dearth of research on modeling and predicting AEs caused by medical errors. Following the design science research paradigm, this study describes the search, design, and evaluation of a novel in-hospital AE prediction model, called Stochastic Autoregressions for Latent Trajectories (SALT). The proposed model uniquely integrates generalized linear mixed model with multitask learning and stochastic time-series processes. Results from our empirical evaluation show that SALT outperforms prior state-of-the-art techniques in predicting AEs during patients’ hospital stays. Through a simulation, we further demonstrate significant cost savings potential when hospitals implement and integrate SALT in their inpatient care. This study contributes to the design science literature by formalizing the in-hospital AE prediction problem, on the one hand, and developing a novel graphical model to address the prediction problem, on the other. For healthcare practitioners and administrators, our predictive analytics approach unveils important insights to minimize AEs.