Abstract: Objective To systematically evaluate the risk prediction model for deep vein thrombosis in hospitalized patients based on machine learning. Methods We conducted literature research in PubMed, Embase, CHINHAL, Cochrane Library, Web of Science, CNKI, and Wanfang databases for literature on risk prediction models for deep vein thrombosis in hospitalized patients constructed by machine learning. The search period spanned from the inception to March 2023. Two researchers completed literature screening and data extraction independently, and used predictive models to construct a research data extraction and quality evaluation checklist (CHARMS) to evaluate the quality of the included literature and screened high-quality literature for discussion. Results Totally 11 high-quality studies were collected, including 28 machine learning models, with an area under the ROC curve ranging from 0.710 to 0.976. Laboratory indicators such as age, VTE history, length of hospital stay, medication history, and D-dimer were are the main predictive factors. Conclusions Risk prediction models constructed using machine learning can accurately identify the risk of DVT events in hospitalized patients, and its predictive performance is superior to traditional risk prediction models. The available literature on the topic exhibits a low overall risk of bias, however, the applicability level of the prediction model is considered average.

Key words: Deep vein thrombosis, DVT, machine learning, nursing, predictive model, systematic review