Reporting of coronavirus disease 2019 prognostic models: the transparent reporting of a multivariable prediction model for individual prognosis or diagnosis statement

作者：Liuqing Yang, Qiang Wang, Tingting Cui, Jinxin Huang, Naiyang Shi, Hui Jin · 发表于：Annals of Translational Medicine · 年份：2021 · DOI：10.21037/atm-20-6933 · 被引用次数：10 · 研究领域：COVID-19 Clinical Research Studies、COVID-19 and healthcare impacts、COVID-19 diagnosis using AI

Evaluation of the validity and applicability of published prognostic prediction models for coronavirus disease 2019 (COVID-19) is essential, because determining the patients' prognosis at an early stage may reduce mortality. This study was aimed to utilize the transparent reporting of a multivariable prediction model for individual prognosis or diagnosis (TRIPOD) to report the completeness of COVID-19-related prognostic models and appraise its effectiveness in clinical practice. A systematic search of the Web of Science and PubMed was performed for studies published until August 11, 2020. All models were assessed on model development, external validation of existing models, incremental values, and development and validation of the same model. TRIPOD was used to assess the completeness of included models, and the completeness of each item was also reported. In total, 52 publications were included, including 67 models. Age, disease history, lymphoma count, history of hypertension and cardiovascular disease, C-reactive protein, lactate dehydrogenase, white blood cell count, and platelet count were the commonly used predictors. The predicted outcome was death, development of severe or critical state, survival time, and length-of-hospital stay. The reported discrimination performance of all models ranged from 0.361 to 0.994, while few models reported calibration. Overall, the reporting completeness based on TRIPOD was between 31% and 83% [median, 67% (interquartile range: 62%, 73%...