Table 1.

Examples of machine learning studies that aim to predict various outcomes for in-patients with confirmed COVID-19 diagnosis. We refer the readers to extensive published literature reviews [2,23,24].

Reference	Outcome	Input Data	Models	Study Location
[29]	Deterioration (intubation or ICU admission or mortality)	Chest X-ray images and clinical data (patient demographics, seven vital-sign variables, and 24 laboratory-test results)	Convolutional neural network for chest X-ray images and gradient boosting model for clinical data	United States
[30]	Mortality	Five laboratory-test results	Support vector machine	United States
[31]	Severe progression (high oxygen flow rate, mechanical ventiliation or mortality)	Chest CT scans, patient demographics, five vital-sign variables, symptoms, comorbidities, 14 laboratory-test results, and chest CT radiology report findings	Deep neural network and logistic regression	France
[32]	Prognostication (intubation or hospital admission, or mortality)	Chest X-ray images, two vital-sign variables, and nine laboratory-test results	Convolutional neural network	United States
[28]	Sepsis	Eight laboratory-test results	Gradient boosting model	China
[25]	AKI	Findings of abdominal CT scans, demographics, vital signs, comorbidities, and three laboratory-test results	Logistic regression	United States
[33]	ARDS	Demgraphics, interventions, comorbidities, 17 laboratory-test results, and eight vital signs	Gradient boosting model	United States