Identification of high-risk COVID-19 patients using machine learning

Abstract

The current COVID-19 public health crisis, caused by SARS-CoV-2 (severe acute respiratory syndrome coronavirus 2), has produced a devastating toll both in terms of human life loss and economic disruption. In this paper we present a machine-learning algorithm capable of identifying whether a given patient (actually infected or suspected to be infected) is more likely to survive than to die, or vice-versa. We train this algorithm with historical data, including medical history, demographic data, as well as COVID-19-related information. This is extracted from a database of confirmed and suspected COVID-19 infections in Mexico, constituting the official COVID-19 data compiled and made publicly available by the Mexican Federal Government. We demonstrate that the proposed method can detect high-risk patients with high accuracy, in each of four identified clinical stages, thus improving hospital capacity planning and timely treatment. Furthermore, we show that our method can be extended to provide optimal estimators for hypothesis-testing techniques commonly-used in biological and medical statistics. We believe that our work could be of use in the context of the current pandemic in assisting medical professionals with real-time assessments so as to determine health care priorities.

A Pilot Study of the Efficacy and Economical Sustainability of Acute Coronavirus Disease 2019 Patient Management in an Outpatient Setting

“…Here we describe an innovative healthcare strategy to optimize the management system of COVID-19 patients while sparing resources for the care of patients with non-COVID-19-related conditions. Similar models have previously been proposed ( 18 , 19 ). MMCOs were designed to fill the gap of care delivery to COVID-19 patients with clinical features that are neither too mild to be managed by the GP in a home-based setting nor too severe to require ED admission or hospitalization.…”

Long-COVID diagnosis: From diagnostic to advanced AI-driven models

“…A deep learning model achieved high sensitivity (90%) and high specificity (96%)in the detection of COVID-19 using chest CT with an area under the curve of 0.96 and an average time for each CT scan of 4,51 s [53] . In addition, an AI algorithm can identify patients who developed severe COVID-19 symptoms [55] , [56] . Quiroz-Juarez has presented a study for the early identification of high-risk patients among those exposed to the SARS-COV-2 virus, using a supervised artificial neural network [55] .…”

“…In addition, an AI algorithm can identify patients who developed severe COVID-19 symptoms [55] , [56] . Quiroz-Juarez has presented a study for the early identification of high-risk patients among those exposed to the SARS-COV-2 virus, using a supervised artificial neural network [55] . The machine learning models were trained using comorbidities, patient demographic data, as well as recent COVID-19-related medical information.…”

“…The machine learning models were trained using comorbidities, patient demographic data, as well as recent COVID-19-related medical information. The authors reported that the disease outcome can be predicted with specificity greater than 82%, sensitivity greater than 86%, and accuracy greater than 84% [55] .…”

Clinical Profiles at the Time of Diagnosis of SARS-CoV-2 Infection in Costa Rica During the Pre-vaccination Period Using a Machine Learning Approach

“…To overcome this situation, clustering or unsupervised machine learning approaches bring an opportunity to extract relevant information by identifying patterns, clusters, or profiles within large volumes of data. Although some machine learning or similar approaches have been implemented to investigate clinical symptoms, risk factors, and other parameters related to COVID-19 (Dixon et al 2021 ; Han et al 2020 ; Sudre et al 2021 ; Tong et al 2020 ; Kim et al 2020 ; Fu et al 2020 ; Alballa and Al-Turaiki 2021 ; Oshinubi et al 2021 , 2022 ; Quiroz-Juárez et al 2021 ; Zoabi et al 2021 ; Li et al 2020 ), to our knowledge, none has been formally reported from Costa Rican cases.…”