skip to main content

Explore Research Products in the NSF-PAR

Title: Predictive Models of Mortality for Hospitalized Patients With COVID-19: Retrospective Cohort Study
Background The novel coronavirus SARS-CoV-2 and its associated disease, COVID-19, have caused worldwide disruption, leading countries to take drastic measures to address the progression of the disease. As SARS-CoV-2 continues to spread, hospitals are struggling to allocate resources to patients who are most at risk. In this context, it has become important to develop models that can accurately predict the severity of infection of hospitalized patients to help guide triage, planning, and resource allocation. Objective The aim of this study was to develop accurate models to predict the mortality of hospitalized patients with COVID-19 using basic demographics and easily obtainable laboratory data. Methods We performed a retrospective study of 375 hospitalized patients with COVID-19 in Wuhan, China. The patients were randomly split into derivation and validation cohorts. Regularized logistic regression and support vector machine classifiers were trained on the derivation cohort, and accuracy metrics (F1 scores) were computed on the validation cohort. Two types of models were developed: the first type used laboratory findings from the entire length of the patient’s hospital stay, and the second type used laboratory findings that were obtained no later than 12 hours after admission. The models were further validated on a multicenter external cohort of 542 patients. Results Of the 375 patients with COVID-19, 174 (46.4%) died of the infection. The study cohort was composed of 224/375 men (59.7%) and 151/375 women (40.3%), with a mean age of 58.83 years (SD 16.46). The models developed using data from throughout the patients’ length of stay demonstrated accuracies as high as 97%, whereas the models with admission laboratory variables possessed accuracies of up to 93%. The latter models predicted patient outcomes an average of 11.5 days in advance. Key variables such as lactate dehydrogenase, high-sensitivity C-reactive protein, and percentage of lymphocytes in the blood were indicated by the models. In line with previous studies, age was also found to be an important variable in predicting mortality. In particular, the mean age of patients who survived COVID-19 infection (50.23 years, SD 15.02) was significantly lower than the mean age of patients who died of the infection (68.75 years, SD 11.83; P<.001). Conclusions Machine learning models can be successfully employed to accurately predict outcomes of patients with COVID-19. Our models achieved high accuracies and could predict outcomes more than one week in advance; this promising result suggests that these models can be highly useful for resource allocation in hospitals.  more » « less
Award ID(s):
1914792 1645681 1664644 2114393
NSF-PAR ID:
10220302
Author(s) / Creator(s):
; ; ; ; ;
Date Published:
Journal Name:
JMIR Medical Informatics
Volume:
8
Issue:
10
ISSN:
2291-9694
Page Range / eLocation ID:
e21788
Format(s):
Medium: X
Sponsoring Org:
National Science Foundation
More Like this
  1. ; ; ; ; ; ; ; ; ( , eLife)
    null (Ed.)
    The new coronavirus (now named SARS-CoV-2) causing the disease pandemic in 2019 (COVID-19), has so far infected over 35 million people worldwide and killed more than 1 million. Most people with COVID-19 have no symptoms or only mild symptoms. But some become seriously ill and need hospitalization. The sickest are admitted to an Intensive Care Unit (ICU) and may need mechanical ventilation to help them breath. Being able to predict which patients with COVID-19 will become severely ill could help hospitals around the world manage the huge influx of patients caused by the pandemic and save lives. Now, Hao, Sotudian, Wang, Xu et al. show that computer models using artificial intelligence technology can help predict which COVID-19 patients will be hospitalized, admitted to the ICU, or need mechanical ventilation. Using data of 2,566 COVID-19 patients from five Massachusetts hospitals, Hao et al. created three separate models that can predict hospitalization, ICU admission, and the need for mechanical ventilation with more than 86% accuracy, based on patient characteristics, clinical symptoms, laboratory results and chest x-rays. Hao et al. found that the patients’ vital signs, age, obesity, difficulty breathing, and underlying diseases like diabetes, were the strongest predictors of the need for hospitalization. Being male, having diabetes, cloudy chest x-rays, and certain laboratory results were the most important risk factors for intensive care treatment and mechanical ventilation. Laboratory results suggesting tissue damage, severe inflammation or oxygen deprivation in the body's tissues were important warning signs of severe disease. The results provide a more detailed picture of the patients who are likely to suffer from severe forms of COVID-19. Using the predictive models may help physicians identify patients who appear okay but need closer monitoring and more aggressive treatment. The models may also help policy makers decide who needs workplace accommodations such as being allowed to work from home, which individuals may benefit from more frequent testing, and who should be prioritized for vaccination when a vaccine becomes available. 
    more » « less
  2. ; ; ; ; ; ; ( , Lancet infections disease)
    Background: Carbapenem-resistant Enterobacteriaceae (CRE) are a global threat. Here, we describe the clinical and molecular characteristics of Centers for Disease Control and Prevention (CDC)-defined CRE in the US. Methods: The second Consortium on Resistance Against Carbapenems in Klebsiella and other Enterobacteriaceae (CRACKLE-2, ClinicalTrials.gov: NCT03646227) is a prospective, multicenter, cohort study. Patients hospitalized in 49 US hospitals, with clinical cultures positive for CDC-defined CRE between 30 April 2016 and 31 August 2017 were included. Primary outcome was desirability of outcome ranking (DOOR) at 30 days. Clinical data and bacteria were collected, and whole genome sequencing (WGS) was performed. Findings: 1,040 patients with unique isolates were included; 449 (43%) with infection and 591 (57%) with colonization. CDC-defined CRE admission rate was 57 CDC-defined CRE admissions/100,000 admissions (95% CI: 45–71). Three subsets of CDC-defined CRE were identified: carbapenemase-producing Enterobacteriaceae (618/1,040, 59%); non-carbapenemase-producing CRE (194/1,040, 19%); and unconfirmed CRE (228/1,040, 22%; initially reported as CRE, but susceptible to carbapenems in two central laboratories). Klebsiella pneumoniae carbapenemase (KPC)-producing clonal group 258 K. pneumoniae was the most common carbapenemase-producing Enterobacteriaceae. In 449 patients with CDC-defined CRE infections, DOOR outcomes were not significantly different in patients with carbapenemase-producing Enterobacteriaceae, non-carbapenemase-producing CRE, and unconfirmed CRE. At 30 days 107/449 (24%, 95% CI 20–28%) patients had died. Interpretation: Among patients with CDC-defined CRE, similar outcomes were observed among three subgroups, including the novel unconfirmed CRE group. CDC-defined CRE represent diverse bacteria, whose spread may not respond to interventions directed to carbapenemase-producing Enterobacteriaceae. 
    more » « less
  3. ; ; ; ; ( , BMC Infectious Diseases)
    Abstract Background

    The Mexican Institute of Social Security (IMSS) is the largest health care provider in Mexico, covering about 48% of the Mexican population. In this report, we describe the epidemiological patterns related to confirmed cases, hospitalizations, intubations, and in-hospital mortality due to COVID-19 and associated factors, during five epidemic waves recorded in the IMSS surveillance system.

     Methods

    We analyzed COVID-19 laboratory-confirmed cases from the Online Epidemiological Surveillance System (SINOLAVE) from March 29th, 2020, to August 27th, 2022. We constructed weekly epidemic curves describing temporal patterns of confirmed cases and hospitalizations by age, gender, and wave. We also estimated hospitalization, intubation, and hospital case fatality rates. The mean days of in-hospital stay and hospital admission delay were calculated across five pandemic waves. Logistic regression models were employed to assess the association between demographic factors, comorbidities, wave, and vaccination and the risk of severe disease and in-hospital death.

     Results

    A total of 3,396,375 laboratory-confirmed COVID-19 cases were recorded across the five waves. The introduction of rapid antigen testing at the end of 2020 increased detection and modified epidemiological estimates. Overall, 11% (95% CI 10.9, 11.1) of confirmed cases were hospitalized, 20.6% (95% CI 20.5, 20.7) of the hospitalized cases were intubated, and the hospital case fatality rate was 45.1% (95% CI 44.9, 45.3). The mean in-hospital stay was 9.11 days, and patients were admitted on average 5.07 days after symptoms onset. The most recent waves dominated by the Omicron variant had the highest incidence. Hospitalization, intubation, and mean hospitalization days decreased during subsequent waves. The in-hospital case fatality rate fluctuated across waves, reaching its highest value during the second wave in winter 2020. A notable decrease in hospitalization was observed primarily among individuals ≥ 60 years. The risk of severe disease and death was positively associated with comorbidities, age, and male gender; and declined with later waves and vaccination status.

     Conclusion

    During the five pandemic waves, we observed an increase in the number of cases and a reduction in severity metrics. During the first three waves, the high in-hospital fatality rate was associated with hospitalization practices for critical patients with comorbidities.

     
    more » « less
  4. ; ; ; ; ; ; ; ; ( , JAMA Network Open)
    Importance

    The frequent occurrence of cognitive symptoms in post–COVID-19 condition has been described, but the nature of these symptoms and their demographic and functional factors are not well characterized in generalizable populations.

     Objective

    To investigate the prevalence of self-reported cognitive symptoms in post–COVID-19 condition, in comparison with individuals with prior acute SARS-CoV-2 infection who did not develop post–COVID-19 condition, and their association with other individual features, including depressive symptoms and functional status.

     Design, Setting, and Participants

    Two waves of a 50-state nonprobability population-based internet survey conducted between December 22, 2022, and May 5, 2023. Participants included survey respondents aged 18 years and older.

     Exposure

    Post–COVID-19 condition, defined as self-report of symptoms attributed to COVID-19 beyond 2 months after the initial month of illness.

     Main Outcomes and Measures

    Seven items from the Neuro-QoL cognition battery assessing the frequency of cognitive symptoms in the past week and patient Health Questionnaire-9.

     Results

    The 14 767 individuals reporting test-confirmed COVID-19 illness at least 2 months before the survey had a mean (SD) age of 44.6 (16.3) years; 568 (3.8%) were Asian, 1484 (10.0%) were Black, 1408 (9.5%) were Hispanic, and 10 811 (73.2%) were White. A total of 10 037 respondents (68.0%) were women and 4730 (32.0%) were men. Of the 1683 individuals reporting post–COVID-19 condition, 955 (56.7%) reported at least 1 cognitive symptom experienced daily, compared with 3552 of 13 084 (27.1%) of those who did not report post–COVID-19 condition. More daily cognitive symptoms were associated with a greater likelihood of reporting at least moderate interference with functioning (unadjusted odds ratio [OR], 1.31 [95% CI, 1.25-1.36]; adjusted [AOR], 1.30 [95% CI, 1.25-1.36]), lesser likelihood of full-time employment (unadjusted OR, 0.95 [95% CI, 0.91-0.99]; AOR, 0.92 [95% CI, 0.88-0.96]) and greater severity of depressive symptoms (unadjusted coefficient, 1.40 [95% CI, 1.29-1.51]; adjusted coefficient 1.27 [95% CI, 1.17-1.38). After including depressive symptoms in regression models, associations were also found between cognitive symptoms and at least moderate interference with everyday functioning (AOR, 1.27 [95% CI, 1.21-1.33]) and between cognitive symptoms and lower odds of full-time employment (AOR, 0.92 [95% CI, 0.88-0.97]).

     Conclusions and Relevance

    The findings of this survey study of US adults suggest that cognitive symptoms are common among individuals with post–COVID-19 condition and associated with greater self-reported functional impairment, lesser likelihood of full-time employment, and greater depressive symptom severity. Screening for and addressing cognitive symptoms is an important component of the public health response to post–COVID-19 condition.

     
    more » « less
  5. ; ; ; ; ; ; ; ( , Infection Control & Hospital Epidemiology)
    Abstract Objective: Current guidance states that asymptomatic screening for severe acute respiratory coronavirus virus 2 (SARS-CoV-2) prior to admission to an acute-care setting is at the facility’s discretion. This study’s objective was to estimate the number of undetected cases of SARS-CoV-2 admitted as inpatients under 4 testing approaches and varying assumptions. Design and setting: Individual-based microsimulation of 104 North Carolina acute-care hospitals Patients: All simulated inpatient admissions to acute-care hospitals from December 15, 2021, to January 13, 2022 [ie, during the SARS-COV-2 ο (omicron) variant surge]. Interventions: We simulated (1) only testing symptomatic patients, (2) 1-stage antigen testing with no confirmatory polymerase chain reaction (PCR) test, (3) 1-stage antigen testing with a confirmatory PCR for negative results, and (4) serial antigen screening (ie, repeat antigen test 2 days after a negative result). Results: Over 1 month, there were 77,980 admissions: 13.7% for COVID-19, 4.3% with but not for COVID-19, and 82.0% for non–COVID-19 indications without current infection. Without asymptomatic screening, 1,089 (credible interval [CI], 946–1,253) total SARS-CoV-2 infections (7.72%) went undetected. With 1-stage antigen screening, 734 (CI, 638–845) asymptomatic infections (67.4%) were detected, with 1,277 false positives. With combined antigen and PCR screening, 1,007 (CI, 875–1,159) asymptomatic infections (92.5%) were detected, with 5,578 false positives. A serial antigen testing policy detected 973 (CI, 845–1,120) asymptomatic infections (89.4%), with 2,529 false positives. Conclusions: Serial antigen testing identified >85% of asymptomatic infections and resulted in fewer false positives with less cost per identified infection compared to combined antigen plus PCR testing. 
    more » « less
  • Citation Formats
  • MLA
  • APA
  • Chicago
  • BibTeX
  • Save / Share this Record
  • Send to Email