Dataset Information

Ruling In and Ruling Out COVID-19: Computing SARS-CoV-2 Infection Risk From Symptoms, Imaging and Test Data.

ABSTRACT: Background: Assigning meaningful probabilities of SARS-CoV-2 infection risk presents a diagnostic challenge across the continuum of care.

Objective: Develop and clinically validate an adaptable, personalized diagnostic model to assist clinicians in ruling-in and ruling-out COVID-19 in potential patients. This study compares the diagnostic performance of probabilistic, graphical, and machine-learning models against a previously published benchmark model.

Methods: We integrated patient symptom and test data using machine learning and Bayesian inference to quantify individual patient risk of SARS-CoV-2 infection. We trained models with 100,000 simulated patient profiles based on thirteen symptoms, estimated local prevalence, imaging, and molecular diagnostic performance from published reports. We tested these models with consecutive patients who presented with a COVID-19 compatible illness at the University of California San Diego Medical Center over 14 days starting in March 2020.

Results: We included 55 consecutive patients with fever (78%) or cough (77%) presenting for ambulatory (n=11) or hospital care (n=44). 51% (n=28) were female, 49% were age <60. Common comorbidities included diabetes (22%), hypertension (27%), cancer (16%) and cardiovascular disease (13%). 69% of these (n=38) were RT-PCR confirmed positive for SARS-CoV-2 infection, 11 had repeated negative nucleic acid testing and an alternate diagnosis. Bayesian inference network, distance metric-learning, and ensemble models discriminated between patients with SARS-CoV-2 infection and alternate diagnoses with sensitivities of 81.6 - 84.2%, specificities of 58.8 - 70.6%, and accuracies of 61.4 - 71.8%. After integrating imaging and laboratory test statistics with the predictions of the Bayesian inference network, changes in diagnostic uncertainty at each step in the simulated clinical evaluation process were highly sensitive to location, symptom, and diagnostic test choices.

Conclusions: Decision support models that incorporate symptoms and available test results can help providers diagnose SARS-CoV-2 infection in real world settings.

Clinicaltrial:

SUBMITTER: D'Ambrosia C

PROVIDER: S-EPMC7746395 | biostudies-literature | 2020 Nov

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

Computing SARS-CoV-2 Infection Risk From Symptoms, Imaging, and Test Data: Diagnostic Model Development.

D'Ambrosia Christopher C Christensen Henrik H Aronoff-Spencer Eliah E

Journal of medical Internet research 20201216 12

<h4>Background</h4>Assigning meaningful probabilities of SARS-CoV-2 infection risk presents a diagnostic challenge across the continuum of care.<h4>Objective</h4>The aim of this study was to develop and clinically validate an adaptable, personalized diagnostic model to assist clinicians in ruling in and ruling out COVID-19 in potential patients. We compared the diagnostic performance of probabilistic, graphical, and machine learning models against a previously published benchmark model.<h4>Metho ...[more]

PMID: 33301417

Similar Datasets

Project description:PurposeTo evaluate the diagnostic value of symptoms used by daycares and schools to screen children and adolescents for SARS-CoV-2 infection, we analyzed data from a primary care setting.MethodsThis cohort study included all patients ≤17 years old who were evaluated at Providence Community Health Centers (PCHC; Providence, U.S.), for COVID-19 symptoms and/or exposure, and received SARS-CoV-2 polymerase chain reaction (PCR) testing between March-June 2020. Participants were identified from PCHC electronic medical records. For three age groups- 0-4, 5-11, and 12-17 years-we estimated the sensitivity, specificity, and area under the receiver operating curve (AUC) of individual symptoms and three symptom combinations: a case definition published by the Rhode Island Department of Health (RIDOH), and two novel combinations generated by different statistical approaches to maximize sensitivity, specificity, and AUC. We evaluated symptom combinations both with and without consideration of COVID-19 exposure. Myalgia, headache, sore throat, abdominal pain, nausea, anosmia, and ageusia were not assessed in 0-4 year-olds due to the lower reliability of these symptoms in this group.ResultsOf 555 participants, 217 (39.1%) were SARS-CoV-2-infected. Fever was more common among 0-4 years-olds (p = 0.002); older children more frequently reported fatigue (p = 0.02). In children ≥5 years old, anosmia or ageusia had 94-98% specificity. In all ages, exposure history most accurately predicted infection. With respect to individual symptoms, cough most accurately predicted infection in <5 year-olds (AUC 0.69) and 12-17 year-olds (AUC 0.62), while headache was most accurate in 5-11 year-olds (AUC 0.62). In combination with exposure history, the novel symptom combinations generated statistically to maximize test characteristics had sensitivity >95% but specificity <30%. No symptom or symptom combination had AUC ≥0.70.ConclusionsAnosmia or ageusia in children ≥5 years old should raise providers' index of suspicion for COVID-19. However, our overall findings underscore the limited diagnostic value of symptoms.

Dataset Information

Ruling In and Ruling Out COVID-19: Computing SARS-CoV-2 Infection Risk From Symptoms, Imaging and Test Data.

Publications

Computing SARS-CoV-2 Infection Risk From Symptoms, Imaging, and Test Data: Diagnostic Model Development.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets