Potential solutions for screening, triage, and severity scoring of suspected COVID-19 positive patients in low-resource settings: a scoping review

COVID-19 screening in low resource settings using artificial intelligence for chest radiographs and point-of-care blood tests

Artificial intelligence (AI) systems for detection of COVID-19 using chest X-Ray (CXR) imaging and point-of-care blood tests were applied to data from four low resource African settings. The performance of these systems to detect COVID-19 using various input data was analysed and compared with antigen-based rapid diagnostic tests. Participants were tested using the gold standard of RT-PCR test (nasopharyngeal swab) to determine whether they were infected with SARS-CoV-2. A total of 3737 (260 RT-PCR positive) participants were included. In our cohort, AI for CXR images was a poor predictor of COVID-19 (AUC = 0.60), since the majority of positive cases had mild symptoms and no visible pneumonia in the lungs. AI systems using differential white blood cell counts (WBC), or a combination of WBC and C-Reactive Protein (CRP) both achieved an AUC of 0.74 with a suggested optimal cut-off point at 83% sensitivity and 63% specificity. The antigen-RDT tests in this trial obtained 65% sensitivity at 98% specificity. This study is the first to validate AI tools for COVID-19 detection in an African setting. It demonstrates that screening for COVID-19 using AI with point-of-care blood tests is feasible and can operate at a higher sensitivity level than antigen testing.

Training and testing of a gradient boosted machine learning model to predict adverse outcome in patients presenting to emergency departments with suspected covid-19 infection in a middle-income setting

COVID-19 infection rates remain high in South Africa. Clinical prediction models may be helpful for rapid triage, and supporting clinical decision making, for patients with suspected COVID-19 infection. The Western Cape, South Africa, has integrated electronic health care data facilitating large-scale linked routine datasets. The aim of this study was to develop a machine learning model to predict adverse outcome in patients presenting with suspected COVID-19 suitable for use in a middle-income setting. A retrospective cohort study was conducted using linked, routine data, from patients presenting with suspected COVID-19 infection to public-sector emergency departments (EDs) in the Western Cape, South Africa between 27th August 2020 and 31st October 2021. The primary outcome was death or critical care admission at 30 days. An XGBoost machine learning model was trained and internally tested using split-sample validation. External validation was performed in 3 test cohorts: Western Cape patients presenting during the Omicron COVID-19 wave, a UK cohort during the ancestral COVID-19 wave, and a Sudanese cohort during ancestral and Eta waves. A total of 282,051 cases were included in a complete case training dataset. The prevalence of 30-day adverse outcome was 4.0%. The most important features for predicting adverse outcome were the requirement for supplemental oxygen, peripheral oxygen saturations, level of consciousness and age. Internal validation using split-sample test data revealed excellent discrimination (C-statistic 0.91, 95% CI 0.90 to 0.91) and calibration (CITL of 1.05). The model achieved C-statistics of 0.84 (95% CI 0.84 to 0.85), 0.72 (95% CI 0.71 to 0.73), and 0.62, (95% CI 0.59 to 0.65) in the Omicron, UK, and Sudanese test cohorts. Results were materially unchanged in sensitivity analyses examining missing data. An XGBoost machine learning model achieved good discrimination and calibration in prediction of adverse outcome in patients presenting with suspected COVID19 to Western Cape EDs. Performance was reduced in temporal and geographical external validation.

External validation of triage tools for adults with suspected COVID-19 in a middle-income setting: an observational cohort study

BACKGROUND

Tools proposed to triage ED acuity in suspected COVID-19 were derived and validated in higher income settings during early waves of the pandemic. We estimated the accuracy of seven risk-stratification tools recommended to predict severe illness in the Western Cape, South Africa.

METHODS

An observational cohort study using routinely collected data from EDs across the Western Cape, from 27 August 2020 to 11 March 2022, was conducted to assess the performance of the PRIEST (Pandemic Respiratory Infection Emergency System Triage) tool, NEWS2 (National Early Warning Score, version 2), TEWS (Triage Early Warning Score), the WHO algorithm, CRB-65, Quick COVID-19 Severity Index and PMEWS (Pandemic Medical Early Warning Score) in suspected COVID-19. The primary outcome was intubation or non-invasive ventilation, death or intensive care unit admission at 30 days.

RESULTS

Of the 446 084 patients, 15 397 (3.45%, 95% CI 34% to 35.1%) experienced the primary outcome. Clinical decision-making for inpatient admission achieved a sensitivity of 0.77 (95% CI 0.76 to 0.78), specificity of 0.88 (95% CI 0.87 to 0.88) and the negative predictive value (NPV) of 0.99 (95% CI 0.99 to 0.99). NEWS2, PMEWS and PRIEST scores achieved good estimated discrimination (C-statistic 0.79 to 0.82) and identified patients at risk of adverse outcomes at recommended cut-offs with moderate sensitivity (>0.8) and specificity ranging from 0.41 to 0.64. Use of the tools at recommended thresholds would have more than doubled admissions, with only a 0.01% reduction in false negative triage.

CONCLUSION

No risk score outperformed existing clinical decision-making in determining the need for inpatient admission based on prediction of the primary outcome in this setting. Use of the PRIEST score at a threshold of one point higher than the previously recommended best approximated existing clinical accuracy.

LMIC-PRIEST: Derivation and validation of a clinical severity score for acutely ill adults with suspected COVID-19 in a middle-income setting

BACKGROUND

Uneven vaccination and less resilient health care systems mean hospitals in LMICs are at risk of being overwhelmed during periods of increased COVID-19 infection. Risk-scores proposed for rapid triage of need for admission from the emergency department (ED) have been developed in higher-income settings during initial waves of the pandemic.

METHODS

Routinely collected data for public hospitals in the Western Cape, South Africa from the 27th August 2020 to 11th March 2022 were used to derive a cohort of 446,084 ED patients with suspected COVID-19. The primary outcome was death or ICU admission at 30 days. The cohort was divided into derivation and Omicron variant validation sets. We developed the LMIC-PRIEST score based on the coefficients from multivariable analysis in the derivation cohort and existing triage practices. We externally validated accuracy in the Omicron period and a UK cohort.

RESULTS

We analysed 305,564 derivation, 140,520 Omicron and 12,610 UK validation cases. Over 100 events per predictor parameter were modelled. Multivariable analyses identified eight predictor variables retained across models. We used these findings and clinical judgement to develop a score based on South African Triage Early Warning Scores and also included age, sex, oxygen saturation, inspired oxygen, diabetes and heart disease. The LMIC-PRIEST score achieved C-statistics: 0.82 (95% CI: 0.82 to 0.83) development cohort; 0.79 (95% CI: 0.78 to 0.80) Omicron cohort; and 0.79 (95% CI: 0.79 to 0.80) UK cohort. Differences in prevalence of outcomes led to imperfect calibration in external validation. However, use of the score at thresholds of three or less would allow identification of very low-risk patients (NPV ≥0.99) who could be rapidly discharged using information collected at initial assessment.

CONCLUSION

The LMIC-PRIEST score shows good discrimination and high sensitivity at lower thresholds and can be used to rapidly identify low-risk patients in LMIC ED settings.

LMIC-PRIEST: Derivation and validation of a clinical severity score for acutely ill adults with suspected COVID-19 in a middle-income setting

Background

Uneven vaccination and less resilient health care systems mean hospitals in LMICs are at risk of being overwhelmed during periods of increased COVID-19 infection. Risk-scores proposed for rapid triage of need for admission from the emergency department (ED) have been developed in higher-income settings during initial waves of the pandemic.

Methods

Routinely collected data for public hospitals in the Western Cape, South Africa from the 27 th August 2020 to 11 th March 2022 were used to derive a cohort of 446,084 ED patients with suspected COVID-19. The primary outcome was death or ICU admission at 30 days. The cohort was divided into derivation and Omicron variant validation sets. We developed the LMIC-PRIEST score based on the coefficients from multivariable analysis in the derivation cohort and existing triage practices. We externally validated accuracy in the Omicron period and a UK cohort.

Results

We analysed 305,564, derivation 140,520 Omicron and 12,610 UK validation cases. Over 100 events per predictor parameter were modelled. Multivariable analyses identified eight predictor variables retained across models. We used these findings and clinical judgement to develop a score based on South African Triage Early Warning Scores and also included age, sex, oxygen saturation, inspired oxygen, diabetes and heart disease. The LMIC-PRIEST score achieved C-statistics: 0.82 (95% CI: 0.82 to 0.83) development cohort; 0.79 (95% CI: 0.78 to 0.80) Omicron cohort; and 0.79 (95% CI: 0.79 to 0.80) UK cohort. Differences in prevalence of outcomes led to imperfect calibration in external validation. However, use of the score at thresholds of three or less would allow identification of very low-risk patients (NPV ≥0.99) who could be rapidly discharged using information collected at initial assessment.

Conclusion

The LMIC-PRIEST score shows good discrimination and high sensitivity at lower thresholds and can be used to rapidly identify low-risk patients in LMIC ED settings.

What is already known on this subject

Uneven vaccination in low- and middle-income countries (LMICs) coupled with less resilient health care provision mean that emergency health care systems in LMICs may still be at risk of being overwhelmed during periods of increased COVID-19 infection.Risk-stratification scores may help rapidly triage need for hospitalisation. However, those proposed for use in the ED for patients with suspected COVID-19 have been developed and validated in high-income settings.

What this study adds

The LMIC-PRIEST score has been robustly developed using a large routine dataset from the Western Cape, South Africa and is directly applicable to existing triage practices in LMICs.External validation across both income settings and COVID-19 variants showed good discrimination and high sensitivity (at lower thresholds) to a composite outcome indicating need for inpatient admission from the ED.

How this study might affect research practice or policy

Use of the LMIC-PRIEST score at thresholds of three or less would allow identification of very low-risk patients (negative predictive value ≥0.99) across all settings assessedDuring periods of increased demand, this could allow the rapid identification and discharge of patients from the ED using information collected at initial assessment.

Differential diagnosis of COVID-19 and influenza

This study uses two existing data sources to examine how patients' symptoms can be used to differentiate COVID-19 from other respiratory diseases. One dataset consisted of 839,288 laboratory-confirmed, symptomatic, COVID-19 positive cases reported to the Centers for Disease Control and Prevention (CDC) from March 1, 2019, to September 30, 2020. The second dataset provided the controls and included 1,814 laboratory-confirmed influenza positive, symptomatic cases, and 812 cases with symptomatic influenza-like-illnesses. The controls were reported to the Influenza Research Database of the National Institute of Allergy and Infectious Diseases (NIAID) between January 1, 2000, and December 30, 2018. Data were analyzed using case-control study design. The comparisons were done using 45 scenarios, with each scenario making different assumptions regarding prevalence of COVID-19 (2%, 4%, and 6%), influenza (0.01%, 3%, 6%, 9%, 12%) and influenza-like-illnesses (1%, 3.5% and 7%). For each scenario, a logistic regression model was used to predict COVID-19 from 2 demographic variables (age, gender) and 10 symptoms (cough, fever, chills, diarrhea, nausea and vomiting, shortness of breath, runny nose, sore throat, myalgia, and headache). The 5-fold cross-validated Area under the Receiver Operating Curves (AROC) was used to report the accuracy of these regression models. The value of various symptoms in differentiating COVID-19 from influenza depended on a variety of factors, including (1) prevalence of pathogens that cause COVID-19, influenza, and influenza-like-illness; (2) age of the patient, and (3) presence of other symptoms. The model that relied on 5-way combination of symptoms and demographic variables, age and gender, had a cross-validated AROC of 90%, suggesting that it could accurately differentiate influenza from COVID-19. This model, however, is too complex to be used in clinical practice without relying on computer-based decision aid. Study results encourage development of web-based, stand-alone, artificial Intelligence model that can interview patients and help clinicians make quarantine and triage decisions.

Características clínicas, imagenológicas y de laboratorio de pacientes con COVID-19 según requerimiento de ingreso a UCI en Cali, Colombia

Introducción. Actualmente, hay pocos estudios en Latinoamérica sobre las características demográficas, clínicas y de laboratorio de pacientes con COVID-19 y con requerimiento de ingreso a unidad de cuidados intensivos (UCI). Objetivo. Comparar las características sociodemográficas, clínicas, imagenológicas y de laboratorio de pacientes diagnosticados con COVID-19 atendidos en el servicio de urgencias de una clínica en Cali, Colombia, según requerimiento de ingreso a UCI. Materiales y métodos. Estudio retrospectivo descriptivo de cohorte única realizado en 49 adultos con COVID-19 atendidos en el servicio de urgencias de un hospital de cuarto nivel de atención en Cali, Colombia, en marzo y abril de 2020, los cuales se dividieron en dos grupos: requerimiento de UCI (n=24) y no requerimiento de UCI (n=25). Se realizaron análisis bivariados para determinar las diferencias entre ambos grupos (pruebas de chi-2, exacta de Fisher, t de Student y U de Mann-Whitney), con un nivel de significancia de p<0.05. Resultados. La edad promedio fue 53 años (DE=13) y 29 pacientes fueron hombres. Se encontraron diferencias significativas entre ambos grupos en las siguientes variables: edad promedio (UCI x̅=58 vs. No UCI x̅=49; p=0.020), presencia de diabetes (8 vs. 1; p=0.010), presencia de dificultad respiratoria (20 vs. 11; p=0.007), presencia uni o bilateral de áreas de consolidación (12 vs. 3; p=0.005), mediana del conteo de leucocitos (Med=7570/mm3 vs. Med=5130/mm3; p=0.0013), de neutrófilos (Med=5980/mm3 vs. Med=3450/mm3; p=0,0001) y linfocitos (Med=865/mm3 vs. Med=1400/mm3; p<0,0001), mediana de proteína C reactiva (Med=141,25mg/L vs. Med=27,95mg/L; p<0,001), ferritina (Med=1038ng/L vs. Med=542,5ng/L; p=0.0073) y lactato-deshidrogenasa (Med=391U/L vs, Med=248,5U/L, p=0,0014). Finalmente, 15 pacientes requirieron ventilación mecánica invasiva, 2 presentaron extubación fallida, y en total, 5 fallecieron. Conclusiones. Se observaron diferencias significativas en los valores de varios marcadores inflamatorios, daño celular y parámetros del hemograma entre los pacientes que requirieron admisión a la UCI y los que no, por lo que estas variables podrían emplearse para desarrollar herramientas que contribuyan a establecer el pronóstico de esta enfermedad.

Predicting Severe Disease and Critical Illness on Initial Diagnosis of COVID-19: Simple Triage Tools

Rationale

This study was conducted to develop, validate, and compare prediction models for severe disease and critical illness among symptomatic patients with confirmed COVID-19.

Methods

For development cohort, 433 symptomatic patients diagnosed with COVID-19 between April 15th 2020 and June 30th, 2020 presented to Tawam Public Hospital, Abu Dhabi, United Arab Emirates were included in this study. Our cohort included both severe and non-severe patients as all cases were admitted for purpose of isolation as per hospital policy. We examined 19 potential predictors of severe disease and critical illness that were recorded at the time of initial assessment. Univariate and multivariate logistic regression analyses were used to construct predictive models. Discrimination was assessed by the area under the receiver operating characteristic curve (AUC). Calibration and goodness of fit of the models were assessed. A cohort of 213 patients assessed at another public hospital in the country during the same period was used to validate the models.

Results

One hundred and eighty-six patients were classified as severe while the remaining 247 were categorized as non-severe. For prediction of progression to severe disease, the three independent predictive factors were age, serum lactate dehydrogenase (LDH) and serum albumin (ALA model). For progression to critical illness, the four independent predictive factors were age, serum LDH, kidney function (eGFR), and serum albumin (ALKA model). The AUC for the ALA and ALKA models were 0.88 (95% CI, 0.86–0.89) and 0.85 (95% CI, 0.83–0.86), respectively. Calibration of the two models showed good fit and the validation cohort showed excellent discrimination, with an AUC of 0.91 (95% CI, 0.83–0.99) for the ALA model and 0.89 (95% CI, 0.80–0.99) for the ALKA model. A free web-based risk calculator was developed.

Conclusions

The ALA and ALKA predictive models were developed and validated based on simple, readily available clinical and laboratory tests assessed at presentation. These models may help frontline clinicians to triage patients for admission or discharge, as well as for early identification of patients at risk of developing critical illness.

Implementation of a COVID-19 screening tool in a southern Nigerian tertiary health facility

Screening for coronavirus disease 2019 (COVID-19) in emergency rooms of health facilities during outbreaks prevents nosocomial transmission. However, effective tools adapted for use in African countries are lacking. This study appraised an indigenous screening and triage tool for COVID-19 deployed at the medical emergency room of a Nigerian tertiary facility and determined the predictors of a positive molecular diagnostic test for COVID-19. A cross-sectional study of all patients seen between May and July 2020 at the Accident and Emergency of the University of Benin Teaching Hospital was conducted. Patients with any one of the inputs- presence of COVID-19 symptoms, history of international travel, age 60 years and above, presence of comorbidities and oxygen saturation < 94%- were stratified as high-risk and subjected to molecular testing for Severe Acute Respiratory Syndrome Coronavirus 2 (SARS-CoV-2). Data was obtained from the screening record book patterned after a modified screening tool for COVID-19, deidentified and entered into IBM-SPSS version 25.0. Binary logistic regression was conducted to determine significant predictors of a positive SARS-CoV-2 test. The level of significance was set at p < 0.05. In total, 1,624 patients were screened. Mean age (standard deviation) was 53.9±18.0 years and 651 (40.1%) were 60 years and above. One or more symptoms of COVID-19 were present in 586 (36.1%) patients. Overall, 1,116 (68.7%) patients were designated high risk and tested for SARS-CoV-2, of which 359 (32.2%) were positive. Additional inputs, besides symptoms, increased COVID-19 detection by 108%. Predictors of a positive test were elderly age [AOR = 1.545 (1.127-2.116)], co-morbidity [AOR = 1.811 (1.296-2.530)] and oxygen saturation [AOR = 3.427 (2.595-4.528)]. This protocol using additional inputs such as oxygen saturation improved upon symptoms-based screening for COVID-19. Models incorporating identified predictors will be invaluable in resource limited settings.