Assessing the capacity of symptom scores to predict COVID-19 positivity in Nigeria: a national derivation and validation cohort study

Abstract

ObjectivesThis study aimed to develop and validate a symptom prediction tool for COVID-19 test positivity in Nigeria.DesignPredictive modelling study.SettingAll Nigeria States and the Federal Capital Territory.ParticipantsA cohort of 43 221 individuals within the national COVID-19 surveillance dataset from 27 February to 27 August 2020. Complete dataset was randomly split into two equal halves: derivation and validation datasets. Using the derivation dataset (n=21 477), backward multivariable logistic regression approach was used to identify symptoms positively associated with COVID-19 positivity (by real-time PCR) in children (≤17 years), adults (18–64 years) and elderly (≥65 years) patients separately.Outcome measuresWeighted statistical and clinical scores based on beta regression coefficients and clinicians’ judgements, respectively. Using the validation dataset (n=21 744), area under the receiver operating characteristic curve (AUROC) values were used to assess the predictive capacity of individual symptoms, unweighted score and the two weighted scores.ResultsOverall, 27.6% of children (4415/15 988), 34.6% of adults (9154/26 441) and 40.0% of elderly (317/792) that had been tested were positive for COVID-19. Best individual symptom predictor of COVID-19 positivity was loss of smell in children (AUROC 0.56, 95% CI 0.55 to 0.56), either fever or cough in adults (AUROC 0.57, 95% CI 0.56 to 0.58) and difficulty in breathing in the elderly (AUROC 0.53, 95% CI 0.48 to 0.58) patients. In children, adults and the elderly patients, all scoring approaches showed similar predictive performance.ConclusionsThe predictive capacity of various symptom scores for COVID-19 positivity was poor overall. However, the findings could serve as an advocacy tool for more investments in resources for capacity strengthening of molecular testing for COVID-19 in Nigeria.

Evaluation of Diagnostic Strategies for Identifying SARS-CoV-2 Infection in Clinical Practice: a Systematic Review and Compliance with the Standards for Reporting Diagnostic Accuracy Studies Guideline (STARD)

“…Three of the articles that used AI ( 12 , 15 , 18 ) and four of the studies that developed an algorithm/probability score ( 19 , 20 , 26 , 27 ) validated the developed model in an independent sample. In addition, other studies internally validated the model through techniques such as the bootstrap technique (a statistical procedure that resamples a single dataset to create many simulated samples) ( 21 , 22 , 24 ) or a k-fold cross-validation process (an approach that randomly divides the set of observations into k groups, or folds, of approximately equal size; the first fold is treated as a validation set, and the method is fit on the remaining k-1 folds) ( 13 , 14 , 17 ) ( Table 1 ).…”

“…Of the 16 studies which focused on symptomatic patients with suspected SARS-CoV-2 infection ( 12 – 27 ), seven included techniques related with artificial intelligence (AI) ( 12 – 18 ): five of them used a cohort study, either retrospective ( 12 , 13 , 17 , 18 ) or prospective ( 14 ), and two used a cross-sectional study ( 15 , 16 ). The other nine studies developed an algorithm ( 19 , 24 , 26 ) or a probability score ( 20 – 23 , 25 , 27 ) to diagnose infection by SARS-CoV-2. Four of them used a cross-sectional study ( 22 – 25 ), four used a retrospective cohort study ( 19 – 21 , 26 ), and one used a prospective cohort study ( 27 ).…”

“…The other nine studies developed an algorithm ( 19 , 24 , 26 ) or a probability score ( 20 – 23 , 25 , 27 ) to diagnose infection by SARS-CoV-2. Four of them used a cross-sectional study ( 22 – 25 ), four used a retrospective cohort study ( 19 – 21 , 26 ), and one used a prospective cohort study ( 27 ). Most of the studies included symptomatic patients who were admitted to a hospital due to suspected COVID-19; Trubiano JA et al’s study ( 24 ) included patients assessed for COVID-19 at a screening clinic due to symptoms (98.3%), contact with known COVID-19-positive patient (17.3%), or those who had traveled abroad (24.6%), and Elimian KO et al ( 26 ) included symptomatic patients and also those concerned about COVID-19 infection.…”

“…For a diagnostic model to be applied in practice, it must be validated in a sample which is independent from that in which it was developed ( 39 ). Only seven of the 16 studies that included symptomatic patients with suspected infection ( 12 , 15 , 18 – 20 , 26 , 27 ) and one of the two studies that included already diagnosed patients ( 33 ) carried out such external validation. Authors stated the need for further validation of the model in only five of the nine studies that did not include an external validation ( 21 – 25 ).…”

IFCnCov: An IoT‐based smart diagnostic architecture for COVID‐19

“…Singh et al16 designed and developed a novel ensemble-based classifier to predict COVID-19 cases at a very early stage so that appropriate action can be taken by patients, doctors, health organizations, and the government. Elimian et al17 developed and validated a symptom prediction tool for COVID-19 test positivity in Nigeria. Rinderknecht and Klopfenstein 18 developed a predictive model within 28 days of a COVID-19 diagnosis that predicts a critical state based on demographics, comorbidities,48 Initial diagnosis and patient monitoring in quarantine or at home may be aided by cloud-based healthcare.…”