Integration of symptomatic, demographical and diet-related comorbidities data with SARS-CoV-2 antibody rapid diagnostic tests during epidemiological surveillance: a cross-sectional study in Jakarta, Indonesia

OBJECTIVES

Affordable options for COVID-19 epidemiological surveillance are needed. Virus detection by reverse transcription-PCR (RT-PCR) is sensitive but costly, and antigen-based rapid diagnostic tests (RDTs) are cheap but with reduced sensitivity; both detect current infection but not exposure. RDT-IgM/IgG antibodies to SARS-CoV-2 detect exposure but have poor sensitivity for current infection. We investigated if the integration of symptomatic, demographical and diet-related comorbidities data with antibody RDTs improves their potential to assess infection rates in addition to exposure, thereby broadening their utility for surveillance.

DESIGN

We conducted a cross-sectional study using data from community surveillance for SARS-CoV-2. Health workers collected nasopharyngeal swabs for RT-PCR and RDT antigen assessments and venous blood for RDT-IgM/IgG from symptomatic and asymptomatic persons. Data on age, gender, contact history, symptoms (ie, fever, cough, runny nose, sore throat, headache, dyspnoea and diarrhoea), diet-related comorbidities (ie, diabetes and hypertension) and chest radiology were collected.

SETTING

High-risk communities in Jakarta, Indonesia, in May 2020.

PARTICIPANTS

343 community members' data were included.

OUTCOME MEASURES

RDT-IgM/IgG sensitivity, specificity and predictive values and area under receiver operating characteristic curve for RT-PCR positivity using RDT results alone and in combination with other predictors, including symptom components derived from principal component analysis.

RESULTS

There were 24 PCR-confirmed infections. RDT-IgM/IgG-positive tests were associated with infection (OR 10.8, 95% CI 4.43 to 26.4, p<0.001) with an area under the curve (AUC) of 0.708% and 50% sensitivity, 91.5% specificity, 30.8% positive predictive value (PPV) and 96.1% negative predictive value (NPV). RDT results combined with age, gender, contact history, symptoms and comorbidities increased the AUC to 0.787 and yielded 62.5% sensitivity, 87.0% specificity, 26.6% PPV and 96.9% NPV.

CONCLUSIONS

SARS-CoV-2 RDT-IgM/IgG results integrated with other predictors may be an affordable tool for epidemiological surveillance for population-based COVID-19 exposure and current infection, especially in groups with outbreaks or high transmission.