A mixture model to estimate SARS-CoV-2 seroprevalence in Chennai, India

Hitchings, Matt D T; Patel, Eshan U.; Khan, Rifa; Srikrishnan, Aylur K.; Anderson, Mark G.; Kumar, K S; Wesolowski, Amy; Iqbal, Hussain Syed; Rodgers, Mary A.; Mehta, Shruti H.; Cloherty, Gavin; Cummings, Derek A. T.; Solomon, Sunil S.

doi:10.1101/2022.02.24.22271002

Cited by 2 publications

(2 citation statements)

References 33 publications

(35 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Optical density (OD) was measured at a wavelength of 450 nm, and normalized values were derived by dividing the OD value of each test sample by the value of the calibrator. All samples with a normalized OD ≥ 0.8 were defined as positive, based on an estimation of optimal cutoffs (Fig B in S1 Appendix ) using a previously described Bayesian mixture model [ 18 ]. We consider a higher cutoff (normalized OD ≥ 1.1) in sensitivity analyses.…”

Section: Methodsmentioning

confidence: 99%

Structural factors associated with SARS-CoV-2 infection risk in an urban slum setting in Salvador, Brazil: A cross-sectional survey

et al. 2022

View full text Add to dashboard Cite

Background The structural environment of urban slums, including physical, demographic, and socioeconomic attributes, renders inhabitants more vulnerable to Severe Acute Respiratory Syndrome Coronavirus 2 (SARS-CoV-2) infection. Yet, little is known about the specific determinants that contribute to high transmission within these communities. We therefore aimed to investigate SARS-CoV-2 seroprevalence in an urban slum in Brazil. Methods and findings We performed a cross-sectional serosurvey of an established cohort of 2,041 urban slum residents from the city of Salvador, Brazil between November 2020 and February 2021, following the first Coronavirus Disease 2019 (COVID-19) pandemic wave in the country and during the onset of the second wave. The median age in this population was 29 years (interquartile range [IQR] 16 to 44); most participants reported their ethnicity as Black (51.5%) or Brown (41.7%), and 58.5% were female. The median size of participating households was 3 (IQR 2 to 4), with a median daily per capita income of 2.32 (IQR 0.33–5.15) US Dollars. The main outcome measure was presence of IgG against the SARS-CoV-2 spike protein. We implemented multilevel models with random intercepts for each household to estimate seroprevalence and associated risk factors, adjusting for the sensitivity and specificity of the assay, and the age and gender distribution of our study population. We identified high seroprevalence (47.9%, 95% confidence interval [CI] 44.2% to 52.1%), particularly among female residents (50.3% [95% CI 46.3% to 54.8%] versus 44.6% [95% CI 40.1% to 49.4%] among male residents, p < 0.01) and among children (54.4% [95% CI 49.6% to 59.3%] versus 45.4% [95% CI 41.5% to 49.7%] among adults, p < 0.01). Adults residing in households with children were more likely to be seropositive (48.6% [95% CI 44.8% to 52.3%] versus 40.7% [95% CI 37.2% to 44.3%], p < 0.01). Women who were unemployed and living below the poverty threshold (daily per capita household income <$1.25) were more likely to be seropositive compared to men with the same employment and income status (53.9% [95% CI 47.0% to 60.6%] versus 32.9% [95% CI 23.2% to 44.3%], p < 0.01). Participation in the study was voluntary, which may limit the generalizability of our findings. Conclusions Prior to the peak of the second wave of the COVID-19 pandemic, cumulative incidence as assessed by serology approached 50% in a Brazilian urban slum population. In contrast to observations from industrialized countries, SARS-CoV-2 incidence was highest among children, as well as women living in extreme poverty. These findings emphasize the need for targeted interventions that provide safe environments for children and mitigate the structural risks posed by crowding and poverty for the most vulnerable residents of urban slum communities.

show abstract

Section: Methodsmentioning

confidence: 99%

Structural factors associated with SARS-CoV-2 infection risk in an urban slum setting in Salvador, Brazil: A cross-sectional survey

et al. 2022

View full text Add to dashboard Cite

show abstract

“…This is a distinct problem from estimating high specificity cutoffs from a sample of validated negative controls, and as such, the methods presented here are not appropriate for this problem. When only estimating prevalence, some methods forgo establishing a cutoff and instead fit a mixture model [32][33][34][35][36] or a latent class model [37][38][39] to the continuous test results. In some situations, training data may be continuously collected.…”

Section: Discussionmentioning

confidence: 99%

Estimating cutoff values for diagnostic tests to achieve target specificity using extreme value theory

Pugh,

Fosdick,

Nehring

et al. 2024

BMC Med Res Methodol

View full text Add to dashboard Cite

Background Rapidly developing tests for emerging diseases is critical for early disease monitoring. In the early stages of an epidemic, when low prevalences are expected, high specificity tests are desired to avoid numerous false positives. Selecting a cutoff to classify positive and negative test results that has the desired operating characteristics, such as specificity, is challenging for new tests because of limited validation data with known disease status. While there is ample statistical literature on estimating quantiles of a distribution, there is limited evidence on estimating extreme quantiles from limited validation data and the resulting test characteristics in the disease testing context. Methods We propose using extreme value theory to select a cutoff with predetermined specificity by fitting a Pareto distribution to the upper tail of the negative controls. We compared this method to five previously proposed cutoff selection methods in a data analysis and simulation study. We analyzed COVID-19 enzyme linked immunosorbent assay antibody test results from long-term care facilities and skilled nursing staff in Colorado between May and December of 2020. Results We found the extreme value approach had minimal bias when targeting a specificity of 0.995. Using the empirical quantile of the negative controls performed well when targeting a specificity of 0.95. The higher target specificity is preferred for overall test accuracy when prevalence is low, whereas the lower target specificity is preferred when prevalence is higher and resulted in less variable prevalence estimation. Discussion While commonly used, the normal based methods showed considerable bias compared to the empirical and extreme value theory-based methods. Conclusions When determining disease testing cutoffs from small training data samples, we recommend using the extreme value based-methods when targeting a high specificity and the empirical quantile when targeting a lower specificity.

show abstract

A mixture model to estimate SARS-CoV-2 seroprevalence in Chennai, India

Cited by 2 publications

References 33 publications

Structural factors associated with SARS-CoV-2 infection risk in an urban slum setting in Salvador, Brazil: A cross-sectional survey

Structural factors associated with SARS-CoV-2 infection risk in an urban slum setting in Salvador, Brazil: A cross-sectional survey

Estimating cutoff values for diagnostic tests to achieve target specificity using extreme value theory

Contact Info

Product

Resources

About