Sentinel seroprevalence of SARS-CoV-2 in the Gauteng province, South Africa August to October 2020

Abstract

Background: Estimates of prevalence of anti-SARS-CoV-2 antibody positivity (seroprevalence) are for tracking the Covid-19 epidemic and are lacking for most African countries. Objectives: To determine the prevalence of antibodies against SARS-CoV2 in a sentinel cohort of patient samples received for routine testing at tertiary laboratories in Johannesburg, South Africa Methods: This sentinel study was conducted using remnant serum samples received at three National Health Laboratory Services laboratories situated in the City of Johannesburg (COJ) district, South Africa. Collection was from 1 August until the 31 October 2020. We extracted accompanying laboratory results for haemoglobin A1c, creatinine, HIV, viral load, and CD4+ T cell count. An anti-SARS -CoV-2 targeting the nucleocapsid (N) protein of the coronavirus with higher affinity for IgM and IgG antibodies was used. We reported crude as well as population weighted and test adjusted seroprevalence. Multivariate logistic regression method was used to determine if age, sex, HIV and diabetic status were associated with increased risk for seropositivity. Results: A total of 6477 samples were analysed; the majority (5290) from the COJ region. After excluding samples with no age or sex stated, the model population weighted and test adjusted seroprevalence for COJ (N=4393) was 27.0 % (95% CI: 25.4-28.6%). Seroprevalence was highest in those aged 45-49 [29.8% (95% CI: 25.5-35.0 %)] and in those from the most densely populated areas of COJ. Risk for seropositivity was highest in those aged 18-49 as well as samples from diabetics (aOR =1.52; 95% CI: 1.13-2.13; p=0.0005) and (aOR=1.36; 95% CI: 1.13-1.63; p=0.001) respectively. Conclusion: Our study conducted during the first wave of the pandemic shows high levels of infection among patients attending public health facilities in Gauteng. 274/400 Words

Estimating the time-varying reproduction number for COVID-19 in South Africa during the first four waves using multiple measures of incidence for public and private sectors across four waves

“…The similarities between public-and private-sector R estimates are of particular interest in light of several seroprevalence studies which suggest that clients of the public sector experienced higher levels of SARS-CoV-2 transmission [7,31,38]. It is also important to note the under-representation of clients of the public sector in all three data endpoints relative to clients…”

“…However, aside from the regular reports we released via the National Institute for Communicable Diseases (NICD) [23], the studies we identified relied on publicly reported time series data, which do not include symptom onset dates and which may be affected by backfilling and additional reporting delays, and used international estimates for the generation interval [1,2,[24][25][26][27][28][29][30]; most studies also use a single measure of incidence and do not cover all of the first four waves. While currently available evidence suggests that clients of the public sector experienced higher levels of transmission during the first two waves [7,31], we did not identify any studies comparing reproduction number estimates in different income groups, or between healthcare sectors.…”

Estimating the time-varying reproduction number for COVID-19 in South Africa during the first four waves using multiple measures of incidence for public and private sectors across four waves

“…The similarities between public- and private-sector R estimates are of particular interest in light of several seroprevalence studies which suggest that clients of the public sector experienced higher levels of SARS-COV-2 transmission [ 7 , 22 , 39 ], as well as the under-representation of clients of the public sector in all three data endpoints relative to clients of the private sector – clients of the private sector make up approximately 17% of the South African population, while only 52% of recorded cases were in the private sector.…”

“…However, aside from the regular reports we released via the National Institute for Communicable Diseases (NICD) [ 14 ], the studies we identified relied on publicly reported time series data, which do not include symptom onset dates and which may be affected by backfilling and additional reporting delays, and used international estimates for the generation interval [ 1 , 2 , 15 – 21 ]; most studies also use a single measure of incidence and do not cover all of the first four waves. While currently-available evidence suggests that clients of the public sector experienced higher levels of transmission during the first two waves [ 7 , 22 ], we did not identify any studies comparing reproduction number estimates in different income groups, or between healthcare sectors. R estimates are typically based on time series data of cases or deaths, though any measure representing an approximately constant proportion of total incidence may be used, in conjunction with data from which to estimate the generation interval [ 23 ].…”