Description: Topics covered in the questionnaire are: household members and their characteristics, household income, availability of services, demographic information, medical history, health risk behaviours, symptom history, access to COVID-19 prevention measures, self-perceived risk, and travel habits or physical distancing.

A total of 10 109 VPs were approached to participate in the survey, of the 10 109 VPs approached, 5 580 agreed to participate, giving a household response rate of 55%. Of all the VPs approached, 43.1% refused to take part in the survey, and 1.6% of the valid VPs were unoccupied or empty.

A total of 16 646 individuals were eligible to participate in the study, of whom 90.87% agreed to be interviewed. Of those who were eligible, 81.9% completed the interview and provided blood specimens.

The data set available for dissemination contains 15 127 cases and 156 variables. Abstract: This dataset is a nationwide household-based population survey of SARS-CoV-2 seroprevalence in adolescents 12 years and older in South Africa. Data was collected on two separate rounds using a cross-sectional multi-stage stratified cluster survey design where SARS CoV-2 antibodies were tested first using the Abbott Architect anti-SARS CoV-2 immunoglobulin class G (IgG) chemiluminescent microparticle immunoassay with the final status determined by the Euroimmun Anti-SARS-CoV-2 ELISA (IgG) Euroimmun ® antibody assay. A subset of specimens positive on the Abbott ® assay were selected for analysis using a pseudotyped neutralization assay. Seroprevalence estimates presented are based on the Euroimmun ®assay. Summary statistics were used to describe SARS-CoV-2 Seroprevalence and characteristics of the study population. Bivariate and multivariate logistics regression analysis were used to identify factors associated with SARS-CoV-2 seropositivity.

In addition, among the samples tested for neutralization less than two thirds neutralized the D614G and the Beta strains, highlighting the importance of vaccination.

Face-to-face interview

Individuals meeting the following criteria were included in the survey:

• A de factor household member aged 12 years and older who provided informed consent (assent and parental or guardian consent were required for those younger than 18 years old);

• Individuals who had previously been tested for COVID-19 comprising of those who tested positive for COVID-19 and were self-isolating and those who received the COVID-19 vaccination.

The sampling procedures used in the study were adapted from the WHO vaccination coverage cluster surveys reference manual. Originally, the design aimed to target 19 620 individuals from 436 Small Area Layers (SALs), resulting in an effective sample size of 13 734 individuals from 6 540 households. This sample size was based on an assumption of a 2% SARS-CoV-2 prevalence in South Africa in June 2020, with a 2% margin of error, and a joint (household and individual) response rate of 70%. However, after data collection from November 2020 to February 2021, the actual response rate was lower than expected at 46%. As a result, the status of the epidemic and survey field conditions were reviewed, and following consultations with the Survey Advisory Committee and the Funder, the sample size was revised to improve estimates in some population sub-groups. Considering the more realistic SARS-CoV-2 prevalence of 20% and a response rate of 64%, a sample size of 12 625 individuals providing a blood specimen for antibody testing was projected to achieve precision levels of 1.5% or lower. Therefore, the survey was conducted in two rounds: Round 1 took place between November 2020 and February 2021, and Round 2 between April and June 2021.

Since the main reason to have a second round of data collection was to supplement the data collected during the first round by gathering more data in all provinces and increase the representation of the White and Indian population groups, which were previously underrepresented. An additional 200 SALs were targeted across all nine provinces to reach 6 930 respondents who would provide a blood specimen (from 9 990 individuals approached). Overall, 15 households were randomly selected for inclusion across 647 SALs nationally; and the survey covered both periods when the original SARS-CoV-2 (Wuhan D614G) virus strain and the Beta variant were circulating in the country. In addition, it is worth noting that data collection occurred before the wide-scale vaccination rollout in South Africa, which commenced on 17 February 2021, initially targeting healthcare workers and subsequently following a phased approach for essential workers, individuals in congregate settings, and by age group, starting with the elderly. By the end of the data collection period on 15 June 2021, approximately 2 067 424 vaccinations had been administered, covering approximately 3.4% of the population.