SARS-CoV-2 Seroprevalence after Third Wave of Infections, South Africa

By November 2021, after the third wave of severe acute respiratory syndrome coronavirus 2 infections in South Africa, seroprevalence was 60% in a rural community and 70% in an urban community. High seroprevalence before the Omicron variant emerged may have contributed to reduced illness severity observed in the fourth wave.

All household members were approached for inclusion in the study, and households were eligible for inclusion if >80% of members consented to be enrolled.
For PHIRST-C, all households who participated in PHIRST from the urban site, and households from the 2017 and 2018 cohort in the rural site were approached for enrolment. To supplement the sample size, additional households at each site were approached using the same methods as for PHIRST. Villages in the rural site were restricted to those four used in 2017 and 2018 of PHIRST. Informed consent was obtained from participating adults or a parent/guardian for children <18 years of age. In addition to parent/guardian consent, assent was also obtained from children aged 7-17 years.
We collected baseline data and blood (blood draw [BD]  The Bushbuckridge Municipality within the Ehlanzeni District is considered as predominantly rural, with main industries being agriculture and tourism (2). The Ehlanzeni District has a population of 1,828,738, with sixty percent of the district's population being >18 years of age. The City of Matlosana Municipality is considered to be 88.2% urban, and is located in the Dr Kenneth Kaunda District, which has a population of 797,715 (3). Sixty-five percent of the district population is aged >18 years.

to-infection Ratio (FIR) by Wave of Infection
We calculated the age-and sex-adjusted total number of infections, age-adjusted diagnosed cases, hospitalizations, deaths, CIR (number of infections compared to diagnosed cases), HIR and in-hospital and excess death FIR as described in below equations. We defined the third wave as March 22, 2021 (week 12) to November 14, 2021 (week 45), starting with BD5, and ending the week before BD9 started. Due to differences between the sex ratios of our cohort and the district population, the infection estimates were also adjusted for sex. Data sources used in these calculations are described in the next section. Age-and sex-standardized estimates for the selected endpoints for each wave were obtained as follows: Where si is the seroprevalence in the cohort in the respective community for age and sex group i and SApi is the South African population for age and sex group i. Calculated for wave 3 as the number of individuals seronegative at BD5 and seropositive at BD9. Furthermore, we also included possible re-infections where the individual was already seropositive at BD5, but had a ≥2-fold higher cutoff index (COI) in BD9 compared to BD5. We therefore included infections and re-infections that occurred during the third wave. Estimates only included participants with a blood draw 5 and 9 pair, and adjusted for sensitivity and specificity of test (4).
The ≥2-fold increase in COI cutoff to define possible re-infections was not externally validated, but based on PCR-confirmed infections from twice-weekly nasal swab collections ≥2-fold increase. We were therefore more likely to have underestimated re-infections.
Where ci is the number of diagnosed cases (RT-PCR and antigen-based tests) from the respective district reported to the NMCSS during wave 3 (March 22, to November 14, 2021) in age group i, pi(d) is the district population for age group i and pi(SA) is the South African population for age group i.
Where hi is the number of hospitalizations from the respective district reported to Where di is the number of in-hospital deaths from the respective districts reported to DATCOV during wave 3 (March 22, to November 14, 2021) (5) in age group i, pi(d) is the district population for age group i and pi(SA) is the South African population for age group i.
Where ExD is the rate of provincial excess deaths adjusted to the South African  (6).
For infections, hospitalizations, in-hospital (minimum) and excess (maximum) deaths, 95% CIs were calculated using the Clopper-Pearson method with a Poisson distribution.
Confidence intervals for infection ratios were calculated as ratios from the 95% confidence intervals of infection, hospitalization and death rates.

Data Sources
Population Denominators (7) Population numbers for each district, by age, were obtained from the StatsSA 2021 mid- year population estimates.

Notifiable Medical Conditions Surveillance System (NMCSS) (8)
Reverse transcription PCR (RT-PCR) testing in South Africa to detect SARS-CoV-2 RNA started on 28 January 2020. Rapid SARS-CoV-2 antigen testing was implemented in

Seroreversions
Seven individuals were seronegative at BD5, and with at least one seropositive result during BD6 and BD8, and who were seronegative again at BD9. The maximum COIs for these individuals were low, ranging from 1.09 to 34.5. This fits with an initial low antibody response and subsequent antibody titer waning. In addition, there were five individuals who were seropositive at BD5, whose antibody titers waned below detection during BD6 to BD8, but who were seropositive again at BD9. Two of these individuals had a lower COI in BD9 than in BD5, and were not included as possible re-infections during wave 3.

Limitations
Our study is limited by a small sample size, reducing the power for accurate seroprevalence estimates in small age strata, and inclusion of only 2 geographic sites, including only households with ≥3 people, with high unemployment in selected households compared to the community unemployment rate (9) and therefore may not be representative of other districts and provinces in South Africa. Our definition of re-infections has not been validated elsewhere and is based on a small number of infections that occurred between BD5 and BD7, and did not consider different variants. Based on data from the same cohort, these reinfections occurred in only a small portion (3%) of the cohort (C. Cohen et al., unpub (6, 1-15) 5/50 (11, 4-21) 7/50 (15, 7-26) 8/44 (20, 9-32)