key: cord-0691590-e2xmygu6
authors: Houngbegnon, P.; Nouatin, O.; Yadouleton, A.; Hounkpatin, B.; Fievet, N.; Atindegla, E.; Deschavanne, S.; Guichet, E.; Ayouba, A.; Pelloquin, R.; Maman, D.; Thaurignac, G.; Peeters, M.; Aviansou, A.; Sourakafou, S.; Delaporte, E.; Massougbodji, A.; Cottrell, G.
title: Interest of seroprevalence surveys for the epidemiological surveillance of the SARS-CoV-2 pandemic in African populations: insights from the ARIACOV Project in Benin
date: 2022-04-28
journal: nan
DOI: 10.1101/2022.04.26.22274330
sha: d4926e8f88f53e3bf3b6af7218e616ab3ac9f208
doc_id: 691590
cord_uid: e2xmygu6

Background Many SARS-CoV-2 seroprevalence surveys since the end of 2020 have disqualified the first misconception that Africa had been spared by the pandemic. Through the analysis of three SARS-CoV-2 seroprevalence surveys carried out in Benin as part of the ARIACOV project, we argue that the integration of epidemiological serosurveillance of the SARS-COV2 infection in the national surveillance package would be of great use to refine the understanding of the COVID-19 pandemic in Africa. Methods Three repeated cross-sectional surveys have been carried out in Benin, two in Cotonou, the economic capital in March and May 2021, and one in Natitingou, a semi-rural city in North in August 2021. The global and by age-groups weighted seroprevalences have been estimated and the risk factors of the infection by SARS-COV-2 have been assessed by using logistic regression. Results In Cotonou, a slight increase in overall age-standardized SARS-CoV-2 seroprevalence from 29.77% (95% CI: 23.12-37.41%) at the first survey to 34.86% (95% CI: 31.57-38.30%) at the second survey was observed. In Natitingou the global adjusted seroprevalence was 33.34% (95% CI: 27.75-39.44%), much higher than expected. Adults over 40 seemed to be more at risk than the youngest during the first survey in Cotonou but no longer in the second survey, showing the persistence of the SARS-COV-2 virus circulation outside the epidemic waves. Conclusions A routine serological surveillance on strategic sentinel sites and / or populations could constitute a cost / effective compromise to better anticipate the onset of new waves and define public health strategies.

Many SARS-CoV-2 seroprevalence surveys since the end of 2020 have disqualified the first misconception that Africa had been spared by the pandemic. Through the analysis of three SARS-CoV-2 seroprevalence surveys carried out in Benin as part of the ARIACOV project, we argue that the integration of epidemiological serosurveillance of the SARS-COV2 infection in the national surveillance package would be of great use to refine the understanding of the COVID-19 pandemic in Africa.

Three repeated cross-sectional surveys have been carried out in Benin, two in Cotonou, the economic capital in March and May 2021, and one in Natitingou, a semi-rural city in North in August 2021. The global and by age-groups weighted seroprevalences have been estimated and the risk factors of the infection by SARS-COV-2 have been assessed by using logistic regression.

In Cotonou, a slight increase in overall age-standardized SARS-CoV-2 seroprevalence from 29 .77% (95% CI: 23. .41%) at the first survey to 34.86% (95% CI: 31.57-38.30%) at the second survey was observed. In Natitingou the global adjusted seroprevalence was 33.34% (95% CI: 27.75-39.44%), much higher than expected. Adults over 40 seemed to be more at risk than the youngest during the first survey in Cotonou but no longer in the second survey, showing the persistence of the SARS-COV-2 virus circulation outside the epidemic waves.

A routine serological surveillance on strategic sentinel sites and / or populations could constitute a cost / effective compromise to better anticipate the onset of new waves and define public health strategies.

. CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)

In Benin, the first case of SARS-CoV-2 infection was detected March, 16 th 2020 (1) . From the very first months of the COVID-19 pandemic, most of African countries implemented public health measures (2,3) to counter the spread of the virus, in particular sanitary cords, border closures and sometimes containment. These health governmental decisions were accompanied by sensitization campaigns on the wearing of a mask, social distancing measures, barrier gestures, hand washing, and also by increasing the capacities of national laboratories in terms of SARS-CoV-2 infection screening by Polymerase Chain Reaction (PCR) (4) . The objective of this strategy was to interrupt the chains of transmission as quickly as possible, knowing that the capacity to handle severe cases in hospital was limited in most countries. Nevertheless, the number of tests carried out in Africa has remained low compared to the rest of the world (5, 6) and most often mainly reserved for capitals or large cities, making it difficult to assess the reality of the circulation of the virus in country sides. So far, the severity of the disease has been less than in other parts of the world (7-9) and the proportion of asymptomatic or paucisymptomatic cases among infected people is probably very high. As a result, data on the number of COVID-19 cases in Africa are scarce (10, 11) and their ability to reflect the reality of the extent of the pandemic is questionable. The low number of detected cases and reported deaths in Africa (12) (13) (14) (15) (16) (17) (18) (19) has sometimes been interpreted as the fact that the continent had been spared from the pandemic. However, many SARS-CoV-2 seroprevalence surveys (following the WHO recommendations (20) ) published since the end of 2020 in different African countries, have now definitively disqualified this misconception (21) . The ARIACOV project (Support for the African Response against COVID-19) which carried out SARS-CoV-2 seroprevalence surveys in 6 West and Central African countries (Benin, Ghana, Cameroon, Democratic Republic of Congo, Guinea and Senegal) together with other initiatives helped to demonstrate that the virus has circulated widely on the continent (4, 11, 21, 22) . All of these results show that, as highly . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)

The copyright holder for this preprint this version posted April 28, 2022. ; https://doi.org/10.1101/2022.04.26.22274330 doi: medRxiv preprint useful as it is to understand the general dynamics of the COVID-19 pandemic, routine epidemiological surveillance in Africa exclusively based on the detection of infected cases by PCR (or antigenic RDT more recently) has proved unable to account for the true extent of the epidemic in the African context (limited screening capacity, proportion of asymptomatic cases probably overwhelming). The integration of epidemiological sero-surveillance based on the detection of antibodies against SARS-CoV-2 in the national surveillance package would certainly be of great use to refine the understanding of the dynamics of the epidemic and to detect the most at risk of infection groups in the populations. The aim of this work is to argue this idea through the analysis of three SARS-CoV-2 seroprevalence surveys carried out in Benin as part of the ARIACOV project.

The first two surveys of the study were conducted in Cotonou, the economic capital of Benin with the port and the airport, with an area of 79 km 2 and an estimated population of 840 313 in 2020 (23) . To assess the COVID-19 epidemic situation in a city where epidemic activity was assumed to be low according to the national surveillance, the third survey was carried out at Natitingou, a middle-size city located more than 500 kilometers at North of Benin. Natitingou has an area of 3045 km 2 with an estimated population of 128 511 in 2020 (23).

We conducted three cross-sectional SARS-CoV-2 seroprevalence studies in households. The The third study took place in Natitingou from August 20 to 26, 2021 (increase of the third . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)

The copyright holder for this preprint this version posted April 28, 2022. ; https://doi.org/10.1101/2022.04.26.22274330 doi: medRxiv preprint wave). These studies were carried out as part of the ARIACOV project and used the WHO population-based age stratified seroepidemiological investigation protocol for COVID-19 infection version 2.0.

A two-stage cluster sampling was employed in each study of Cotonou. At the first stage, 50 neighborhoods among 143 were randomly drawn with a probability proportional to their population size, and at the second stage 12 households were selected per neighborhood. In order to balance the age groups (to be able to compare the 2 surveys by age groups), in half of the households all the residents were invited to participate in the study and in the remaining 50%, only the residents aged 40 years and older were invited to participate. In Natitingou, a two-stage cluster sampling was also carried out, where 30 neighborhoods among 77 were randomly drawn with a probability proportional to their size, and then 5 households were randomly drawn in each neighborhood. In Natitingou, all the residents of any household were invited to participate in the study. In each study, the eligible participants were proposed to fill an electronic questionnaire aiming to collect on smartphones general (age, gender…) and socio-enonomic (occupation, individual family home, common yard…) data, as well as experience of COVID-19 confirmed disease or related symptoms in the past 3 months and/or hospitalization (defined asymptomatic if no symptom). The participants were eventually proposed to be performed a blood sample for SARS-CoV-2 serological analysis in the laboratory.

Ethics approval (N°131/MS/DRFMT/CNERS/SA) and Statistical approval (N°26/2020/MPD/INSAE/DCSFR) were obtained respectively from the Comité d'Ethique pour la Recherche en Santé and the Conseil National de la Statistique. All residents (adults and children) were informed about the study objectives and procedures. Adults provided written consent to participate in the study and to be tested for SARS-CoV-2 serology prior to starting . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted April 28, 2022. ; https://doi.org/10.1101/2022.04.26.22274330 doi: medRxiv preprint the interview. Written parental consent and children assent were obtained prior to enrollment of young participants. At the end of the study, each participant was informed about his serology status.

Venous blood samples (2-5ml) collected in EDTA tube from participants were carefully transported to the laboratory of the Clinical Research Institute of Benin (IRCB). After centrifugation and inactivation at 56°C during 45 min, plasmas samples were used for SARS-CoV-2 serological analyses. In this study, the Luminex-based assay was used to simultaneously detect IgG antibodies to two viral antigen, i.e recombinant nucleocapsid (NC) and spike (SP) proteins of SARS-CoV-2, as previously described (24) . This dual target strategy for SARS-CoV-2 IgG antibodies detection give a great sensitivity (100%) and specificity (99,7%) that was previously validated on an panel of more than 1 000 samples from Africa (including 160 samples from Benin) before COVID-19 (22, 24) . Results were expressed as the median fluorescence intensity (MFI) for 100 beads. Cutoff values were also previously determined using pre-COVID-19 samples and those from COVID-19 positive and hospitalized patients (24) . An experiment was validated when (1) the MIF value of the blank is less than 50 MFI regardless of the antigen and (2) the MFI value of the negative control is less than 100. A sample is considered positive for anti-SARS-CoV-2 IgG when positive both for anti-SP IgG and anti-NP IgG, and is declared negative when negative for this both antigens.

The statistical analysis was performed using STATA 14 (StataCorp, College Station, TX, USA). In Cotonou the weighted estimations of the SARS-CoV-2 seroprevalence (and 95% confidence intervals) were performed in the 2 surveys by age-standardizing the crude . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted April 28, 2022. ; https://doi.org/10.1101/2022.04.26.22274330 doi: medRxiv preprint seroprevalence using available national census data (23). We estimated the overall SARS-CoV-2 seroprevalence and the seroprevalence stratified by sex and age groups in each of the 3 surveys. A weighted multivariable logistical regression was performed to determine the factors associated with positive serology anti-SARS-COV-2. The age group and symptom variables were forced into the model.

In Cotonou, during the first and second surveys, 617 and 606 households were visited respectively, with a total of 1596 and 1842 eligible individuals respectively (Table 1) . Among them, 1416 (89%) and 1465 (80%) individuals were present and agreed to participate in the first and second survey respectively and finally, 1405 and 1460 compliant blood samples with epidemiological data were collected, respectively. The average size of the households surveyed was 3.26 ± 1.98 for the first survey and 3.97 ± 1.81 for the second survey. In Natitingou, 155 households were visited for a total of 815 eligible individuals, 742 (91%) individuals were present and agreed to participate in the study, of which 735 did provide a compliant blood sample. The average size of the households surveyed was 6.06 ± 2.16. The general characteristics of the households and participants are described in Table 1 .

. CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted April 28, 2022 . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted April 28, 2022 

In Cotonou, a slight increase in overall age-standardized SARS-CoV-2 seroprevalence from 29.77% (95% CI: 23.12-37.41%) at the first survey to 34.86% (95% CI: 31.57-38.30%) at the second survey was observed (Table 2 ). In Natitingou, the global adjusted seroprevalence was also as high as 33.34% (95% CI: 27.75-39.44%). The seroprevalence stratified by sex is presented Table 2 . The proportion of symptomatic subjects according to the SARS-COV-2 serology is shown Figure 1 . The first survey in Cotonou showed a lower seroprevalence among the youngest age groups (less than 40 years) compared to 40 years and above (Figure 2 ), but this trend was no longer observed at the second survey. In Natitingou, a higher seroprevalence was observed among people aged over 60 years or over but the sample size in this age group was quite low ( Figure 3 ).

The multivariate logistic regression (Table 3 ) presents the covariates for which a clear association with a SARS-CoV-2 seropositivity was found in at least one of the surveys. The three surveys showed an OR greater than 1 in women (higher risk of positive serology, with a . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted April 28, 2022. ; https://doi.org/10.1101/2022.04.26.22274330 doi: medRxiv preprint p <0.05 only for the second Cotonou survey). A consistent trend towards a higher risk in more populated households or in households living in common yards (vs. more isolated structures) was observed during the first survey in Cotonou, but this was not found in the others surveys.

Adults over 40 seemed to be more at risk than the youngest (with p> 0.05 however) during the first survey in Cotonou, while in the second survey the ORs (and 95% CI) in the different age groups were more homogeneous. In Natitingou, a clear tendency for increased risk with age was found, with an OR of 2.67 for people aged 60 and over (p <0.05). No association was found between symptomatic status during the last 3 months and SARS-CoV-2 positive serology in surveys in Cotonou. In Natitingou, an incread-risk OR in symptomatic people was observed (OR = 1.33), although this association was not very clear (p = 0.25).

. CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted April 28, 2022 is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted April 28, 2022. in Cotonou whereas just after the first survey in Cotonou, the number of cases diagnosed by PCR nationwide was 6 501 (only one case detected for every 40 infections). This ratio is in accordance with numerous seroprevalence surveys carried out in Africa (4, 11, 25, 26) .The seroprevalences we found in our 3 surveys fit also well with a recent meta-analysis showing a combined seroprevalence of 25% (95% C.I. [13% -39%] in West Africa (21) . Some studies, for example in Mali or in Zimbabwe have measured seroprevalences before and after epidemic waves, demonstrating the substantial impact of these waves on seroprevalence estimates (11, 27) . Unlike these studies, the two surveys in Cotonou were both carried out during and after second wave (end of January to end of March) and only 2 months apart. This largely explains the smaller increase in seroprevalence between the two surveys (from 29.77% to 34.86%). Interestingly in the second survey in Cotonou, the prevalence among the youngest caught up with the prevalence of the older age groups. This could suggest an impact of the sensitization discourse (particularly oriented towards older age groups), associated to a greater respect for distancing measures between the two surveys by older age groups compared to younger ones. It reflects also the persistence of virus circulation even outside of waves, so even when virological PCR monitoring does not seem to suggest any increase in COVID-19 cases incidence.

. CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted April 28, 2022 Based on the findings in Cotonou, our hypothesis was that the prevalence of infected people with the SARS-CoV-2 virus was undoubtedly underestimated throughout the country. The goal of the investigation in Natitingou was then to assess the situation in an area where available data suggested very low circulation of the virus, despite the start of the third COVID-19 epidemic wave in Benin (August to October 2021). The estimation of age-groups prevalences was not a main objective, which justifies the non age-group stratified design in Natitingou. Our hypothesis was confirmed, with an estimated overall adjusted prevalence of 33.34%, well above the low PCR positive rate detected by the epidemiological surveillance in the study area (data not shown), and in agreement with the high prevalence observed in semi-rural areas after the second wave in Mali for example (26) .

Across the three surveys, the presence of symptoms in the past 3 months has not shown any relationship with positive serology, as observed in several other studies (4, (28) (29) (30) (31) . This supports the hypothesis that the vast majority of the infected people have been asymptomatic or pauci-symptomatic, even if the information regarding the symptoms may be affected by a memorization bias. The proportion of people (even symptomatic) having tested for COVID-19 was very low (less than 5%, data not shown), which shows the limits of surveillance based on PCR tests intended for people presenting spontaneously.

Our study has limitations. In Cotonou, the design was repeated cross-sectional rather than longitudinal surveys. Consequently, the differences between the prevalence by age groups in the two surveys may be because the samples were carried out different neighborhoods of Cotonou. Nonetheless, the relatively small increase in prevalence between the two surveys is consistent with the relatively narrow time interval (2 months) with no epidemic wave in between. Another limitation is that in Natitingou the design does not allow to estimate the prevalences by age group as precisely as in Cotonou. However, all age groups seem to be highly . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted April 28, 2022. ; https://doi.org/10.1101/2022.04.26.22274330 doi: medRxiv preprint affected, which is consistent with what was observed in Cotonou and in other studies including the ARIACOV project (11) .

All of these results show the essential value of seroprevalence surveys in providing a more accurate view of the spread and the extent of the COVID19 epidemic than the reported cases from virological PCR-based surveillance only. Virological surveillance seems suitable for describing epidemic waves when they are in progress, but turns out to be inadequate neither for capturing a precise quantitative estimate of the extent and spread of the epidemic in populations.

It fails also to apprehend clearly the persistence of the virus and its basal dynamics between epidemic waves. Our results show, along with others in Africa, that a more systematic use of serological surveillance to support an overall strategy of routine epidemiological surveillance would certainly be useful in overcoming these shortcomings. The interest of associating seroprevalence studies with PCR (or RDT) screening as part of the epidemiological surveillance of the pandemic was pointed out in early 2020 by international health institutions such as the World Health Organization (32) or the Center of Disease Control and Prevention (33) especially given the high proportion of infected people with few or no symptoms. Their rapid integration into the national packages for monitoring the pandemic in northern countries has effectively made it possible to better estimate the real extent of the transmission of the virus, the risk factors for infection and the foci of transmission. In particular before mass vaccination of populations, knowledge of the true proportion and sub-groups of the population with anti-SARS-COV-2 antibodies is an essential element to better guide the national vaccine strategies.

These considerations are particularly important in an African context where many countries are constrained to lighten, sometimes significantly, their epidemiological surveillance systems because of its logistical burden and high cost, especially between epidemic waves when the epidemic situation seems stable. However, as shown here, the virus probably continues to diffuse quietly, contributing to the emergence of new variants and new waves, particularly in . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)

The copyright holder for this preprint this version posted April 28, 2022. ; https://doi.org/10.1101/2022.04.26.22274330 doi: medRxiv preprint populations that are still poorly vaccinated. It is therefore urgent to think about a way to maintain a routine epidemiological surveillance adapted to this context. To minimize the cost, a routine serological surveillance on strategic sentinel sites and / or populations could constitute a cost / effective compromise to better anticipate the onset of new waves and characterize the most vulnerable subgroups of populations. In order to take into account the logistical constraints, the use of an easy-to-use diagnostic tool such as serological RDTs , well known to health workers as well as to populations, as it is already used for the COVID-19 case detection (antigenic RDTs) and in other pathologies like malaria, could prove interesting.

. CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)

The copyright holder for this preprint this version posted April 28, 2022. ; https://doi.org/10.1101/2022.04.26.22274330 doi: medRxiv preprint

Informations coronavirus (covid-19) | Gouvernement de la République du Bénin

The COVID-19 epidemic

High prevalence of anti-SARS-CoV-2 antibodies after the first wave of COVID-19 in Kinshasa, Democratic Republic of the Congo: results of a cross-sectional household-based survey

African Union rolls out Partnership to Accelerate COVID-19 Testing

COVID-19: Unpacking the low number of cases in Africa. Public Health Pract

Latest updates on the COVID-19 crisis from Africa CDC

SARS-CoV-2 diagnostic testing in Africa: needs and challenges

Modelling COVID-19 transmission in Africa: countrywise projections of total and severe infections under different lockdown scenarios

Serological evidence of human infection with SARS-CoV-2: a systematic review and meta-analysis

Community SARS-CoV-2 seroprevalence before and after the second wave of SARS-CoV-2 infection in Harare

The Conundrum of Low COVID-19 Mortality Burden in sub-Saharan Africa: Myth or Reality? Glob Health Sci Pract

WHO | Regional Office for Africa. Coronavirus (COVID-19)

Americas Region COVID-19 Dashboard

COVID-19 situation in the WHO South-East Asia Region

Regional Office for Europe. COVID-19 situation in the WHO European Region

Regional Office for the Eastern Mediterranean. COVID-19 situation in the Region -total reports

Western Pacific Region. COVID-19 Situation in WHO -Western Pacific Region

World Health Organizatio(WHO)

World Health Organization. Population-based age-stratified seroepidemiological investigation protocol for coronavirus 2019 (COVID-19) infection

Seroprevalence of anti-SARS-CoV-2 antibodies in Africa: A systematic review and meta-analysis

ARIACOV : projet de recherche-action en appui à la riposte africaine à l'épidémie de Covid-19 | Site Web IRD

Multiplex detection and dynamics of IgG antibodies to SARS-CoV2 and the highly pathogenic human coronaviruses SARS-CoV and MERS-CoV

Estimated SARS-CoV-2 infection rate and fatality risk in Gauteng Province, South Africa: a population-based seroepidemiological survey

Séroprévalence SARS-CoV-2 au Mali : résultats d'une enquête transversale

The epidemiological characteristics of an outbreak of 2019 novel coronavirus diseases (COVID-19) in China

Evidence of SARS-CoV-2 Infection in Returning Travelers from Wuhan, China

Clinical features of patients infected with 2019 novel coronavirus in Wuhan, China

Limited Specificity of Serologic Tests for SARS-CoV-2 Antibody Detection

Challenges in interpreting SARS-CoV-2 serological results in African countries

WHO. Unity Studies: Early Investigation Protocols

Centers for Disease Control and Prevention. 2020

Special gratitude to all the study participants. We acknowledge all the technicians and the field surveyors. We highlighted the facilities offer by the administrative and health authorities at central and local levels. We acknowledge Inès Boko, Elisée ADIMI, Romuald AKOHO and Jean-François Etard for their appreciated contribution to the study.

Research reported in this publication was funded by the Agence Française de Développement, (AFD) as part of the project "Appui à la Risposte Africaine de l'Epidémie à COVID-19 (ARIACOV)" carried out in Benin by the Institut de Recherche Clinique du Bénin.

All authors no reported conflicts of interest. All authors have submitted the ICMJE form for Disclosure of Potential Conflicts of Interest.

It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted April 28, 2022