key: cord-0899619-iqq8kmn0 authors: Uyoga, S.; Adetifa, I. M. O.; Karanja, H. K.; Nyagwange, J.; Tuju, J.; Wanjiku, P.; Aman, R.; Mwangangi, M.; Amoth, P.; Kasera, K.; Ng'ang'a, W.; Rombo, C.; Yegon, C. K.; Kithi, K.; Odhiambo, E.; Rotich, T.; Orgut, I.; Kihara, S.; Otiende, M.; Bottomley, C.; Mupe, Z. N.; Kagucia, E. W.; Gallagher, K.; Etyang, A.; Voller, S.; Gitonga, J.; Mugo, D.; Agoti, C. N.; Otieno, E.; Ndwiga, L.; Lambe, T.; Wright, D.; Barasa, E.; Tsofa, B.; Bejon, P.; Ochola-Oyier, L. I.; Agweyu, A.; Scott, J. A. G.; Warimwe, G. M. title: Seroprevalence of anti-SARS-CoV-2 IgG antibodies in Kenyan blood donors date: 2020-07-29 journal: nan DOI: 10.1101/2020.07.27.20162693 sha: 8460110c88b62121e561cc00e4576caf7719349a doc_id: 899619 cord_uid: iqq8kmn0 Background There are no data on SARS-CoV-2 seroprevalence in Africa though the COVID-19 epidemic curve and reported mortality differ from patterns seen elsewhere. We estimated the anti-SARS-CoV-2 antibody prevalence among blood donors in Kenya. Methods We measured anti-SARS-CoV-2 spike IgG prevalence by ELISA on residual blood donor samples obtained between April 30 and June 16, 2020. Assay sensitivity and specificity were 83% (95% CI 59, 96%) and 99.0% (95% CI 98.1, 99.5%), respectively. National seroprevalence was estimated using Bayesian multilevel regression and post-stratification to account for non-random sampling with respect to age, sex and region, adjusted for assay performance. Results Complete data were available for 3098 of 3174 donors, aged 15-64 years. By comparison with the Kenyan population, the sample over-represented males (82% versus 49%), adults aged 25-34 years (40% versus 27%) and residents of coastal Counties (49% versus 9%). Crude overall seroprevalence was 5.6% (174/3098). Population-weighted, test-adjusted national seroprevalence was 5.2% (95% CI 3.7, 7.1%). Seroprevalence was highest in the 3 largest urban Counties; Mombasa (9.3% [95% CI 6.4, 13.2%)], Nairobi (8.5% [95% CI 4.9, 13.5%]) and Kisumu (6.5% [95% CI 3.3, 11.2%]). Conclusions We estimate that 1 in 20 adults in Kenya had SARS-CoV-2 antibodies during the study period. By the median date of our survey, only 2093 COVID-19 cases and 71 deaths had been reported through the national screening system. This contrasts, by several orders of magnitude, with the numbers of cases and deaths reported in parts of Europe and America when seroprevalence was similar. -5 - Africa accounts for 17% of the global population, 1 but has reported about 5% of global COVID-19 cases and 2.5% of global COVID-19 deaths to date. 2 This disparity has been variously attributed to limited capacity for diagnosis, timely implementation of stringent containment measures, a younger population structure 3 and a predominance of asymptomatic and mild infections. 4 The first case of COVID-19 in Kenya was detected on March 12, 2020. Within one week the government instituted several containment measures to limit the spread of the virus and the number of fatalities. 5 While the number of new cases continues to rise (14, 168 cases and 250 deaths as at 21 st July 2020) 6 , the rate is notably slower than in Wuhan, China or Europe. The proportion of COVID-19 cases that are asymptomatic varies by setting. 7 However, in countries like Kenya, with a young population, it is likely to be high. 3 In addition, testing algorithms have ascertained only a fraction of the symptomatic cases in most countries. 8 Good public health planning requires information on the potential burden of the pandemic on the healthcare system and the collateral effects of any mitigation measures. In countries affected early, serological surveillance was used to define cumulative incidence. For example, at the release of lockdown in Wuhan, 9.6% Several studies conducted early in the epidemic curve have found seroprevalence substantially below 5%, which emphasises the need for a highly specific antibody assays. 12 IgG and IgM antibodies rise almost simultaneously within the first 1-3 weeks after symptom onset and 90% of cases are seropositive by 14 days. 12 IgG responses are weaker in pauci-symptomatic or asymptomatic cases and may wane rapidly. 13 A wide range of antibody assays have been developed principally against the spike and nucleocapsid proteins of SARS-CoV-2. To facilitate comparisons across studies, the World Health Organisation (WHO) convened a network to share a panel of test sera and evaluate inter-laboratory variation. 14 Physical distancing and movement restrictions imposed as part of the COVID-19 response make random population sampling impracticable and therefore several countries have monitored seroprevalence in convenience samples from blood transfusion donors [15] [16] [17] or expectant mothers attending ante-natal clinics. 18 Here we report the results of a pragmatic national serological survey using residual blood samples from transfusion donors across Kenya. This study was carried out at the KEMRI-Wellcome Trust Research Programme (KWTRP) in Kilifi, Kenya in collaboration with the Kenya National Blood Transfusion Service (KNBTS). The KWTRP is the government designated SARS-CoV-2 testing laboratory for the Coastal Region of Kenya. All rights reserved. No reuse allowed without permission. (which was not certified by peer review) 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 July 29, 2020. conjugated goat anti-human IgG antibody (KPL-SeraCare) diluted 1:10,000 in wash buffer was added to the plates, incubated for 1h at room temperature, washed 3 times All rights reserved. No reuse allowed without permission. (which was not certified by peer review) 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 July 29, 2020. . We explored different approaches to define seropositivity using well-characterised populations from before and during the COVID-19 pandemic ( Table 1, Table S1 , Supplementary text). For both the RBD and spike ELISA specificity was higher when using OD ratios (sample OD/negative control OD) rather than the raw sample OD plus 3 standard deviations. By OD ratio, each ELISA correctly classified 901 of the 910 pre-pandemic samples as seronegative and 15 of the 18 known reverse transcriptase polymerase chain reaction (RT-PCR) positive samples as seropositive (Tables 1, S1 ). However, the spike ELISA detected more seropositive individuals among those who were investigated for SARS-CoV-2 during the pandemic but were RT-PCR negative ( Figure S2 , panels A & B). Based on these data, we defined anti-SARS-CoV-2 IgG seropositivity as an OD ratio >2 and selected the spike ELISA for this study; the sensitivity and specificity of this threshold was 83% (95% CI: 59-96%) and 99.0% (95% CI 98.1-99.5%), respectively (Table 1, Figure S3 panels A & B). -9 -2019 census. We used 2 methods of standardisation; 1) direct standardisation on the observed prevalence and population weights in 80 region-age-sex strata, and 2) multilevel regression and post-stratification (MLRP) adjusted for the sensitivity and specificity of the assay. 21 22 To predict stratum prevalence for MLRP we fitted a Bayesian logistic regression model that included sex as a fixed effect and age and region as random effects. 23 The Bayesian analysis (see Supplement) was done using the rjags package in R version 3.6.1. 24 We used vague or weakly informative priors for all parameters except the sensitivity (83.0%) and specificity (99.0%) which took bounds from previous studies (62.0%-92.0% and 98.6%-99.8%, respectively). 21 . We plotted sample distribution and seroprevalence over time to illustrate potential time-trends and tabulated the distribution and seroprevalence across different donor groups (FRDs vs VNRDs) to explore potential biases. This study was approved by the Scientific and Ethics Review Unit (SERU) of the Kenya Medical Research Institute (Protocol SSC 3426). Before the blood draw, donors gave individual consent the use of their samples for research. All rights reserved. No reuse allowed without permission. (which was not certified by peer review) 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 July 29, 2020. . https://doi.org/10.1101/2020.07.27.20162693 doi: medRxiv preprint A total of 3,174 samples were collected between April 30 and June 16, 2020, from individuals aged 15-66 years at four regional centres; approximately half of the samples were drawn in Mombasa, and the remainder were evenly distributed between Nairobi, Kisumu and Eldoret ( Figure 1 ). We excluded 18 duplicate samples, 56 records missing data on age or collection date and 2 records from individuals aged Overall, 174 of 3098 samples were positive for SARS-CoV-2 anti-Spike IgG giving a crude seroprevalence of 5.6% (95% CI 4.8-6.5%). Crude seroprevalence varied by age (p=0.046), ranging between 3.4-7.0% among adults 15-54 years; all 71 donors aged 55-64 years were seronegative (Table 3) . Crude seroprevalence did not vary significantly by sex (p=0.50) but did by region, ranging from 1.9% in the Rift Valley region to 10.0% in the Western region (p=0.002, Table 3 ). All rights reserved. No reuse allowed without permission. (which was not certified by peer review) 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 July 29, 2020. . The Bayesian population-weighted and test-adjusted seroprevalence for Kenya was 5.2% (95% CI 3.7-7.1%, Table 3 ) and the posterior sensitivity and specificity estimates were 82.5% (95% CI 69.6-91.2%) and 99.2 (95% CI 98.7-99.6%), respectively. Seroprevalence was higher (5.1-6.2%) in the younger age groups years) and then declined in the subsequent age groups (45-64 years) but was similar for both sexes. Seroprevalence was highest for Mombasa, Nairobi and the Western region though the number of observations for the Western region was small. The directly-standardised seroprevalence estimates are presented in Table S2 and the directly-standardised national seroprevalence (5.4%) did not differ substantially from the Bayesian population-weighted, test-adjusted estimate. Seroprevalence was also calculated for Counties that had at least 120 donors sampled. The three largest urban Counties of Mombasa, Nairobi, and Kisumu had SARS-CoV-2 seroprevalence of 9.3% (95% CI 6.4-13.2%), 8.5% (95% CI 4.9-13.5%) and 6.5% (95% CI 3.3-11.2%), respectively (Table S3 ). The frequency of blood donor sampling and crude seroprevalence estimates increased with time over the 7-week study period ( Figure 2 ). The median sample date was May 30, 2020 while the mid-point of the study was May 24, 2020. We did not adjust for sample date because the period of sampling varied for residents of different counties ( Figure 2C ) instead we show the variation in crude prevalence over time (Figure 2A ). The VNRDs were 7.6% (236/3098) of donors while the remainder were FRDs. The two groups did not differ significantly by age (p=0.15) or sex (p=0.51, Table S4 ). Crude seroprevalence was 8.5% (20/236) for VNRDs and 5.4% (154/2862) for FRDs. All rights reserved. No reuse allowed without permission. (which was not certified by peer review) 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 July 29, 2020. . https://doi.org/10.1101/2020.07.27.20162693 doi: medRxiv preprint However, the median sample date for VNRDs (June 14, 2020) was two weeks later than the median sample date for FRDs (May 14, 2020). In this anti-SARS-CoV-2 IgG seroprevalence study of blood donors in Kenya, the crude prevalence was 5.6% and the population-weighted test-adjusted seroprevalence was 5.2%. To the best of our knowledge, this is the first published report of SARS-CoV-2 seroepidemiology in Africa. Surprisingly, population exposure across the country is considerably higher than previously thought. As expected, seroprevalence was highest in the three urban counties; Mombasa (9.3%), Nairobi (8.5%) and Kisumu (6.5%). Consistent with other studies, seroprevalence did not vary significantly by sex; 25 it peaked in 35-44-yearolds and was lowest for those ≥ 45 years, which is also consistent with other studies reporting lower seroprevalence in older adults 10 11 Antibody responses are related to the severity of disease 25 and asymptomatic infections, which elicit the least detectable immune responses, are more common in younger populations 4 26 . In older populations, antibody responses may be modified by pre-existing HCoV-elicited immunity 27 28 and seroprevalence may also be reduced by lower risk of exposure ('shielding' at home). The prevalence of SARS-CoV-2 antibodies in our study is comparable to estimates from large population-based serosurveys in China, Switzerland, Spain and the USA after the initial epidemic peak. 10 11 29 30 Our results are also comparable to other surveys All rights reserved. No reuse allowed without permission. (which was not certified by peer review) 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 July 29, 2020. Serological surveys can also be used to estimate the state of population immunity. However, the low seroprevalence observed in countries that have experienced a substantial first wave of the epidemic implies that relying on herd immunity is unlikely to terminate the pandemic. At the same time, these observations have provoked a lively debate on alternative mechanisms of immunity to SARS-CoV-2 including cell-mediated immunity [33] [34] [35] and the role of pre-existing HCoV-elicited immunity 27 28 as well as the function and duration of antibody responses following All rights reserved. No reuse allowed without permission. (which was not certified by peer review) 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 July 29, 2020. . https://doi.org/10.1101/2020.07.27.20162693 doi: medRxiv preprint infection 13 36-38 . Despite our work showing HCoVs circulate in Kenya 39 , we did not identify evidence of cross-reactive antibodies to endemic coronaviruses in the validation of our assay. Other reports suggest that anti-spike antibody responses may be insensitive as a correlate of exposure to the virus or as a correlate of immunity to the diseases. 40 A key strength of this study is the rigorous validation that included samples from the target population as well as reference plasma from the UK NIBSC as part of a WHO coordinated effort on SARS-CoV-2 seroepidemiology. In addition, we adopted a conservative seropositivity threshold to prioritise specificity over sensitivity. This is relevant in settings like ours, where seroprevalence is under 10% and a similar approach has been implemented by others. 10 Although blood donors are not representative of the Kenyan population as a whole, we adjusted for bias in the sample structure by standardisation against the age, sex, Blood donors may also be unrepresentative of the general population because their risk of exposure differs. This bias could work two ways; volunteers are excluded from giving blood if they have been ill during the last two weeks so the sample may All rights reserved. No reuse allowed without permission. (which was not certified by peer review) 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 July 29, 2020. . https://doi.org/10.1101/2020.07.27.20162693 doi: medRxiv preprint underestimate the population SARS-CoV-2 antibody prevalence; on the other hand, people who are shielding at home are unlikely to be captured in our sample leading to an overestimate of seroprevalence. Our exploration of the two distinct populations of blood donors, FRDs and VNRDs, suggests variation in the seroprevalence by donor group but, of note, 92% (2862/3098) of our sample came from the group with lower seroprevalence. Our study provides the first national and regional estimates of population exposure to SARS-CoV-2 in an African country. The highest prevalence was seen in the 3 urban counties. The results suggest about 1 in 20 people aged 15-64 years have been exposed to SARS-CoV-2 which is in sharp contrast with the very small numbers of COVID-19 cases and deaths reported during the same period. (which was not certified by peer review) 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 July 29, 2020. (which was not certified by peer review) 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 July 29, 2020. . (which was not certified by peer review) 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 July 29, 2020. . United Nations DoEaSA Ministry of Health K. Press Statement on the update of the coronavirus in the country and response measures Ministry of Health K. COVID-19 Situation Reports (SITREP) Estimating the extent of asymptomatic COVID-19 and its potential for community transmission: systematic review and metaanalysis Prevalence of Asymptomatic SARS-CoV-2 Infection: A Narrative Review Serological tests facilitate identification of asymptomatic SARS-CoV-2 infection in Wuhan Prevalence of SARS-CoV-2 in Spain (ENE-COVID): a nationwide, population-based seroepidemiological study Seroprevalence of anti-SARS-CoV-2 IgG antibodies in Geneva, Switzerland (SEROCoV-POP): a population-based study Clinical and immunological assessment of asymptomatic SARS-CoV-2 infections Solidarity II" global serologic study for COVID-19 SARS-CoV-2 seroprevalence trends in healthy blood donors during the COVID-19 Milan outbreak Seroprevalence of anti-SARS-CoV-2 among blood donors in Rio de Janeiro PREPRINT (Version 1) available at Research Square Bayesian analysis of tests with unknown specificity and sensitivity Poststratification into many categories using hierarchical logistic regression Data Analysis Using Regression and Multilevel/Hierarchical Models Bayesian Graphical Models using MCMC. R package version 4-10 A systematic review of antibody mediated immunity to coronaviruses: antibody kinetics, correlates of protection, and association of antibody responses with severity of disease Comparison of Clinical Characteristics of Patients with Asymptomatic vs Symptomatic Coronavirus Disease Pre-existing immunity to SARS-CoV-2: the knowns and unknowns Pre-existing and de novo humoral immunity to SARS-CoV-2 in humans Seroprevalence of Antibodies to SARS-CoV-2 in 10 Sites in the United States Ministry of Health K. COVID-19 Outbreak in Kenya Global, regional, and national estimates of the population at increased risk of severe COVID-19 due to underlying health conditions in 2020: a modelling study Targets of T Cell Responses to SARS-CoV-2 Coronavirus in Humans with COVID-19 Disease and Unexposed Individuals SARS-CoV-2-specific T cell immunity in cases of COVID-19 and SARS, and uninfected controls Robust T cell immunity in convalescent individuals with asymptomatic or mild COVID-19 Rapid Decay of Anti-SARS-CoV-2 -21 Humoral Response Dynamics Following Infection with SARS-CoV-2 Surveillance of endemic human coronaviruses (HCoV-NL63, OC43 and 229E) associated with pneumonia in Kilifi Severe Acute Respiratory Syndrome Coronavirus 2-Specific Antibody Responses in Coronavirus Disease Patients