SARS-CoV-2 seroprevalence in the urban population of Qatar: An analysis of antibody testing on a sample of 112,941 individuals

Summary The study objective was to the assess level of detectable severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) antibodies in the urban population of Qatar. Antibody testing was performed on residual blood specimens for 112,941 individuals (∼10% of Qatar's urban population) attending for routine/other clinical care between May 12 and September 9, 2020. Seropositivity was 13.3% (95% confidence interval [CI] = 13.1–13.6%) and was independently associated with sex, age, nationality, clinical care encounter type, and testing date. Median optical density (antibody titer) among antibody-positive persons was 27.0 (range = 1.0–150.0), with higher values associated with age, nationality, clinical care encounter type, and testing date. Seropositivity by nationality was positively correlated with the likelihood of having higher antibody titers (Pearson correlation coefficient = 0.85; 95% CI = 0.47–0.96). Less than two in every 10 individuals in Qatar's urban population had detectable antibodies against SARS-CoV-2, suggesting this population is still far from herd immunity and at risk of subsequent infection waves. Higher antibody titer appears to be a biomarker of repeated exposures to the infection.

Less than 20% of Qatar's urban population had detectable antibodies for SARS-CoV-2 Qatar's urban population is still at risk of future epidemic waves Antibody titers varied by age, nationality, and over time Higher antibody titers appear to be a biomarker of re-exposures to the infection

INTRODUCTION
With the breakthrough development of highly efficacious vaccines against the severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) (Polack et al., 2020;Jackson et al., 2020;Voysey et al., 2020), determining the population's cumulative infection exposure and current immunity level is critical to inform national vaccine roll-out strategies.
Qatar, located in the Arabian Peninsula, with a multinational population of 2.8 million people, nearly all living in the capital city, Doha, had a significant first wave of coronavirus disease 2019 (COVID-19) that peaked in late May 2020 (Planning and Statistics Authority-State of Qatar, 2020; Planning and Statistics Authority-State of Qatar, 2019). As of December 23, 2020, >60,000 infections per million population (number of infections divided by total population) had been laboratory confirmed (Hamad Medical Corporation, 2020b;Ministry of Public Health-State of Qatar, 2021). Qatar has a unique socio-demographic structure, in which single-unit and family households, including children, adults, and/or older adults, account for only 40% of the total population, with adults in this ''urban population'' often being part of the professional or service workforce (Planning and Statistics Authority-State of Qatar, 2020; Ministry of Interior-State of Qatar, 2020; Planning and Statistics Authority-State of Qatar, 2017). The remaining 60% of the population consists of craft and manual workers (CMWs) (Planning and Statistics Authority-State of Qatar, 2020; Ministry of Interior-State of Qatar, 2020; Planning and Statistics Authority-State of Qatar, 2017)-mostly single, young men working in development projects (Planning and Statistics Authority-State of Qatar, 2017) and typically living in large, shared accommodations (De Bel-Air, 2018).
Infection transmission in Qatar was first documented among CMWs on March 6, 2020 (Al , who were subsequently most affected by this epidemic (Abu-Raddad et al., 2021a). A recently completed nationwide, population-based survey assessing ''every'' infection among the CMW population found that six out of every ten persons had detectable antibodies against SARS-CoV-2 (Al-Thani et al., 2021;Jeremijenko et al., 2021), suggesting that this population is at or near herd immunity for the variants circulating at this time (Jeremijenko et al., 2021;Anderson et al., 2020;Britton et al., 2020). In the present study, the first objective was to assess the level of infection exposure among the rest of the population of Qatar, that of the ''urban population'' of this country. The urban population was defined as the complement of the CMW population, that is, the population that lives in single-unit or family households (not shared accommodations) and is part of the professional or service workforce. This part of the population is highly diverse and includes over 150 nationalities in addition to Qataris. The second objective was to identify predictors for infection and for having higher antibody titers.

RESULTS
In all, 112,941 individuals were tested for SARS-CoV-2 antibodies, representing 10% of the urban population of Qatar (Planning and Statistics Authority-State of Qatar, 2020) (Table 1). Of these, 51.6% were men. Two-thirds (66%) of tested persons were 20-49 years of age. Qatari (25.8%) and Indian nationals (16.5%) were most heavily represented in the sample, reflecting their representation in the urban population (Ministry of Interior-State of Qatar, 2020; Priya Dsouza Communications, 2019; Planning and Statistics Authority-State of Qatar, 2019). Blood specimens were collected in the course of routine clinical care during home care visits (34.2%), outpatient visits (28.5%), inpatient hospital stays (21.0%), and emergency department visits (16.4%). Overall, the sample mirrored the urban population demographics (Table S1 of Figure S1 of supplemental information shows seropositivity month by month, which was largely stable over the study duration. Figure 1 illustrates the distribution of antibody titers (optical density values) among the 18,844 antibodypositive persons. Optical density values ranged from 1.0 to 150.0 with a median of 27.0. Having higher antibody titers than the median was not associated with sex, but in the multivariable regression analysis, they were independently associated with age, nationality, clinical care type, and the calendar date of the antibody test (Table 2). Compared to those aged 20-29 years, the AOR was higher in children <10 years and adults aged 40-79 years. There were significant differences by nationality. AOR was 1.68 (95% CI: Qataris. Compared to emergency department attendees, inpatients had an AOR for higher antibody positivity of 0.38 (95% CI: 0.34-0.43), while no difference was found for outpatients or for patients with home care visits or follow-up consultations. Having higher antibody titers increased with time (Tables 2 and S3 of supplemental information), with an AOR (per day) of 1.011 (95% CI: 1.010-1.013; Table 2).

DISCUSSION
The above results indicate that <20% of the urban population of Qatar, which constitutes 40% of the total population and includes nearly all older adults, manifests evidence of prior infection. This seroprevalence is substantially less than that in the CMW part of the population, which was estimated recently in a nationwide survey at 55.3% (Al-Thani et al., 2021), with the leading risk factor being living in a large shared accommodation (Al-Thani et al., 2021;Abu-Raddad et al., 2021a;Jeremijenko et al., 2021;Al Kuwari et al., 2020).
This finding suggests that the lockdown and imposed social and physical distancing restrictions have been more successful in slowing transmission in the urban population compared to the CMW population. Building on the totality of evidence on the Qatar epidemic (Abu-Raddad et al., 2021a; Al Al-Thani et al., 2021;Ayoub et al., 2021;Jeremijenko et al., 2021), this appears to be due to the dwelling structure, in that the urban population lives mostly in single-unit or family households that each includes a small number of individuals. Meanwhile, the CMW population lives mostly in large shared accommodations that each includes a large number of individuals. While the lockdown forced individuals to stay more at place of residence, it is typical to have more social contacts every day in a large shared accommodation than in a single-unit or family household. This outcome highlights the role of the ''boarding school'' effect in respiratory infection transmission, seen often in the intense influenza outbreaks in regular and boarding schools (Jackson et al., 2013;Glatman-Freedman et al., 2012). This effect has been also seen in the large SARS-CoV-2 outbreaks in nursing homes in Europe and the United States (Arons et al., 2020;Burton et al., 2020;Ladhani et al., 2020).
With a seroprevalence of <20%, the urban population of Qatar remains far below the herd immunity threshold, estimated at 60-70% infection exposure (Anderson et al., 2020;Britton et al., 2020;Jeremijenko et al., 2021). Accordingly, there should exist potential for subsequent waves of infection in this part of the population, but no second wave materialized from May 2020 up to end of this year (Abu-Raddad et al., 2021a; Ayoub et al., 2021), that is, before the introduction and expansion of the B.1.1.7 and B.1.351 variants of concern (Abu-Raddad et al., 2021b). On the contrary, only a slow increase in seroprevalence has occurred since the peak of the first wave (Table S2 of   iScience Article lockdown and easing of many social distancing restrictions, may be explained by an ''immunity shield'' effect (Weitz et al., 2020) arising from the social mixing between the urban and CMW populations. With most CMWs being immune, the infection transmission chains had difficulty in sustaining themselves, as they were interrupted by the presence of immune persons who were not getting reinfected (Abu-Raddad et al.,  (Ayoub et al., 2021).
There were significant differences in seropositivity by sex, age, and nationality. These are probably not due to biological differences but to differences in the likelihood of exposure to the infection. Indeed, a small proportion of the specimens tested in this study belonged to CMWs who had a higher risk of exposure to the infection than the urban population (Al-Thani et al., 2021;Jeremijenko et al., 2021). While Hamad Medical Corporation (HMC) provides healthcare primarily to the urban population and other providers cater to the CMW population, HMC is a main tertiary care center in Qatar and was also the nationally designated provider for COVID-19 healthcare needs. Thus, it is likely that a small proportion of specimens, which cannot be estimated precisely, were drawn from CMWs who were hospitalized for COVID-19 or other reasons. This may explain the higher antibody positivity of young Bangladeshi, Indian, and Nepalese men (Table 1), who form the bulk of the CMW population (Al-Thani et al., 2021;Jeremijenko et al., 2021). This may also explain the higher seroprevalence in the blood specimens drawn during inpatient or emergency clinical care, which are more likely to be COVID-19 related, than those drawn during outpatient or home care/follow-up consultation clinical care (Table 1). The higher exposure among men aged 20-69 years probably reflects their more frequent work and other activities outside the home, whereas men aged R70 years, urged through public health messaging to remain at home, were more likely to do so, out of concern about infection severity.
The proportion of those antibody positive who had a PCR-confirmed diagnosis prior to the antibody-positive test was 47.1% (Table 3), much higher than the 9.3% in the CMW population (Al-Thani et al., 2021), and that estimated for the total population of Qatar (11.6%) (Ayoub et al., 2021). This is probably because study specimens were drawn from individuals receiving healthcare, including those hospitalized for COVID-19, people more likely to have been tested for the infection. This fact, along with the difference in age structure between the urban and CMW populations (Planning and Statistics Authority-State of Qatar, 2020; Ministry of Interior-State of Qatar, 2020; Planning and Statistics Authority-State of Qatar, 2017; Al-Thani et al., 2021), may have resulted in higher estimates of infection severity, criticality, and fatality rates in this study (Table  3), compared to the study of the CMW population (Al-Thani et al., 2021), or model predictions for the entire population of Qatar (Seedat et al., 2020).
Strikingly, having a higher antibody titer varied by nationality, clinical care type, and time (Table 2). Variation by nationality is probably an indirect biomarker of re-exposure to infection, resulting in repeated immune system  iScience Article reactivation. This is suggested by the very strong positive correlation between the odds of having a higher antibody titer and seroprevalence across the nationalities (Figure 2). Lower antibody titers were found in inpatients, but this may reflect COVID-19 hospitalizations for recent infections so that there was not sufficient time for higher antibody titers to develop. There was a trend of increasing ''higher antibody titers'' over time, which may reflect the growing pool of infected persons who have had more time to develop higher levels of detectable antibodies after infection or alternatively to being re-exposed to the infection.

Limitations of the study
This study has some limitations. The sample included individuals attending HMC for routine or other clinical care, but this population may not necessarily be representative of the wider urban population of Qatar. Though HMC is the main public healthcare provider and is widely accessible at minimal cost to nationals and residents, the sample may still have missed persons inclined to seek healthcare in the private sector or who avoided contact with healthcare during the COVID-19 pandemic. Some specimens may have been drawn from CMWs, who were not part of our intended study sample. However, the large sample size, equivalent to 10% of the urban population of Qatar, as well as the probabilistic weighting used in the analysis may have reduced inherent biases in our sample.
Laboratory methods were based on high quality, validated commercial platforms, such as the Roche platform used for serological testing (The Roche Group, 2020;Jahrsdö rfer et al., 2020). The Roche platform is one of the most extensively used and investigated commercial platforms, with a specificity R99.8% (The Roche Group, 2020;Public Health England, 2020;Oved et al., 2020) and a sensitivity R89% (Jahrsdö rfer et al., 2020;Abu-Raddad et al., 2021a;Oved et al., 2020). However, it is possible that the less than perfect sensitivity, especially for those with recent infections, may have underestimated the actual seroprevalence as it may take up to few weeks before recently infected individuals develop antibodies at detectable levels (Nasrallah et al., 2020;Wajnberg et al., 2020). Indeed, a recent investigation of the performance of three automated, commercial, serological platforms in Qatar, including the Roche platform, found that each of them missed R20% of individuals with past or current infections (Nasrallah et al., 2020). Factoring the less than perfect sensitivity and specificity (Sempos and Tian, 2021) would have increased the measured seroprevalence to 14.8% instead of 13.3%. The fatality rate may have been underestimated in this study as some individuals may have died before being tested or before developing detectable antibody levels. In conclusion, fewer than two in every 10 individuals in the urban population of Qatar had detectable antibodies against SARS-CoV-2, suggesting that this population is still well below the herd immunity threshold and is potentially at risk from a subsequent epidemic wave. This emphasizes the need to maintain current social and physical distancing restrictions while SARS-CoV-2 vaccinations are being scaled up throughout the country. The findings also suggest that higher antibody titers appear to be a biomarker of repeated exposures to the infection.

STAR+METHODS
Detailed methods are provided in the online version of this paper and include the following:

ACKNOWLEDGMENTS
We thank Her Excellency Dr. Hanan Al Kuwari, Minister of Public Health, for her vision, guidance, leadership, and support. We also thank Dr. Saad Al Kaabi, Chair of the System Wide Incident Command and Control (SWICC) Committee for the COVID-19 national healthcare response, for his leadership, for analytical insights, and for his instrumental role in enacting data information systems that made these studies possible. We further extend our appreciation to SWICC Committee and Scientific Reference and Research Taskforce (SRRT) members for their informative input, scientific technical advice, and enriching discussions. We also thank Dr. Mariam Abdulmalik, CEO of the Primary Health Care Corporation and the Chairperson of the Tactical Community Command Group on COVID-19, as well as members of this committee, for providing support to the teams that worked on the field surveillance. We further thank Dr. Nahla Afifi, Director of Qatar Biobank (QBB), Ms. Tasneem Al-Hamad, Ms. Eiman Al-Khayat, and the rest of the QBB team for their unwavering support in retrieving and analyzing samples and in compiling and generating databases for COVID-19 infection, as well as Dr. Asmaa Al-Thani, Chairperson of the Qatar Genome Programme Committee and Board Vice Chairperson of QBB, for her leadership of this effort. We also acknowledge the dedicated efforts of the Clinical Coding Team and the COVID-19 Mortality Review Team, both at Hamad Medical Corporation, and the Surveillance Team at the Ministry of Public Health. The authors are grateful for support from Hamad Medical Corporation, the Ministry of Public Health, and the Biomedical Research Program and the Biostatistics, Epidemiology, and Biomathematics Research Core, both at Weill Cornell Medicine-Qatar. The statements made herein are solely the responsibility of the authors.

AUTHOR CONTRIBUTIONS
P.V.C. conceived and designed this study and led its implementation and antibody testing. H.C. managed the databases, performed the statistical data analyses, and co-wrote the first draft of the manuscript. L.J.A. led the statistical analyses and co-wrote the first draft of the article. All authors contributed to development of the study protocol, data collection, and acquisition, database development, discussions and interpretation of the results, and to the writing of the manuscript. All authors have read and approved the final manuscript. iScience 24, 102646, June 25, 2021 iScience Article