key: cord-1026266-tksl0v7h
authors: Mostafa, Aya; Kandil, Sahar; El-Sayed, Manal H.; Girgis, Samia; Hafez, Hala; Yosef, Mostafa; Saber, Saly; Ezzelarab, Hoda; Ramadan, Marwa; Algohary, Eman; Fahmy, Gehan; Afifi, Iman; Hassan, Fatmaelzahra; Elsayed, Shaimaa; Reda, Amira; Fattuh, Doaa; Mahmoud, Asmaa; Mansour, Amany; Sabry, Moshira; Habeb, Petra; Ebeid, Fatma S.E.; Elanwar, Ali; Saleh, Ayman; Mansour, Ossama; Omar, Ashraf; El-Meteini, Mahmoud
title: SARS-CoV-2 seroconversion among 4040 Egyptian healthcare workers in 12 resource-limited healthcare facilities: A prospective cohort study
date: 2021-01-20
journal: Int J Infect Dis
DOI: 10.1016/j.ijid.2021.01.037
sha: e1d62aef1f8d42a5cc1139e541475db3c4073c3e
doc_id: 1026266
cord_uid: tksl0v7h

BACKGROUND: We examined SARS-CoV-2 seroconversion incidence and its risk factors 21 days after baseline screening among healthcare workers(HCWs) in a resource-limited setting. METHODS: A prospective cohort study of 4040 HCWs took place between April and June 2020 at 12 university healthcare facilities in Cairo, Egypt. Follow-up exposure and clinical data were collected using an online survey. SARS-CoV-2 testing was done using rapid IgM and IgG serological tests; RT-PCR was done for those with positive serology. Cox proportional hazards modelling was used to estimate adjusted hazard ratios(HR) of seroconversion. RESULTS: Of the 3870/4040(95.8%) HCWs who tested negative for IgM, IgG, and PCR at baseline, 2282(59.0%) returned for follow-up. Seroconversion incidence (positive IgM and/or IgG) was 100/2282(4.4%,95%:CI:3.6-5.3),mostly asymptomatic(64/100,64.0%), with a daily hazard = 0.21%(95%-CI:0.17-0.25)/48746 person-days of follow-up. Seroconversion was 4.0%(64/1596;95%CI:3.1-5.1) among asymptomatic and 5.3%(36/686;95%CI:3.7-7.2) among symptomatic HCWs. Seroconversion was independently associated with older age(HR:2.71,95%CI:1.31-5.58); lower education(HR:3.89,95%CI:1.89-8.00); contact with a confirmed case>15 minutes(HR:2.20,95%CI:1.18-4.10); chronic kidney disease(HR:4.42,95%CI:1.03-18.96); pregnancy(HR:3.5,95%CI:1.05-11.85); change/loss of smell(HR:3.17,95%CI:1.48-6.78); and negatively associated with contact at workplace(HR:0.48,95%CI:0.28-0.82). CONCLUSIONS: Most seroconversions were asymptomatic, emphasizing the need for regular universal testing. Seropositivity was three-fold that observed at baseline. Cumulative infections increased nationally by a similar rate, suggesting that HCWs’ infections reflect community not nosocomial transmission during the first wave of COVID-19 epidemic in Egypt.

As of 15 November 2020, more than 54.5 million cases of coronavirus disease 2019 and 1.3 million deaths have been documented globally (Johns Hopkins University of Medicine, 2020). Severe Acute Respiratory Syndrome Coronavirus 2 (SARS-CoV-2) infection in healthcare workers (HCWs) reports, mainly in high-income settings, varied widely (0.4%-57.1%) with a pooled prevalence of 11% and 7% by reverse transcription-polymerase chain reaction (RT-PCR) and serology testing, respectively (Gómez-Ochoa et al., 2020) .

Studies examining seroconversion in HCW cohorts, particularly in resource-limited settings, are scarce. One UK study reported 20% seroconversion among HCWs within one month of follow-up (Houlihan et al., 2020) . Silent (subclinical or asymptomatic) seroconversion was reported in a limited number of non-HCW (6/9, 67.0%) (Hung et al., 2020) and HCW cohorts (11/25, 44.0%) (Hains et al., 2020) . Asymptomatic SARS-CoV-2 carriage reached up to 68.2% (Mostafa et al., 2020) and 86% (Mandić-Rajčević et al., 2020) among HCWs, contributing to the ongoing community and nosocomial transmission, and challenging the overwhelmed healthcare systems.

Serological testing can help in diagnosis of suspected cases with negative RT-PCR test results and in early identification of asymptomatic infections (Long et al., 2020) . Antibodies specific to SARS-CoV-2 appear within two weeks and peak three to four weeks after infection, (Seow et al., 2020) thus can help detect cases otherwise undetectable during the acute phase. Universal regular screening of HCWs enables prompt detection, isolation, and management of cases, including asymptomatic infections, hence, protecting HCWs and patients.

In Egypt, 110767 total COVID-19 cases have been detected through the national symptom-based testing using PCR and 6453 deaths have occurred as of 15 November 2020 (Ministry of Health and Population, 2020). By mid-October 2020, it was estimated that 3,576 COVID-19 cases (1.6%) and 188 deaths (5.3%) have occurred among 220,000 Egyptian physicians (Alarabiya News, 2020) .

To understand the extent of SARS-CoV-2 infection among a cohort of HCWs in a resource-limited setting, we conducted a prospective investigation including baseline and follow-up screening and risk assessment of HCWs (SARAH: NCT04348214). The baseline screening phase of 4040 HCWs in Ain Shams University (ASU) medical campus, including 12 university healthcare facilities in Cairo, Egypt has been described previously (Mostafa et al., 2020) . The infection proportion among HCWs at baseline (positive IgM, or IgG, or PCR) was 4.2% (170/4040, 95%CI 3.6-4.9). The baseline seroprevalence (positive IgM and/or IgG) was 1.3% (n=53/4040, 95%CI 1.0-1.7). In this paper, we describe the follow-up screening phase. We present the incidence of seroconversion after 21 days of follow-up using rapid serological tests and the associated risk factors among HCWs who tested negative for SARS-CoV-2 by IgM, IgG, and PCR at baseline. We also report the follow-up test results of the 170 HCWs who were infected at baseline.

We conducted a prospective cohort study in 12 ASU healthcare facilities, where follow-up procedures took place between 14 May and 10 June 2020, after 21 days from the baseline screening (World Health Organization, 2020a) . The baseline study design, setting, participants, tools, J o u r n a l P r e -p r o o f procedures, and findings have been previously described (Mostafa et al., 2020) . Baseline screening was conducted between April 22 and May 14 2020 and included an online survey plus laboratory testing to assess SARS-CoV-2 seroprevalence. Follow-up screening also included an online survey to assess the occurrence of symptoms and exposure to a confirmed case during the follow-up period plus laboratory testing to assess the incidence of SARS-CoV-2 seroconversion.

All HCWs who participated at baseline were informed about the follow-up screening procedures and schedule, for which they had provided written informed consent during the baseline recruitment process. They were all invited to participate at the follow-up phase with no exclusion criteria, i.e. follow-up screening targeted all hospital staff regardless of their baseline test results, the type of care they provided (clinical or non-clinical), presence of symptoms (asymptomatic HCWs were included), or place of work within the healthcare facility (no units/wards were prioritized for follow-up screening). This study was approved by the Ethics Review Committee, Faculty of Medicine, ASU (FMASUP18b/2020).

Participation was voluntary and all HCWs on-the-job were informed and scheduled for follow-up screening through formal announcements by their healthcare facility administration. Also, all participants received phone call reminders to attend their scheduled follow-up appointments and were re-scheduled within a period of three days to provide flexibility according to their workload if needed. At each of the 12 healthcare facilities, HCWs visited the two workstations dedicated for follow-up screening procedures. Teams consisting of the hospital management, infection control focal points, nurses, technicians, and administrative staff coordinated the screening activities.

J o u r n a l P r e -p r o o f Workstation 1. Online survey: HCWs completed the follow-up survey through available PCs with internet connection or their cellphones to reduce the screening time and unnecessary contact with shared surfaces. Staff were available to interview HCWs who needed assistance in filling the survey. Data were entered using the assigned study ID to facilitate anonymous linkage with laboratory results.

Workstation 2. Laboratory sampling: At follow-up, a 5-ml venous blood sample was collected by venipuncture into a plain vacutainer for qualitative detection of SARS-CoV-2 IgM and IgG antibodies. For those who tested positive for serology, an appointment was scheduled for RT-PCR, where combined nasal and oropharyngeal swabs were collected in a single tube containing viral transport medium for detecting viral RNA by RT-PCR (Centers for Disease Control and Prevention, 2020a). Swabs were transported to the laboratory in ice boxes at 4C.

Questions were adapted from relevant World Health Organization protocols and interim guidance (World Health Organization, 2020b; World Health Organization, 2020c) . Details of the questionnaire development and data on background characteristics were previously described (Mostafa et al., 2020) . The follow-up survey consisted of two sections including questions on exposures and symptoms that occurred in the period since baseline screening:

J o u r n a l P r e -p r o o f Section 1: HCWs were provided an exhaustive list of symptoms suggestive of COVID-19 and were asked to report the occurrence, type, and severity of symptoms since baseline screening: fever <38C, fever ≥38C, chills, fatigue, muscle pain, joint pain, sore throat, dry cough, cough with sputum, runny nose/nasal congestion, sneezing, shortness of breath, wheezing, chest pain, other respiratory symptoms, nausea, vomiting, abdominal discomfort/pain, diarrhea, headache, confusion, loss of appetite, change/loss of taste or smell, skin rash, conjunctivitis.

Section 2: HCWs reported community and nosocomial exposure through close contact with a confirmed COVID-19 case and when and where the contact had occurred. In the nosocomial exposure section, we asked about direct care provision for a confirmed case, face-to-face contact within 2 meters and the duration of contact.

Detection of SARS-CoV-2 IgM and IgG antibodies was done using the lateral flow immunochromatographic assay Artron® One Step COVID-19 IgM/IgG Antibody Test, the same method used at baseline (Artron Laboratories Inc., Canada). The assay has estimated sensitivity and specificity of 83.3% and 100% , respectively by Lassaunière et al. (Lassaunière et al., 2020) and 94.4% and 98.2% respectively by Zhang and Zheng, 2020 (Zhang and Zheng, 2020) . RT-PCR testing was done only for seropositive cases at follow-up, using the same baseline technique previously described (Mostafa et al., 2020 ; Centers for Disease Control and Prevention, 2020a).

Case definition: Seroconversion was defined as having a positive test result at the follow-up phase for IgM and/or IgG among HCWs who tested negative for all three tests (IgM, IgG and PCR) at baseline (n=3870) (Figure 1 ).

For describing the study sample, we calculated median and interquartile ranges (IQR) for continuous variables or counts and proportions for categorical variables. To calculate the incidence proportion and the hazard of seroconversion, the number of IgM and/or IgG positive HCWs was divided by the number of participants and person-days of follow-up, respectively. Multivariable Cox proportional hazard regression modelling was used to determine the independent determinants of seroconversion and to estimate the adjusted hazard ratios and 95% confidence intervals (CI).

Effect estimates with confidence intervals and exact P-values are provided. All analyses were done using SPSS version 25.

Between 14 May and 10 June 2020, 2282 (59.0%) HCWs returned for testing at follow-up out of the 3870 HCWs who tested negative at baseline for IgM, IgG, and PCR ( Figure 1 ). The median follow-up period was 25 days (IQR 21-27). The incidence of SARS-CoV-2 seroconversion (positive IgM and/or IgG) was 100/2282 (4.4%, 95%CI 3.6-5.3) and the daily hazard was 0.21% (95%CI 0.17-0.25) over 48746 person-days of follow-up. Of the 100 seroconverted HCWs, 59

(59.0%) tested positive only for IgM, 6 (6.0%) tested positive only for IgG, and 11 (11.0%) tested positive for both IgM and IgG. Twenty-one (21.0%) tested positive for both PCR and IgM, one J o u r n a l P r e -p r o o f (1.0%) tested positive for both PCR and IgG, and two (2.0%) had all three tests positive ( Figure   1 ).

The overall follow-up group's description is presented in Table 1 Only 36/100 (36.0%) seroconverters reported being symptomatic during the follow-up period;

3/36 (8.3%) reported severe symptoms (Table 2) . Experiencing symptoms suggestive of SARS-CoV-2 infection were more frequently reported during follow-up than at baseline (Figure 2) , such as sneezing, muscle pain, change/loss of smell, shortness of breath, change/loss of taste, abdominal discomfort, loss of appetite, and conjunctivitis. Symptoms that were not reported at baseline, but only at the follow-up stage among seroconverters were chest pain, other respiratory symptoms, vomiting, wheezing, and dizziness ( Figure 2 ). The incidence of seroconversion was 4.0%

(64/1596, 95%CI 3.1-5.1) among asymptomatic and 5.3% among symptomatic HCWs (36/686, 95%CI 3.7-7.2).

Overall, HCWs reported contact with a confirmed case more frequently at follow-up (1131/2282, 49.6%) than at baseline (1098/4040, 27.2%) Also, a higher proportion of exposures at home/residential area (5/39, 12.8% vs 2/68, 2.9%) and providing direct care to a confirmed case J o u r n a l P r e -p r o o f (29/100, 29.0% versus 22/170, 12.9%) were observed among seroconverters at follow-up than among infected HCWs at baseline (Supplementary Table 2 ).

In the bivariate analysis, a higher risk of seroconversion was associated with the following factors:

older age, female gender, marriage, lower level of education, working as a nurse or non-clinical care, presence of a pre-existing medical condition, exposure to a confirmed case at home/residential area, and a duration of contact >15 minutes with a confirmed case during health care provision (Tables 1, 2). The pre-existing medical conditions that were associated with a higher risk of seroconversion were diabetes, chronic kidney disease, and pregnancy Table 3 ). Seroconversion risk was not associated with the history of contact with a confirmed case, provision of direct care to a confirmed case, the interaction with a confirmed case within 2 meters, being symptomatic, or the severity of symptoms (Tables 1, 2) . When analyzed separately, the following symptomst were associated with a higher risk of seroconversion: fever <38°C, muscle pain, joint pain, sneezing, shortness of breath, other respiratory symptoms, loss of appetite, change/loss of taste, change/loss of smell, and conjunctivitis (Supplementary Table 4 ).

In the multivariable Cox proportional hazards model, the risk of seroconversion among HCWs of older age was a double to triple the risk among those aged 18-29 years. HCWs with lower educational levels were also at about double or triple the risk of seroconversion compared with HCWs with university or higher education. HCWs who reported contact with a confirmed case within the workplace were at a lower risk of seroconversion compared with those who reported contact at home/residential area (adjusted HR 0.48, 95%CI 0.28-0.82). During health care J o u r n a l P r e -p r o o f provision, it is the longer time of exposure rather than the close contact within 2 meters that increased the risk of seroconversion. Exposure to a confirmed case for >15 minutes doubled the risk of seroconversion as compared to exposure for <5 minutes (adjusted HR 2.20, 95%CI 1.18-4.10). Having chronic kidney disease and being pregnant were independently associated with seroconversion (adjusted HR were 4.42, 95%CI 1.03-18.96 and 3.52, 95% CI 1.05-11.85, respectively). Out of an exhaustive list of symptoms, change/loss of smell was the only symptom that predicted seroconversion (adjusted HR 3.17, 95%CI 1.48-6.78) ( Table 3) .

Apart from this follow-up group of HCWs (n=2282) who tested negative for all three tests (IgM, IgG, and PCR) at baseline, 73/170 (42.9%) participants who tested positive for one or more of these three tests at baseline returned for follow-up. Details of their baseline and follow-up test results are described in Figure 3 and Supplementary Figure 1 . Notably, 30/45 (66.7%) who were PCR-positive only at baseline did not develop antibodies (Figure 3, Supplementary Figure 1) .

The median follow-up period for this group was 26 days (IQR 25-29).

The incidence of seroconversion for SARS-Co-V2 in this HCW cohort was 4.4% (5.3% in symptomatic and 4.0% in asymptomatic HCWs) using rapid serological IgM and IgG tests among those who tested negative (IgM, IgG, and PCR) for SARS-Co-V2 at baseline and returned after 3 weeks for follow-up screening. Seropositivity at follow-up (100/2282, 4.4%) was three-fold the seroprevalence observed at baseline (53/4040, 1.3%). During the follow-up phase, cumulative infections in Egypt have also increased by a similar rate, approximately tripling from 10,829 to May to 10 June 2020 and ended just 10 days before the peak of the first wave. The highest number of daily cases (1774) detected through the national symptom-based PCR testing was recorded on 20 June. After that point, the daily number of cases in Egypt started to decline reaching the rate detected at the launch of baseline screening about 6 weeks later (Figure 4) . The same mitigation measures adopted at ASU Hospitals and nationally continued during the follow-up phase. This suggests that infections among HCWs might continue to reflect community rather than nosocomial transmission during the first wave of the COVID-19 epidemic in Egypt.

Approximately two-thirds of the seroconversions in the current study were silent. Asymptomatic infections were also high (68.2%) in this HCW cohort at baseline (Mostafa et al., 2020) . Similarly, 6/9 (67%) of the quarantined diamond cruise ship infected passengers remained asymptomatic for 14-days, although they seroconverted and had a high viral load (Hung et al., 2020) . Subclinical seroconversion has been reported after 3 weeks among 11/25 (44.0%) asymptomatic HCWs in a pediatric dialysis unit in Indiana, USA; one HCW seroconverted on day 21 despite three negative PCR test results (Hains et al., 2020) . In Norway, serology testing of household members six weeks after the index case had a positive PCR test revealed that detecting seroconversion might more accurately identify attack rates in households than intermittent PCR testing (Cox et al., 2020) .

These findings reinforce the need for expanding regular universal testing using a combination of PCR and serology in finding cases and tracing contacts, particularly in settings with high rates of asymptomatic infections, to enable prompt isolation and management.

Consistent with our baseline screening findings, (Mostafa et al., 2020) most seroconverted HCWs were nurses (54/100, 54.0%). However, being a nurse was not an independent determinant of seroconversion. Also, more infected HCWs providing non-clinical care tested positive at followup (33.0%) than at baseline (16.5%); HCWs' exposure to COVID-19 patients through community transmission during the follow-up phase when cases were soaring in Egypt may be a plausible explanation for the increased proportion of positive tests among these HCW groups. A metaanalysis of COVID-19 in HCWs reported that those working in hospital nonemergency wards and nurses were the most commonly infected personnel (Gómez-Ochoa et al., 2020). Among this Egyptian HCW cohort, reporting exposure to a confirmed case doubled from a quarter at baseline to half of the overall participants at follow-up. However, the proportion reporting contact with a confirmed case among the infected HCWs remained the same (about two-fifths).

Nosocomial exposure does not seem to be associated with seroconversion in this Egyptian HCW cohort because two-thirds of those who seroconverted did not report any obvious epidemiological link with a confirmed case. Also, the incidence of seroconversion among HCWs reporting household exposure was five-fold that reported from exposure at the workplace. However, among a cohort of patient-facing HCWs in London, UK, a much higher seroprevalence (25%) at enrolment and seroconversion rate (20%) within one month of follow-up were reported with a peak of HCW cases between 30 March and 5 Aprila rate at least twice that of the general population, (Houlihan et al., 2020) indicating the possibility of nosocomial rather than community transmission at that stage of the epidemic in London. Nguyen et al also reported that highest infection rates among their UK HCW cohort study were around London and the Midlands . The difference in the magnitude of infections among the UK and the Egyptian HCW cohorts may depend on how much overlap occurred between study timeframes and the epidemic curves in the two countries.

A weak or even lack of seroconversion after initial PCR positivity was reported among some COVID-19 patients during variable periods of follow-up, leaving these individuals susceptible to infection or relapse (Papachristodoulou et al. 2020) . For example, 22% of 27 patients did not develop a serologic response after 60 days of follow-up in Germany; (Fill Malfertheiner et al., 2020) 20% of 173 Chinese patients did not develop IgG after two weeks of onset; (Zhao et al., 2020) 6% did not show any antibody response two weeks after hospital discharge and 30% of 175 patients showed very low neutralizing antibody titres in Shanghai, China; and one of 14 (7%) HCWs had not seroconverted after 17 days of follow-up in the UK. (Houlihan et al., 2020) Among this Egyptian HCW cohort, 66.7% of 45 seronegative HCWs who tested PCRpositive only at baseline did not seroconvert after three weeks of follow-up. Furthermore, there was weak IgG seroconversion among the seropositive group at baseline after 3 weeks of followup: only 2/20 (10.0%) developed IgG, 11/20 (55.0%) cleared IgM without developing IgG, and 6/8 (75.0%) cleared IgG antibodies. However, the sensitivity of the rapid serological test used in this study was low (83.3%), thus high false negatives rate may have been encountered. The identification of false-negative results among HCWs and the current variation in performance of point-of-care serological assays requires caution in interpreting IgG results. (Pallett et al., 2020) Per 1000 tested, Deeks et al. assumed that if the sensitivity of IgG/IgM was 91.4% at 15 to 21 days, at a prevalence of 5%, 4 cases (3 to 7) would be missed. (Deeks et al., 2020) Despite this concern, our findings have important implications for the level of long-term immunity and vaccine development in settings with predominantly asymptomatic SARS-CoV-2 infections.

HCWs ≥50 years old had approximately a three-fold risk of seroconversion compared with HCWs of younger age groups. Also, older HCWs were reported to have higher IgG titres in an Italian HCW cohort (Calcagno et al., 2020) and a higher seroprevalence in Idaho, the USA (Bryan et al., 2020) . Having chronic kidney disease was associated with a four-fold risk of seroconversion among HCWs in the current study. Being older than 40 years and having chronic kidney disease were independent determinants of having a positive SARS-CoV-2 test among 587 primary care patients in the UK (de Lusignan et al., 2020) . Pregnant HCWs were at a four-fold risk of seroconversion in the current study compared to nonpregnant HCWs. Pregnant women in the USA were reported to be at risk for severe COVID-19 illness and its complications (Centers for Disease Control and Prevention, 2020b). 28 Change/loss of smell independently predicted seroconversion among HCWs in this study at follow-up and seropositivity at baseline. Similarly, recent loss of smell was associated with a three-fold increase in the risk of seropositivity for SARS-CoV-2 in a community-based cohort in the UK (Makaronidis et al., 2020) . Also, a meta-analysis reported that olfactory dysfunctions were found in 41% of 8438 COVID-19 patients from 13 countries (Agyeman et al., 2020) . Being a highly specific symptom of COVID-19, it could guide early testing and case isolation.

Approximately two-fifths of the HCWs in this longitudinal study were lost to follow-up and some did not return for their PCR appointments. HCWs were redistributed among newly established isolation wards/hospitals during the follow-up phase that preceded the peak of the epidemic in Egypt, therefore, many were not able to return for follow-up. Also, some HCWs were quarantined or self-isolating. Second, in contrast to the baseline phase, PCR testing was not conducted universally for HCWs at follow-up; instead, it was restricted to those with positive serology. This decision was taken as a precautionary measure to preserve the PCR stock and meet the expected increase in testing needs, given the foreseen rise in hospital admissions at that stage of the epidemic in Egypt. As a result, some infected HCWs who were in the early stage of infection may have been missed using serological tests alone. PCR is costly and requires certain skill in sample collection.

These challenges may affect the quality and sustainability for universal regular testing of HCWs.

To avoid critical delay in diagnosis, management, and quarantine, using serological tests (in isolation or combined with PCR when possible)especially rapid point-of-care tests that are continuously being developed with higher accuracy to meet the parallel development in vaccines (Pallett et al., 2020; Pickering et al., 2020; Amanat et al., 2020 )-may be a more reasonable assay choice for high-throughput use, less workload, and a faster turnaround time in resource-limited settings. Third, despite the published study reports addressing the performance of the used serological assay testing kits (Zhang and Zheng, May 21, 2020; Artron Laboratories, Inc. Oct. 28, 2020) , the commercial manufacturer voluntarily withdraw the test from the U.S Food and Drug Administration (FDA) notification list of commercial manufacturers distributing serology test kits under the Policy for Coronavirus Disease-2019 Tests and hence, included in the FDA removal list on 25 May 2020. This action may be due to not providing the necessary documentation in a timely manner, as could be concluded from the dates of reports publishing. The two reports documented sensitivity of 94.4% and 93.3%; and specificity of 98.2% and 100% (Zhang and Zheng, May 21, 2020; Artron Laboratories, Inc. Oct. 28, 2020) . Based on the evidence we have; the used testing kits has an acceptable level of validity that most probably did not jeopardize the validity of our study results. 

Smell and Taste Dysfunction in Patients With COVID-19: A Systematic Review and Meta-analysis

Syndicate: 3576 cases and 188 deaths of Corona among Egyptian physicians

A serological assay to detect SARS-CoV-2 seroconversion in humans

IgG/IgM Antibody Test

Performance Characteristics of the Abbott Architect SARS-CoV-2 IgG Assay and Seroprevalence in

Risk of SARS-CoV-2 infection in healthcare workers

Centers for Disease Control and Prevention 2020a. Coronavirus Disease

Resources for Labs. Interim Guidelines for Collecting, Handling, and Testing Clinical Specimens for COVID-19

Centers for Disease Control and Prevention 2020b. Coronavirus Disease

Cases, data & surveillance. Data on COVID-19 during Pregnancy: Severity of Maternal Illness

Seroconversion in household members of COVID-19 outpatients

Risk factors for SARSCoV-2 among patients in the Oxford Royal College of General Practitioners Research and Surveillance Centre primary care network: a cross-sectional study

Antibody tests for identification of current and past infection with SARS-CoV-2. Cochrane Database Syst Rev

Immune response to SARS-CoV-2 in health care workers following a COVID-19 outbreak: A prospective longitudinal study

COVID-19 in Healthcare Workers: A Living Systematic Review and Meta-analysis of Prevalence, Risk Factors, Clinical Characteristics, and Outcomes

Asymptomatic Seroconversion of Immunoglobulins to SARS-CoV-2 in a Pediatric Dialysis Unit

Pandemic peak SARS-CoV-2 infection and seroconversion rates in London frontline health-care workers

seroconversion among passengers quarantined after disembarking a cruise ship: a case series

Evaluation of Nine Commercial SARS-CoV-2 Immunoassays. medRxiv

Antibody responses to SARS-CoV-2 in patients with COVID-19

Seroprevalence of SARS-CoV-2 antibodies in people with an acute loss in their sense of smell and/or taste in a community-based population in London, UK: An observational cohort study

Contact tracing and isolation of asymptomatic spreaders to successfully control the COVID-19 epidemic among in Milan (Italy)

Universal COVID-19 screening of 4040 health care workers in a resource-limited setting: an Egyptian pilot model in a university with 12 public hospitals and medical centers

Risk of COVID-19 among front-line health-care workers and the general community: a prospective cohort study

Point-ofcare serological assays for delayed SARS-CoV-2 case identification among health-care workers in the UK: a prospective multicentre cohort study

Long-term and herd immunity against SARS-CoV-2: implications from current and past knowledge

Comparative assessment of multiple COVID-19 serological technologies supports continued evaluation of point-of-care lateral flow assays in hospital and community healthcare settings

Longitudinal observation and decline of neutralizing antibody responses in the three months following SARS-CoV-2 infection in humans

Population-based age-stratified seroepidemiological investigation protocol for COVID-19 virus infection

World Health Organization 2020b. Risk Assessment and Management of Exposure of Health Care Workers in the Context of COVID-19

Protocol for Assessment of Potential Risk Factors for 2019-novel Coronavirus (2019-nCoV) Infection Among Health Care Workers in a Health Care Setting 2020c

Evaluating the Association of Clinical Characteristics With Neutralizing Antibody Levels in Patients Who Have Recovered From Mild COVID-19 in

IgM/IgG Antibody Test Diagnostic Sensitivity and Specificity Study Report

Antibody responses to SARS-CoV-2 in patients of novel coronavirus disease 2019