Prevalence and risk factors of SARS-CoV-2 antibody responses among healthcare workers (June 2020–November 2021)

Abstract Background To assess SARS-CoV-2 seroprevalence in healthcare workers (HCW) with sampling in June and October 2020 and April and November 2021. Methods Observational and prospective study in 2455 HCW with serum sampling. Antibodies against SARS-CoV-2 nucleocapsid and occupational, social and health risk factors were assessed at each time point. Results Seropositivity for SARS-CoV-2 in HCW increased from 11.8% in June 2020 to 28.4% in November 2021. Of those with a positive test in June 2020, 92.1% remained with a positive test, 6.7% had an indeterminate test and 1.1% had a negative test in November 2021. Non-diagnosed carriers represented 28.6% in June 2020 and 14.6% in November 2021. Nurses and nursing assistants showed the highest prevalence of seropositivity. Close contact (at home or in the hospital) with Covid-19 cases without protection and working in the frontline were the main risk factors. A total of 88.8% HCW were vaccinated, all with a positive serological response in April 2021, but levels of antibodies decreased about 65%, and two vaccinated persons presented a negative serological test against spike protein in November 2021. Levels of spike antibodies were higher in those vaccinated with Moderna compared with Pfizer and the percentage of antibody reduction was higher with Pfizer vaccine. Conclusions This study shows that seroprevalence of SARS-CoV-2 antibodies among HCW doubled that of the general population and that protection both at the workplace and in the socio-familial field was associated with a lower risk of infection, which stabilized after vaccination.


Introduction
T he first confirmed case of COVID-19 in Spain was 31 January 2020.4][5][6][7][8][9] A recent systematic review of SARS-CoV-2 seroprevalence studies in Europe included data on HCW and the community and showed a wide heterogeneity in the reported prevalence of seroconversions within and between these two populations. 10Regarding HCW, studies from Spain reported a seroprevalence very close to 10% between March and August 2020 [3][4][5] but some studies reported higher seroprevalences in other European countries. 8Community studies performed in Europe during the same period reported very heterogeneous results, ranging from 0.42% in Greece 11 to 23% in Italy 12 but, in general, lower seroprevalences in comparison to HCW. 10 Results of a nationwide, population-based, seroepidemiological study in Spain (ENE-COVID), performed between April and June 2020, reported a seroprevalence of 5.2%, and a third of individuals who had developed SARS-Cov-2 antibodies remained asymptomatic. 13The data are quite different because populations, geographic areas, epidemic situations, study dates and serological techniques vary between studies.
It is essential to know the risk factors for HCW to contract COVID-19 in order to design effective preventive measures.Likewise, assessing the prevalence of non-diagnosed carriers is crucial in order to develop effective strategies to prevent the spread of the disease from healthcare centers.In June 2020, the Maresme Health Consortium (CSdM), a medical consortium that manages a general hospital, a subacute hospital, a nursing home and three primary care centers, launched a prospective study to monitor seroprevalence against SARS-CoV-2 among its workers.This study not only determined SARS-CoV-2 antibodies at several time points but also the rate of negativizations and the evolution of SARS-CoV-2 antibodies after vaccination.
The main objectives of this study were: (i) to assess the evolution of the prevalence of exposure to SARS-CoV-2 of the 2455 professionals working at CSdM, Catalonia, Spain, over time (June 2020-November 2021); (ii) to determine the prevalence of non-diagnosed carriers; (iii) to explore occupational, social and health risk factors associated with SARS-CoV-2 infection; and (iv) to monitor the antibody responses following mRNA vaccination.

Study design and population
This observational and prospective study followed the cohort of all 2455 active CSdM workers from June 2020 to November 2021.All of them were informed about the study and invited to participate by email and by announcements on the CSdM website.Once workers gave their informed consent, they answered an electronic selfadministered questionnaire on personal sociodemographic, clinical and labor characteristics through the CSdM website, and scheduled a blood extraction for serological tests.This procedure was carried out in June 2020, October 2020, April 2021 and November 2021.

Data gathering
The self-administered electronic questionnaire on risk factors for COVID-19 included sociodemographic characteristics, health status, comorbidities and chronic medication, current symptoms and labor characteristics, which included exposure to COVID-19 patients and use of protective measures.A summary of the epidemiological survey is presented in Supplementary appendix S1.The venous blood samples obtained in June 2020 and October 2020 were sent to the laboratory for the following serological tests: (i) chemiluminescence Immunoassay (CLIA) (Elecsys Cobas, Roche Diagnostics) to detect SARS-CoV-2 nucleocapsid-protein antibodies and, for those positive, (ii) a qualitative enzyme-linked Immunosorbent Assay (ELISA) to differentiate IgG from IgM and IgA antibodies for SARS-CoV-2 (COVID-19 ELISA IgG G1032; COVID-19 ELISA IgMþIgA MA1032, Vircell Microbiologists).In the case of positive IgMþIgA, a nasopharyngeal swab was obtained for the real-time reverse transcriptase-polymerase chain reaction (TaqPath COVID-19 CE-IVD RT-PCR, Thermo Fisher Scientific, Pleasanton, CA, USA) to detect asymptomatic carriers of the virus and participants who had had COVID-19 and still harbored the virus.If the results of the total antibodies CLIA test were undetermined and ELISA IgMþIgA and IgG were also undetermined, the serology test was repeated 10 days later.The venous blood samples obtained in April and November 2021 were sent to the laboratory for the following serological tests: (i) the same CLIA to detect SARS-CoV-2 nucleocapsid-protein antibodies as in June and October 2020 and (ii) a test to quantify SARS-CoV-2 spike-protein antibodies (DiaSorin LIAISON, TrimericS IgG assay) to determine post-vaccine seroconversion.All serum aliquots were stored in a sample collection registered in the 'Registro Nacional de Biobancos' (https://biobancos.isciii.es/ListadoColecciones.aspx) under the code: C0006111.The data that support the findings of this study are not openly available due to confidentiality norms, but are available from the corresponding author upon reasonable request and Ethical Committee approval.

Data analysis
Seroprevalence against SARS-CoV-2, expressed as percentages and 95% confidence intervals (95% CI), and the prevalence of nondiagnosed carriers in June 2020, October 2020, April 2021 and November 2021 were estimated.An individual was considered positive if CLIA was positive (independently of IgA/IgM and IgG results), negative if CLIA was negative, indeterminate if CLIA was indeterminate, and false positive if CLIA was positive and both IgA/ IgM and IgG were negative.The range of measurement of IgG SARS-CoV-2 SPIKE-TRIMERIC was 4.81-2080 BAU/ml (measurements >33.8 BAU/ml were considered positive).Cross-sectional (June 2020) and longitudinal (June 2020-April 2021) exploratory analyses of the study factors associated with a positive serological test result were performed using the Chi-square test or Fisher exact test and bivariate and multivariate logistic regression analyses.All variables associated with a positive serological test in the bivariate analysis (P-values <0.10) were considered for the multivariate analysis using the stepwise method.Gender was also included in the multivariate model (although not reaching a P-value <0.10 in the bivariate analysis).Factors related to non-diagnosed carriers in June 2020 were also analyzed using the same above-mentioned statistical tests (Chisquare or Fisher exact test and logistic regression analysis).Immunological vaccination response was assessed by describing the prevalence of HCW with positive antibodies against SARS-CoV-2 spike protein (as a percentage and 95% CI) in April 2021 and November 2021.Levels of anti-spike protein antibodies were compared between April 21 and November 21 using the Wilcoxon test for paired data.As multiple comparisons were made, the Bonferroni correction was applied and statistical significance was established at a P-value <0.0014.Statistical analysis was performed with the IBM SPSS Statistics software (version 28.0.0.0).

Description of the study sample
Figure 1 presents the flow chart of participants in the study over time and indicates that the study cohort was dynamic with some workers leaving (retirements) and others entering (new incorporations) throughout the study period.At baseline (June 2020), the mean age of the study sample was 42.3 (612.0) years (SD) and 1328 (78.5%) were women.Clinical, social and labor characteristics of the study sample are presented in table 1.Details on comorbidities and medication of the study sample are presented in Supplementary appendix S2.

Immunological response after mRNA vaccination
Mean (SD) levels of antibodies against spike protein were 1617.6 (555.0)BAU/ml among correctly vaccinated HCW who had not suffered COVID-19.All presented a positive serological test (>33.8BAU/ml) against the spike protein of SARS-Cov2.Among vaccinated HCW who had not suffered COVID-19 since April 2021, mean levels of antibodies against spike protein decreased from 1613.8 BAU/ml (555.0) in April 2021 to 589.9 BAU/ml (523.1) in November 2021 (P < 0.001).Two vaccinated persons presented a negative serological test against spike protein in November 2021, both vaccinated with Pzifer.Figure 3 presents mean levels of antibodies against spike protein according to the type of vaccine in vaccinated HCW who had not suffered COVID-19.Levels of spike antibodies were higher in those vaccinated with Moderna compared with those vaccinated with Pfizer, both in April and November 2021 (P < 0.001 in both cases).The percentage of reduction in the level of antibodies was also higher in HCW that received the Pfizer vaccine compared with Moderna (65.1% vs. 48.7%,respectively, P ¼ 0.005).

Discussion
Main study results indicate that the prevalence of SARS-CoV-2 infection among HCW increased from 11.8% in June 2020 to 28.4% in November 2021 and affected, above all, HCW in direct contact with COVID-19 patients (nurses, nursing assistants and physicians).Results also indicate a higher risk of COVID-19 among nurses, nurse assistants and physicians compared with other HCW, and that adherence to protective measures may prevent SARS-CoV-2 infection.Approximately a quarter of professionals who were infected with SARS-CoV-2 had not been diagnosed, which would have favored the spread of the virus and some of the in-hospital outbreaks observed during the study period.The vaccine generated a positive immune response in all cases, which remained positive in 99.9% of cases 7 months later.However, anti-spike antibody levels decreased by 65% after 7 months, with still unknown clinical relevance.The growth of SARS-CoV-2 infection decreased sharply after the mRNA vaccines, which suggests their effectiveness.
4][5] A meta-analysis that included 25 seroprevalence studies worldwide for the year 2020 in HCW showed results that varied from 1.3% to 31.6% (mean 8%). 6he ENE-COVID was a nationwide population-based study aimed to estimate the seroprevalence of SARS-CoV-2 infection in Spain, which was carried out from 27 April to 11 May 2020. 13It showed a seroprevalence of 4.6% for the overall Spanish population and 6.8% for the population of the province of Barcelona. 4These data indicate that the seroprevalence of SARS-CoV-2 in HCW nearly doubled that of the general population.We have found only one longitudinal study that followed a cohort of workers during the first 2 years of the pandemia.It is a study with retrospective data collection from primary care electronic clinical notes in Barcelona (Spain), and shows an accumulated incidence of SARS-CoV-2 infection (from 20 March to 21 September) of 28% in HCW and of 14% in all other occupations. 7This result is consistent with our finding.Seroprevalence of CSdM HCW increased over time but became stabilized after April 2021, probably due to the vaccination program initiated in January and February 2021.8][9] These data raise the issue of systematic and periodic screening among HCW in an epidemic situation of high transmissibility to prevent spread among them and to preserve the functioning of the healthcare system.
The main factors associated with a positive serological test were to work as a physician, nurse or nurse assistant in the frontline with  COVID-19 patients, and to be in close contact without protection with family members infected by SARS-Cov-2.The occupations with the highest overall seroprevalence were nurses and nursing assistants, who are the professional groups with the closest and most prolonged contact with patients.Likewise, unprotected contact with relatives with COVID-19 seems to have a similar effect, so it should be recommended not to relax protection and isolation measures at home.However, some differences were observed between study phases for the factors related to a positive serological test.Feeling unprotected working in the frontline or contact with patients without protection were factors related to a positive serological test in June 2020, but not in April 2021.At the beginning of the pandemic there was a shortage of appropriate material and protective equipment, which could explain why the feeling of being unprotected at work were associated with COVID-19 in June 2020 but not in April 2021.These results suggest the effectiveness of protective measures in the prevention of SARS-CoV-2 infection, but also highlight that this effectiveness is not absolute and that, despite such measures, HCW with the closest contact with COVID-19 patients had a more than 2-fold increase in the risk of SARS-CoV-2 infection over other HCW.A review showed that gender and inadequate/lack of protective personal equipment performing tracheal intubation were major risk factors of COVID-19 in HCW. 14 On the other hand, tobacco consumption showed an independent association with SARS-CoV-2 infection with an OR of 0.45.6][17] However, it cannot be ruled out that this is a spurious association.There is controversy over the effect of smoking on SARS-CoV-2 infection, so further studies are needed to evaluate and clarify it.Furthermore, there is growing evidence that smokers have worse outcomes after contracting the virus than non-smokers, 18 and a large-scale observational study suggested a causal effect of smoking on COVID-19 severity. 19More aged professionals were less likely to become infected, which may be explained by longer professional experience and better knowledge and use of preventive measures, which further reinforce the importance of knowledge, training, use and supervision of protection and prevention measures in the workplace.
Regarding SARS-CoV-2 vaccination, its acceptance by CSdM HCW was high (88%) and similar or slightly higher than that presented by other authors (between 70% and 80%). 20All vaccinated individuals presented positive anti-spike antibody levels, which is consistent with the results published by other authors. 21We observed that anti-spike antibody levels were higher among workers vaccinated with Moderna than those vaccinated with Pfizer, a difference that has also been reported in other studies. 22However, the clinical relevance of this difference is not known.Neutralizing antibody levels are predictive of greater immune protection from symptomatic infection, 23 which suggests that higher levels may have some clinical relevance, especially in terms of maintaining longer-term protection as antibody levels decrease over time.Thus, at 7 months of vaccination, there was a 65% drop in anti-spike antibody levels in those vaccinated with Moderna and 68% in those vaccinated with Pfizer.Despite this, anti-spike antibodies remained positive in 99.9% of vaccinated individuals at 7 months follow up.These results agree with those presented by Doria et al. 24 in relation to the Moderna vaccine.Finally, it should be noted that the seroprevalence of SARS-CoV-2 infection increased by 130% (from 11% to 27% of HCW) between June 2020 and April 2021 (before vaccination) but increased by <4% (from 27% to 28%) between April and November 2021 (after vaccination), which strongly suggests the effectiveness of mRNA vaccines. 25he relatively large sample size, high participation rate in all phases of the study with four blood samples over time, high vaccination acceptance and follow up of the study cohort for 17 months (6-7 months before and 10-11 months after vaccination) are important strengths of the study.However, it has some limitations.The fact that the study had an open cohort (with entrances and exits) made it possible to obtain an accurate picture of the epidemiological characteristics of the pandemic among CSdM HCW in real life, but made it difficult to assess risk factors that showed changes over time.Moreover, this is not a clinical trial but an observational study, so it is not the best design to assess the efficacy of a vaccine but it allowed us to assess its effectiveness in terms of immunological response in real conditions.The epidemic has also undergone changes over time, with outbreaks and virus mutations with variants with different clinical and epidemiological characteristics.We have presented epidemiological data in the context of the epidemic and its various outbreaks in our territory until the appearance of the omicron variant, which has radically changed the clinical features of COVID-19.With omicron, a new scenario arises in which epidemiological surveillance of new variants, new vaccinations or booster doses of vaccine, protection of the most vulnerable (such as aged or immunocompromised) or the search for new effective oral antivirals are some of the main challenges to face.Nevertheless, continuing to monitor the evolution of the immunological response against SARS-CoV-2 among HCW is important from the occupational health point of view and will be useful to assess immunological response protection against COVID-19.

Figure 1
Figure 1 Flow chart of HCW participants in the seroprevalence study.S þ Q, serological test þ questionnaire; OS, only serological test; OQ, only questionnaire; S, serological test (with and without questionnaire)

Figure 2
Figure 2 Prevalence of positive serological tests against the nucleocapsid proteins of SARS-Cov2 by professional categories at each time point

Figure 3
Figure 3 Mean levels of antibodies against spike protein according to the type of vaccine in vaccinated HCW who had not suffered COVID-19.***P < 0.001 Participation rates in the seroprevalence study were 77.9%, 77.1%, 84.3% and 68.9% of the total target population, respectively.Most CSdM workers (88.8%) received two doses of vaccine from January to March 2021, all of them with mRNA vaccines (74.4% Moderna COVID-19 Vaccine, Moderna Biotech Spain, 25.6% Comirnaty BioNTech, Pfizer).A figure in Supplementary material gives the epidemiological time context, showing the number of confirmed cases of COVID-19 treated at Mataro ´hospital from March 2020 to January 2022, the various phases of the study and the vaccination periods (first, second and third doses).The study protocol was approved by the Ethics Committee (CEIm, CSdM 56/20).The study was conducted according to the principles and rules stated in the Declaration of Helsinki and following the regulations established by Spanish biomedical research law (LIB 14/2007), the Spanish law of protection of personal data (LOPD 3/2018) and Regulation (EU) 2016/679 of the EU Parliament and of the Council of 27th April 2016.The study was registered in the ClinicalTrials.govwebsite under code: NCT04425759.

Table 1
Factors associated with a positive serological test on June 20.Bivariate analysis