Incidence and risk factors of SARS-CoV-2 infection among workers in a public health laboratory in Tunisia

Abstract

The aim of this study was to measure the extent of severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) infection among workers at the Institut Pasteur de Tunis (IPT), a public health laboratory involved in the management of the COVID-19 pandemic in Tunisia, and to identify risk factors for infection in this occupational setting. A cross-sectional survey was conducted on IPT workers not vaccinated against coronavirus disease 2019 (COVID-19). Participants completed a questionnaire that included a history of reverse transcription-polymerase chain reaction (RT-PCR)-confirmed SARS-CoV-2 infection. Immunoglobulin G antibodies against the receptor-binding domain of the spike antigen (anti-S-RBD IgG) and the nucleocapsid protein (anti-N IgG) of the SARS-CoV-2 virus were detected by enzyme-linked immunoassay (ELISA). A multivariate analysis was used to identify factors significantly associated with SARS-CoV-2 infection. A total of 428 workers were enrolled in the study. The prevalence of anti-S-RBD and/or anti-N IgG antibodies was 32.9% [28.7–37.4]. The cumulative incidence of SARS-CoV-2 infection (positive serology and/or previous positive RT-PCR test) was 40.0% [35.5–44.9], while the proportion with asymptomatic infection was 32.9%. One-third of the participants with RT-PCR-confirmed infection tested seronegative more than 90 days postinfection. Participants aged over 40 and laborers were more susceptible to infection (adjusted OR [AOR] = 1.65 [1.08–2.51] and AOR = 2.67 [1.45–4.89], respectively), while tobacco smokers had a lower risk of infection (AOR = 0.54 [0.29–0.97]). The SARS-CoV-2 infection rate among IPT workers was not significantly different from that detected concurrently in the general population. Hence, the professional activities conducted in this public health laboratory did not generate additional risk to that incurred outside the institute in day-to-day activities.

Introduction

Health care workers (HCWs), especially those in contact with and/or who care for coronavirus disease 2019 (COVID-19) patients, have an increased risk of severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) infection compared to the general population [1]. Data collected by the World Health Organization (WHO) during the early stages of the pandemic, primarily from European and American countries, indicated that nearly 14% of COVID-19 cases were among health workers [1]. These figures were mainly based on reverse transcription polymerase chain reaction (RT-PCR) testing and most likely underestimated the true extent of infection. In fact, many infected people are asymptomatic or only express minor symptoms and therefore do not get tested [2]. Thus, it is necessary to estimate the invisible ‘iceberg’ of mild infections to assess the real burden of SARS-CoV-2 infections [3].

Seroepidemiological studies are key to understanding the extent of infection in a population. However, uncertainties in estimates may be related to test performance or the proportion of infected persons who do not develop antibodies or whose antibody levels fall under the cutoff value at the time of testing [4–7]. The real extent of SARS-CoV-2 infection could be better estimated by measuring the cumulative incidence of infection based on serology and RT-PCR testing, a goal mainly jeopardized by the unavailability of mass testing in countries that lack resources.

Institut Pasteur de Tunis (IPT) is a scientific research institution focused on human health that is primarily involved in the investigation of infectious diseases. Public health laboratory activities and diagnostic tests for outpatients are also performed at the institute. IPT has been involved in the surveillance and monitoring of the COVID-19 pandemic in Tunisia since its inception, by establishing a protocol for the molecular diagnosis of SARS-CoV-2 infection. This has resulted in a heavy workload, with many IPT staff members volunteering to help the virology department by collecting nasopharyngeal swabs from consultants and travelers, managing the sample circuit, and performing RT-PCR testing. Therefore, a committee was appointed within IPT to monitor and mitigate the risk of infection among staff members. A nasal swab RT-PCR test is offered to any person presenting symptoms suggestive of COVID-19 or having had contact with a positive case, and the test is repeated periodically in some departments of IPT. Although many cases of SARS-CoV-2 infection have been detected among IPT staff, even in those not involved in COVID-19 activities, the actual degree of spread of the infection among the workforce is unknown.

The aims of this study were to use PCR and serology to measure the cumulative incidence of SARS-CoV-2 among IPT staff from the start of the pandemic in Tunisia (March 2020) until the end of the second wave and the start of the third wave in the country (March 2021), to determine the proportion of asymptomatic infections, and to study risk factors for infection.

Materials and methods

Study design and study population

A cross-sectional survey was conducted during the last two weeks of March 2021, which coincided with the end of the second wave of the COVID-19 pandemic in Tunisia and the start of the third wave, prior to widespread vaccination among HCWs in the country. All IPT staff and PhD students who had not received the COVID-19 vaccine were invited to participate in the survey. All Pasteurian community members with an institutional email address were informed of the study objectives and modality of participation in an email sent by the principal investigator. In addition, to ensure a high participation rate, trained investigators visited the IPT departments before the effective start date of data collection to meet eligible persons and raise awareness of study participation.

Sample and data collection

After providing written informed consent, each participant underwent venous blood sampling for SARS-CoV-2 antibody testing and was asked to complete a paper questionnaire. Trained interviewers administered the questionnaire to individuals who had difficulty completing the questionnaire by themselves. The questionnaire inquired about demographic information, professional activities, previous COVID-19 testing and results, history of COVID-19 or signs suggestive of the disease, exposure to SARS-CoV-2, training in COVID-19 prevention and control, and compliance with prevention measures. The reported data concerning previously confirmed SARS-CoV-2 infection were verified for all participants based on COVID-19 records held by the committee appointed to monitor the risk of infection among IPT staff. All precautions were taken to preserve the confidentiality of data collected from study participants. An identification number was assigned to each participant for the labeling of questionnaires and biological samples. The link between this identification number and the identity of participants was maintained only by the principal investigator, who was responsible for informing them about their serology results.

Laboratory testing

To assess the prevalence of SARS-CoV-2 antibodies, we used two enzyme-linked immunoassays (ELISA) developed in-house, one of each that targets antibodies to the nucleocapsid (N) protein and the receptor-binding domain of the spike antigen (S-RBD) of SARS-CoV-2 [8]. In brief, the ELISA assays were optimized using sera from 108 COVID-19 patients and 72 pre-pandemic patients. A receiver operating curve showed that the N- and S-RBD-based ELISA assays had very high performances (AUC: 0.966 and 0.98, respectively; p < 0.0001). This resulted in a sensitivity of 94% and specificity of 93% for the anti-N test and a sensitivity of 95% and specificity of 93% for the anti-S-RBD test.

Statistical analysis

EpiInfo software version 7.2.2.6 (developed by the Centers for Disease Control and Prevention, Atlanta, GA, USA) was used for data entry and analysis. Continuous variables are presented as the mean ± standard deviation (SD), and categorical variables as counts and percentages. The crude prevalence of SARS-CoV-2 antibodies was estimated. Then, the adjusted prevalence according to the performance of the assays used was calculated using the following formula [9]:

$$\text{Adjusted prevalence}=\frac{\text{crude prevalence}+\text{specificity}-1}{\text{sensitivity}+\text{specificity}-1}$$

The cumulative incidence of SARS-CoV-2 infection was estimated as the proportion of participants who tested positive for IgG antibodies (anti-S-RBD IgG and/or anti-N IgG) and/or who reported a previous positive RT-PCR test. Fisher’s exact χ² test was used for univariate analysis, and a backward stepwise logistic regression was used for multivariate analysis to identify factors significantly associated with SARS-CoV-2 infection. SARS-CoV-2 infection was defined as a positive result by serology and/or a previous RT-PCR test. Variables with a p-value less than or equal to 0.25 in the univariate analysis were included in the multivariate analysis.

Results

Among the 592 IPT employees (all professional categories included) and PhD students, 36 who had received the COVID-19 vaccine were excluded, and 556 eligible individuals were approached to participate in the survey. Among the eligible individuals, 128 refused to participate and 428 were included, giving a response rate of approximately 77%. Participants were aged between 22 and 69 years, and the mean age was 41.7 ± 10.5 years. Nearly two-thirds of participants were female, 33.1% were scientists and medics, and 22% had a chronic disease. The main diseases reported were diabetes (5.7%), hypertension (5.1%), chronic respiratory disease (3.5%), and heart disease (3.1%). Approximately one-fifth of those surveyed (17.8%) had participated in COVID-19 activities within the IPT, and 26.4% reported contact with outpatients or human biological specimens. Only 10.5% of participants attended training sessions on preventive measures concerning COVID-19, while 66.0% reported that they always or usually adhered to these measures. More than two-thirds (72.1%) of respondents had previous contact with a confirmed COVID-19 case, and only 9.8% had travelled outside Tunisia since December 2019. The characteristics of the participants are detailed in Table 1.

Table 1.

Main characteristics of participants (N = 428)

	n (%)
Gender (N = 428)
Men	141 (32.9)
Women	287 (67.1)
Age class (years, N = 426)
20–39	198 (46.5)
40–59	209 (49.1)
≥60	19 (4.4)
Occupation (N = 426)
Scientists and medics	141 (33.1)
Technicians	130 (30.5)
Administrative staff	65 (15.3)
Laborers	90 (21.1)
Chronic diseases (N = 428)
Yes	94 (22.0)
No	334 (78.0)
Tobacco consumption (N = 422)
Yes	85 (20.1)
No	337 (79.9)
Alcohol consumption (N = 419)
Yes	35 (8.4)
No	384 (91.6)
Involvement in COVID-19 activities within the IPT (N = 426)
Yes	76 (17.8)
No	350 (82.2)
Contact with outpatients or their biological specimens (N = 424)
Yes	112 (26.4)
No	312 (73.6)
Training in COVID-19 prevention and control (N = 420)
Yes	44 (10.5)
No	376 (89.5)
Compliance with preventive measures (N = 427)
Always/ usually	282 (66.0)
Never/ occasionally	145 (34.0)
Contact with a confirmed case of COVID-19 (N = 394)
Yes	284 (72.1)
No	110 (27.9)
Travel outside Tunisia since December 2019 (N = 428)
Yes	42 (9.8)
No	386 (90.2)

Table 2 shows the distribution of survey participants according to the results of serology and previous RT-PCR testing. More than a quarter (25.7%) of respondents reported that they were previously diagnosed with COVID-19 by RT-PCR. Among the RT-PCR-confirmed cases, 30 individuals tested negative for both anti-S and anti-N antibodies (27.3%).

Table 2.

Distribution of participants according to results of serology and previous RT-PCR testing (N = 428)

	Anti-S+ & anti-N− n (%)	Anti-N+ & anti-S- n (%)	Anti-S+ & anti-N+ n (%)	Anti-S− & anti-N− n (%)	Total N
No previous RT-PCR testing	8 (11.9)	2 (3.0)	9 (13.5)	48 (71.6)	67
Previous RT-PCR testing negative	23 (9.2)	5 (2.0)	14 (5.5)	209 (83.3)	251
Previous RT-PCR testing positive	15 (13.6)	0 (0)	65 (59.1)	30 (27.3)	110

The crude prevalence of SARS-CoV-2 antibodies (anti-N IgG and/or anti-S-RBD IgG) was 32.9% [28.7–37.4]; 22.2% [18.2–25.9] of participants had anti-N IgG antibodies, and 31.3% [26.9–35.7] had anti-S-RBD IgG antibodies (Table 3). The adjusted prevalence of anti-N and anti-S antibodies was 17.5% [12.9–21.7] and 27.6% [22.6–32.6], respectively.

Table 3.

Results of serology and previous RT-PCR tests among study participants (N = 428)

	n	% (CI 95%)
Presence of anti-N antibodies	95	22.2 [18.2–25.9]
Presence of anti-S-RBD antibodies	134	31.3 [26.9–35.7]
Presence of anti-N and/or anti-S-RBD antibodies	141	32.9 [28.7–37.4]
Previous positive RT-PCR	110	25.7 [21.7–29.9]
Positive serology and/or RT-PCR	171	40.0 [35.5–44.9]

CI: Confidence interval

The cumulative incidence of SARS-CoV-2 infection among our study population was 40.0% [35.5–44.9]. Among previously infected individuals, the proportion with asymptomatic infection was 33.3%.

For RT-PCR-confirmed cases, the prevalence of antibodies against SARS-CoV-2 (anti-N IgG and/or anti-S-RBD IgG) was 72.7%; the prevalence of anti-S-RBD IgG antibodies was 72.7%, and that of anti-N IgG antibodies was 59.1%. The elapsed time between the date of COVID-19 diagnosis by RT-PCR and the date of serology testing in the present study ranged from 18 to 213 days (median duration = 106.5 days, interquartile range = 86.0–174.2). The prevalence of anti-N antibodies decreased significantly over time (p = 0.016), while the prevalence of anti-S-RBD antibodies did not differ significantly with regard to time since COVID-19 diagnosis. Among participants who had been diagnosed with COVID-19 more than three months prior, less than half (48.5%) had anti-N IgG antibodies, while 66.2% had anti-S-RBD IgG antibodies. More details are shown in Table 4.

Table 4.

Serology results according to the time period between COVID-19 diagnosis date and serology blood sampling date among participants previously diagnosed with COVID-19 (positive RT-PCR) (N = 110)

Duration (days)	N	Presence of anti-N IgG n (%)	p-value	Presence of anti-S-RBD IgG n (%)	p-value	Absence of anti-N & anti S-RBD IgG n (%)
0–90	42	32 (76.2)	0.016	35 (83.3)	0.118	7 (16.7)
91–180	47	23 (48.9)		30 (63.8)		17 (36.2)
> 180	21	10 (47.6)		15 (71.4)		6 ( 28.6)
Total	110	65 (59.1)		80 (72.7)		30 (27.3)

In the bivariate analysis, the odds of SARS-CoV-2 infection were higher among participants aged over 40 years (crude OR [COR] = 1.57 [1.06–2.33]) and among laborers compared to scientists and medics (COR = 2.13 [1.24–3.67]). Tobacco smokers and those who reported alcohol consumption had lower odds of SARS-CoV-2 infection (COR = 0.52 [0.31–0.87] and COR = 0.41 [0.18–0.94], respectively) (Table 5). SARS-CoV-2 infection was not significantly associated with other factors, including involvement in COVID-19 activities within the IPT, training in preventive measures, and the degree of compliance with these measures (Table 5).

Table 5.

Risk factors for SARS-CoV-2 infection: results of univariate and multivariate* analysis

	SARS-CoV-2 infection n (%)	Crude OR (95% CI)	p-value	Adjusted OR (95% CI)	p-value
Gender			0.208		0.076
Men	50 (35.5)	0.75 [0.49–1.14]		0.63 [0.38–1.5]
Women	121 (42.2)	1		1
Age class (years)			0.029		0.019
≥40	103 (45.2)	1.57 [1.06–2.33]		1.65 [1.08–2.51]
<40	68 (34.3)	1		1
Occupation			0.045		0.014
Laborers	45 (50.0)	2.13 [1.24–3.67]	0.006	2.67 [1.45–4.89]	0.002
Technicians	55 (42.3)	1.56 [0.95–2.57]	0.077	1.66 [0.99–2.80]	0.054
Administrative staff	24 (36.9)	1.25 [0.67–2.31]	0.479	1.29 [0.68–2.45]	0.425
Scientists	45 (31.9)	1		1
Chronic diseases			0.634
Yes	40 (42.6)	1.15 [0.72–1.83]
No	131 (39.2)	1
Tobacco consumption			0.013		0.039
Yes	24 (28.2)	0.52 [0.31–0.87]		0.54 [0.29–0.97]
No	145 (43.0)	1		1
Alcohol consumption			0.031		0.432
Yes	8 (22.9)	0.41 [0.18–0.94]		0.70 [0.29–1.69]
No	160 (41.7)	1		1
Involvement in COVID-19 activities within the IPT			0.365
Yes	34 (44.7)	1.29 [0.78–2.13]
No	135 (38.6)	1
Contact with outpatients or their biological specimens			0.312
Yes	49 (43.8)	1.26 [0.81–1.95]
No	119 (38.1)	1
Training in COVID-19 prevention and control			0.329
No	152 (40.4)	1.45 [0.76–2.83]
Yes	14 (31.8)	1
Compliance with preventive measures			0.404
Never/ occasionally	62 (42.8)	1.20 [0.80–1.81]
Always/ usually	108 (38.3)	1
Contact with a confirmed COVID-19 case			0.646
Yes	114 (40.1)	1.13 [0.72–1.77]
No	41 (37.3)	1
Travel outside Tunisia since December 2019			0.869
Yes	16 (38.1)	0.91 [0.47–1.76]
No	155 (40.2)	1

CI: Confidence interval

Variables introduced in the initial model of multivariate analysis were gender, age class, occupation, tobacco consumption, and alcohol consumption.

*Variables introduced in the initial model of multivariate analysis were gender, age class, occupation, tobacco consumption, and alcohol consumption.

In the multivariate analysis, only age, occupation, and tobacco status were significantly associated with the risk of SARS-CoV2 infection. Specifically, participants aged over 40 years and laborers were more susceptible to infection (adjusted OR [AOR] = 1.65 [1.08–2.51] and AOR = 2.67 [1.45–4.89], respectively), while tobacco smokers had a lower risk of infection (AOR = 0.54 [0.29–0.97]) (Table 5).

Discussion

In the present study, we assessed the cumulative incidence of SARS-CoV-2 infection based on RT-PCR and serology as well as related risk factors among IPT staff and PhD students. Our results revealed a relatively high incidence of infection among the study population (40.0% [35.5–44.9]). Nearly one-third of confirmed infections (positive serology and/or RT-PCR) were asymptomatic. Laborers had a higher risk of SARS-CoV-2 infection, while a lower risk of infection was found among young adults and smokers.

Among the surveyed population, the prevalence of SARS-CoV-2 antibodies was 32.9% [28.7–37.4]. A study carried out by our team on the general population of the city of Tunis during the same period and using the same serological tests [10] found a comparable prevalence of SARS-CoV-2 antibodies in the age range of 22 to 69 years (34.7% [32.0-37.7]). These results indicate that the risk of infection in the specific setting of a public health laboratory such as IPT is not different from that in the community. This information is encouraging because it implies that the professional activities of the participants, who were in contact with outpatients, other consultants, and infected samples, did not generate any additional risk beyond that incurred outside the institute in day-to-day life; this was confirmed by the univariate and multivariate analysis of our data. Indeed, in our study, COVID-19 circuit activities within the IPT were not found to be associated with an increased risk of infection. Similarly, other studies [11, 12] found no increased risk of infection among HCWs caring for COVID-19 patients, perhaps because of the higher perception of risk among frontline HCWs [11, 12]. Interestingly, persons having contact with intubated patients while using personal protective equipment (PPE) were found to have a decreased risk of infection, highlighting that appropriate use of PPE and strict measures can decrease the transmission of SARS-CoV-2 [13]. This is consistent with the fact that SARS-CoV-2 is mainly transmitted through the airway and less frequently through the manipulation of blood sample or tracheal nasal swabs. Hence, the use of simple barrier measures as recommended at IPT (wearing masks and gloves) efficiently limits these professional risks. Unfortunately, the overall high level of infection in the general population (including IPT staff) points to the fact that these barrier measures are still neglected in everyday life. A systematic review commissioned by WHO on the epidemiology and risk factors of COVID-19 and other coronaviruses among HCWs around the world [14] showed that estimates of SARS-CoV-2 infection vary significantly across studies. In fact, the incidence of PCR-positive SARS-CoV-2 infection varied from 0.4–49.6%, and the prevalence of SARS-CoV-2 seropositivity ranged from 1.6–31.6% [11]. However, comparing the results of seroepidemiological studies may not be appropriate, since several factors contribute to variability in estimates, such as differences in healthcare settings, the intensity of community transmission where healthcare facilities are located, different study designs, the use of different serology tests with varying sensitivity and specificity, and different time periods of data collection [1, 15].

Of the participants in this study who had been diagnosed with COVID-19 more than three months prior, less than half had anti-N IgG antibodies, while nearly two-thirds had anti-S-RBD IgG antibodies. This is in accordance with the results of previous studies showing that anti-S antibody titers tend to stabilize over time, while anti-N IgG titers decay at a faster rate [16, 17]. Gallais et al. [16] have suggested that serological assays targeting the nucleocapsid protein should not be used preferentially for seroprevalence studies, despite their ability to distinguish between natural infection and post-vaccinal immunity.

The proportion of asymptomatic infections in the present study was 33.3%. This result is similar to the findings of meta-analyses performed by Sah et al. [18] and Ma et al. [19], which found that the pooled percentage of asymptomatic infections was 35.1% and 40.5%, respectively. The high prevalence of asymptomatic COVID-19 cases suggests that screening based only on symptoms will fail to identify a considerable proportion of SARS-CoV-2 infections, increasing the threat of rapid dissemination of the disease [20]. Consequently, scaling-up of testing and periodic screening are recommended as appropriate control measures to tackle transmission of the infection [21].

Our study revealed that the incidence of SARS-CoV-2 infection was lower among young adults. According to a systematic review of the seroprevalence of SARS-CoV-2 among HCWs before the vaccine was introduced [22], many studies [23–27] have reported no significant association between seroprevalence rates and age, while other studies [28–31] showed a significant association between younger age and seropositivity. The susceptibility of younger HCWs was explained by a higher rate of community transmission among this age category and a more active role during patient care to protect older HCWs [22]. The findings of the present study may be explained by the results of studies carried out by Varma et al. [32] and Wong et al. [33]. The former study found that young people may be more vulnerable to stress and anxiety during the COVID-19 pandemic [32], and the latter revealed that higher anxiety was associated with an increased adherence to preventive measures against COVID-19 [33].

This study found higher odds of SARS-CoV-2 infection among laborers compared to scientists and medics. Similarly, Saba Villarroel et al. [34] found that cleaners had a higher SARS-CoV-2 seroprevalence than other occupational groups in healthcare and suggested that this increased risk for low-income occupational groups may be due to their frequent use of public transportation, lower adherence to preventive measures, and living in overcrowded housing. Our result may also be attributed to differences in education levels between the two groups. Indeed, some studies have reported that persons with lower levels of education considered themselves to be at a lower risk of becoming infected with SARS-CoV-2 and were therefore less likely to adhere to COVID-19 safety recommendations [35–40].

We found lower odds of SARS-CoV-2 infection among smokers. Paleiron et al. [41] reported a paradoxical link between smoking and becoming infected with SARS-CoV-2 in the literature, as smoking appears to reduce the susceptibility to infection but increase the risk of developing severe disease. Some hypotheses have been advanced to explain the puzzling protective role of tobacco. Some authors have proposed that nicotine has an influence on the biosynthesis of angiotensin conversion enzyme 2 (ACE2), the receptor for SARS-CoV-2 cell adhesion [42]. Other authors have suggested that the resistance of tobacco smokers to SARS-CoV-2 infection may be immunologically mediated by chronic exposure to the common tobacco-dwelling virus: the tobacco mosaic virus (TMV) [43]. The presence of TMV virions and related RNA may induce the production of interferons and other cytokines, which will be present in case of exposure to SARS-CoV-2 and may limit infection [43]. Nevertheless, the reasons for the lower susceptibility of smokers to SARS-CoV-2 remain unknown and require further investigation [41]. Moreover, smoking clearly should not be recommended as a public health measure [41].

In the present study, we did not find an association between the degree of compliance with preventive measures and the occurrence of SARS-CoV-2 infection. Since nearly two-thirds of participants reported that they adhered to preventive measures, a prevarication bias cannot be ruled out, in particular if a proportion of individuals who had already been infected with SARS-CoV-2 felt uncomfortable admitting negligent behavior because they considered themselves responsible for setting a good example in terms of compliance and for raising awareness among the general population about the importance of these measures. In addition, since only 10.5% of surveyed persons were trained in COVID-19 prevention and control, some participants might claim to apply preventive measures when in reality they are not performing them correctly. Insufficient infection prevention and control (IPC) training, lack of monitoring of IPC routines, and poor hand hygiene have been identified as risk factors for contracting SARS-CoV-2 in healthcare settings [14, 44, 45], while proper use of PPE has proven to be protective in other studies [46, 47].

This study is the first to report the incidence of SARS-CoV-2 infection among HCWs in Tunisia. Unlike several previous studies that were limited to the estimation of the prevalence of anti-SARS-CoV-2 antibodies, we thought that it would be more judicious to determine the cumulative incidence of the infection based on the results of both serology and RT-PCR. This is because the actual extent of SARS-CoV-2 infection can be underestimated if only serology results are considered, since antibodies can become undetectable over time and a proportion of infected persons may not develop antibodies, as revealed in the present study (27.3% of participants with RT-PCR-confirmed infection had negative serology). Specimens collected from the surveyed IPT staff prior to vaccination would be very useful for post-vaccination follow-up surveys. In fact, the post-vaccination humoral response was found to be stronger in previously infected individuals [48].

Our study had some limitations. First, the relatively high rate (23%) of individuals who declined to participate in the survey may have led to an underestimation or overestimation of the cumulative incidence of SARS-CoV-2 infection depending on the frequency of past infection in non-respondents. Second, a prevarication bias may have occurred. Indeed, since the interviewers involved in the study were among the IPT staff, some participants may have felt embarrassed to answer some questions honestly, especially about their past medical history, the consumption of addictive substances, and compliance with preventive measures. Third, the cross-sectional design of our study did not allow the inference of causal relationships between dependent and independents variables.

Our results showed a similar rate of SARS-CoV-2 antibodies in IPT workers and the general population. As such, the Pasteurian community may be considered as a sentinel group that is more accessible to longitudinal monitoring and may deliver valuable information to assess changes in trends of SARS-CoV-2 transmission in the general population.

Authors contribution

Ghassen KHARROUBI: data curation, methodology, investigation, formal analysis, writing – original draft. Ines CHERIF: data curation, methodology, investigation, writing – original draft. Wissem GHAWAR: data curation, investigation. Nawel DHAOUADI: data curation, investigation. Rihab YAZIDI: data curation, investigation. Sana CHAABANE: project management and coordination, data curation. Mohamed Ali SNOUSSI: data curation, investigation. Sadok SALEM: data curation, investigation. Wafa BEN HAMMOUDA: data curation, investigation. Sonia BEN HAMMOUDA: data curation, investigation. Adel GHARBI: data curation, investigation. Nabil BEL HAJ HMIDA: data curation, investigation. Samia ROUROU: data curation, investigation. Koussay DELLAGI: conception of the work, resources, funding acquisition, supervision, validation, visualization, review & editing. Mohamed-Ridha BARBOUCHE: validation, review & editing. Chaouki BENABDESSALEM: supervision, validation, visualization, review & editing. Melika BEN AHMED: supervision, validation, visualization, review & editing. Jihène BETTAIEB: conception of the work, data curation, formal analysis, investigation, methodology, project administration, resources, supervision, validation, visualization, writing – review & editing. All authors read and approved the final manuscript.

Funding

The study was conducted in the frame of the COVID Africa Repair Project, a multipartner research program of the Pasteur Network, grouping the 10 Institutes Pasteur established in Africa. Repair was funded by the French Ministry for Europe and Foreign Affairs (MEAE) grant number: SC/projet_REPAIR N°57/2020.ipt. There was no involvement from the study sponsors in any part of preparing, conducting, writing and submitting the manuscript.

Availability of data and materials

The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request.

Declarations

Conflict of interest

The authors declare that they have no competing interests.

Ethical approval  and consent to participate

The study was conducted according to the guidelines of the Declaration of Helsinki and approved by the Biomedical Ethics Committee of the IPT (registration number: 2021/02/I/LR16IPT). All participants were fully informed about the study and signed a written informed consent to be enrolled. Testing was available at no cost to all participants on a voluntary and uncompensated basis.