Coronavirus Disease 2019 (COVID-19) is caused by severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), a new beta-coronavirus that shares 88% of sequence identity from SARS-CoV, the virus that caused an important outbreak in 2002-2003.1-2 COVID-19 is spread worldwide, and Brazil was the country with the third highest number of accumulated cases in the world by the time this study was conducted, according to the Brazilian Ministry of Health´s official report.3

SARS-CoV-2 is transmitted mainly through inhalation or contact with infected droplets. Health care workers (HCWs) have been one of the groups most affected by the pandemic,4,5 due to their position on the front-line, especially those who work in intensive care units where contact with infected SARS-CoV-2 patients is more direct and frequent.6 A similar situation was observed in the 2002 SARS outbreak when 21% of infected patients were HCWs.7 By September 2020, a total of 1405 Brazilian HCWs had been infected by SARS-CoV-2 and 315 died due to COVID-19. The most frequently infected HCWs were nurse technicians (109,955; 34.1%), nurses (47,339; 14.7%), doctors (33,032; 10.3%), community health workers (16,546; 5.1%), and health units´ receptionists (14,024; 4.4%).8 In addition, the population of HCWs also involves general service personnel involved in cleaning, transportation, and food preparation, which are essential workers for the maintenance of hospital´s functioning but generally are not included in the statistics on infected HCWs.9

The high exposure to SARS-CoV-2, and the increased stress levels that characterize the work in pandemic situations have been well-described.5,10,11 Many COVID-19 studies focused on HCWs mental health 11-12 but we still have gaps in the knowledge of specific characteristics of HCWS and the main risk factors related to COVID-19 infection in this occupational group. Identifying the risk factors for COVID-19 acquisition in hospital settings is an important step to design preventive strategies and to mitigate the impact of COVID-19 on HCWs.13 We aimed to estimate the incidence of SARS-CoV-2 infection, to identify factors associated, and to describe symptoms reported by healthcare workers at a tertiary hospital in Salvador, Brazil.

The University Hospital has around 2,650 HCWs; 2090 were interviewed, but only 2,083 completed sociodemographic and symptom questionnaires and were included in the study. Fig. 1 summarizes the study procedures. Overall, 71.8% were women, 77.8% were self-identified as non-white, and the mean age was 41 ± 10 years. Regarding the current function, 65% were front-line health care workers (doctor, nurse, multiprofessional, nursing technician), 13.3% ancillary workers, 10.8% administrative, 6.3% of students, and 4.6% worked with diagnostic support. Most participants (72.0%) had a higher education level (graduate or postgraduate), and 153 (7.5 %) declared no religion. A total of 271 (13.0%) and 25 (1.2%) HCWs tested positive for IgG and IgM SARS-CoV-2 antibodies, respectively, and three HCWs had a detectable SARS-CoV-2 RT-PCR test, all of them had also IgG positive antibodies.

Fig. 1.

Health care workers SARS-CoV-2 incidence study, flow diagram.

(0.31MB).

The SARS-CoV-2 incidence was 13.0% (271/2083) in the five-month period. The IPR for SARS-CoV-2 was higher among those presenting lower socioeconomic level (Table 1). In addition, there wee differences in SARS-CoV-2 incidence according to religion, with a higher incidence among Catholics and Protestants than in those who professed no religion. Table 1 depicts the difference in incidence of SARS-CoV-2 infection according to main characteristics of HCWs. There was an inverse association between the socioeconomic level, and SARS-CoV-2 incidence.

The logistic regression model (X2 (8) = 4.612, p > 0.798), showed a good fit for the data, which explained 16.0%. Nagelkerke R2 of the variance in COVID-19 correctly classified 88.0% of cases.19 The regression analysis, completed based on 1,887 observations, was performed to ascertain the association between HCWs functional group, education, race, family income, and religion and COVID-19 incidence.

Ancillary workers were almost 5-fold more likely to have COVID-19 than medical students (OR: 4.96, 95% CI: 1.95–12.63). Lower level, elementary education (OR: 2.91, CI 95: 1.19–7.13), and high level of schooling (OR: 2.89 95% CI: 1.61–5.18), remained independently associated with an increased likelihood of presenting a COVID-19 positive test. In addition, Catholics (OR: 2.16, 95% CI: 1.01–4.62) were 2-fold more likely to have a positive COVID-19 test than those without religion (Table 2).

Table 3 displays the main clinical characteristics reported by the study participants. Ageusia (OR: 7.41 95% CI: 6.15–8.94) and anosmia (OR: 8.51 95% CI: 7.09–10.22) were the most frequent symptoms associated with COVID-19 incidence.

This study describes the main characteristics of 2,083 HCWs at a tertiary hospital in Salvador-Brazil, who were tested for SARS-CoV-2 antibodies. A total of 271 (13.0%) HCWs tested positive for SARS-CoV-2 antibodies, and the sociodemographic markers of social vulnerability were predictive of a higher likelihood of infection by SARS-CoV-2. In addition, Catholic religion, and clinical findings of anosmia, and/or ageusia were strongly associated to SARS-CoV-2 infection.

Socially vulnerable HCWs showed the highest COVID-19 incidence, reflecting similar pattern observed among non-HCW population.20,21 We infer that factors like low income, informal work status, low level of education, and black or mixed race, are markers of social vulnerability and are indirect markers of use of mass public transportation, poor access to basic services and probably inadequate use of personal protection equipment. We also found that having no religion was associated with lower risk of COVID-19 infection. Some religious practices can contribute to the spread of COVID-19 due to extended transmission during communal religious prayers and large attendance at religious gatherings and festivals.22,23 This association was poorly studied, especially during lockdown periods, in a retrospective study people declaring no religion had the lowest risk of COVID-19 related death, before and after lockdowns.23

The detected associations between social vulnerability and a higher risk of COVID-19 demonstrate that the risk of infection by SARS-CoV-2 is not only determined by the exposure. Non-white race, lower educational level and lower family income are strongly associated with a higher risk of infection, in a HCW population. In this study, it clearly indicates that the risk of COVID-19 in a hospital is not a direct consequence of a higher exposure to infected patients but is strongly influenced by the HCW´s educational level, and social position. Probably these characteristics are a proxy of lower adherence to preventive actions, like masks use, social distancing and hands hygiene.24

Social disparities were previously described in a Brazilian population-based study, showing that black and mixed individuals had infection rates 81% and 45% higher than whites, with a higher percentage of deaths occurring in primary-care or isolated emergency-care units and predominantly in public institutions, as a reflex of barriers to healthcare access. A positive gradient was found for all indicators of socioeconomic status and increases in disparities (denoted by less education, more household crowding, lower income, and higher concentration of subnormal areas) were associated with higher mortality rates.25 The same pattern seems to be repeated in other countries with different sociodemographic, cultural and health system characteristics, like the United States and the United Kingdom.25-26 Indirectly, the higher rates of infection in vulnerable populations, contributes to increase on contagion, saturation of public health systems or difficulty of access to private health systems and consequently higher mortality.

When the contemporary Brazilian epidemiological report13 is compared to the dates in which our study was carried out, we observe similar proportions of infected nursing technicians, nurses, and doctors, with nursing technicians being the most affected group. This indicates that regardless of the type of service, whether this hospital is a referral for COVID-19 or not, professionals with a lower degree of education could be at higher risk not only due to the greater exposure, but also due to educational factors and socioeconomic limitations. In addition, it was not possible to determine if the contagion occurred during the intra-hospital service or in the community. The low frequency of HCWs with a detectable RT-PCR was expected, as all interviewed participants were on duty within the hospital and were asymptomatic at the moment of testing.

The initial reports described fever and cough as the most frequent symptoms related to COVID-19.27,28,29 However, following the increasing number of reports on the clinical characteristics of the disease, ageusia and anosmia became common symptoms, and highly suggestive of COVID-19.19 In the current study, anosmia and ageusia were the most frequent symptoms presented by infected HCWs, and their presences were significantly associated with COVID-19 incidence. Usually such symptoms are not accompanied by nasal obstruction or rhinitis symptoms; this probably occurs due to direct damage caused by the virus on the olfactory and gustatory receptors 30 or due to altered mucosal immune response in the upper airways.31Table 3 shows that COVID-19 related symptoms, ageusia and anosmia, were strongly associated with COVID-19 positivity.

Our study has some limitations: the cross-sectional design does not allow to determine causality. However, we investigated all active HCWs of a tertiary hospital, which makes the detected associations consistent for the target population and minimizes selection bias. Given the circumstances and the study design, we were unable to adjust for factors such as community and occupational exposures, including the frequency of direct contact with COVID-19 patients or compliance to PPE (personal protective equipment) use. Even with the use of two tests to determine the serological status, there is a likelihood of false positive and false negative results. Some patients also are not able to develop a detectable production of antibodies against SARS-CoV-2, as we were not able to predict the exact number of cases on immunological window, which could also indicate a lower than actual seroprevalence rate. However, the testing approach we used (two screening tests, ELISA and PCR when necessary) was robust enough to provide a safe estimate of seroprevalence rate.

The findings from this large sample reinforce the need of a better understanding of the factors associated with an increased risk of COVID-19 in the HCW population. In conclusion, we observed that among HCWs at a tertiary hospital, markers of social vulnerability like: low mean family income, lower level of schooling, ancillary work, or being black had a significantly higher likelihood of testing positive for SARS-CoV-2 antibodies, those HCWs characteristics seems to be a more important risk factor for COVID-19 than the occupational exposure to infected patients. This knowledge can provide important insights for development and implementation of preventive strategies aiming at minimizing the risk of infection by SARS-CoV-2 in similar institutions, especially in a continental-wide country, like Brazil.

“This study was financed in part by the Coordenação de Aperfeiçoamento de Pessoal de Nível Superior – Brasil (CAPES) – Finance Code 001”. The funder had no role in study design, data collection, data analysis, data interpretation, or writing of the report. The last author is a CNPq researcher (process number: 311095/2020-8).