Occupational risk factors of contracting COVID-19 among health workers: A systematic review

2.1Search strategy

This study aimed to review, extract, and summarise the risk factors of infectious diseases for HWs. We selected the Web of Science Core Collection for the literature search, and used the “Topic” research for retrieving published reports, reviews, and primary observational studies (case-control, cross-sectional, and cohort). To gather as much additional information as possible, every retrieval item was exported with “Full records and cited references” in “.txt” format from Web of Science. The time span was limited to 2010 to 2021. The keywords used for the initial data collection included “Covid-19”, “SARS-CoV-2”, “infectious disease”, “health care worker”, “nurse”, “infectious factor”, and “risk factor”.

2.2Selection criteria

We used a three-stage selection process to filter the literature. In the first stage, duplicate and irrelevant items were identified and deleted, and the initial data obtained was used for subsequent bibliometric visualisation analysis. In the second stage, the titles and abstracts of the remaining articles were screened to exclude obviously irrelevant studies. In the third stage, the full texts were briefly browsed to further screen the literature. After further screening, the final articles for information extraction were obtained. Inclusion criteria were all reports, reviews, and primary observational studies describing occupational risk factors for infectious diseases among HWs.

2.3Bibliometric analysis

Bibliometrics uses mathematical and statistical methods to quantitatively analyse published articles. Its history can be traced back to the beginning of the 20th century. Its importance in research evaluation is increasing, and it has become a widely used method for analysing academic exchanges and evaluating research results [8]. It compares the citation influence of researchers, research groups and institutions at different times and in various disciplines. Since these two factors are not related to the quality of publications, these academic studies can be quantitatively evaluated [9].

Many software packages have been developed for bibliometric analysis, such as CiteSpace, HistCite, BibExcel, and CitNetExplorer. It is convenient to use CiteSpace to derive a sequence of co-citation networks from a series of equal length time interval slices. These time registered networks are merged and visualised in a panoramic view in such a way that intellectually significant articles can be identified based on their visually salient features [10]. Therefore, in this study, we used CiteSpace (5.7.R3) for our bibliometric analysis.

3.1Analysis of institutions

Institute statistics can provide information about the location and organisation where the research was completed. We used CiteSpace to extract the author’s institution and its location information and counted the frequency of each institution.

Table 1 lists the top 15 contributing institutions. Huazhong University of Science and Technology in China ranks the best in this regard, with a total of 15 occurrences in the 619 studies included. In terms of geographic location, the United States has the strongest overall research strength in this area, with seven institutions, such as Columbia University in the City of New York and Harvard Medical School appearing in Table 1. In addition, the institutions listed in the table include not only world-class universities, but also national medical organizations, such as the Centers for Disease Control and Prevention. From this point of view, the research on infectious diseases and HWs is relatively widespread in terms of geographic location and institutional type.

3.2Analysis of co-citation journals

An analysis of co-citation journals can help us understand the most cited journals and the research fields covered by the literature. Based on CiteSpace, we derived the co-citation information of the source journals of 15,853 references out of the 619 articles collected.

Table 2 lists the top 15 co-citation journals and frequency statistics. It shows that in the co-citation network formed by the 15,853 research studies, these journals have been cited the most times at the same time and have the greatest influence in this research field. Among them, Lancet has been co-cited the most, up to 221 times, and New England Journal of Medicine has been co-cited 174 times, which illustrates the important position of these journals in the medical field. In addition, most of the co-cited journals are journals in the fields of medicine, infection, or diseases in medicine or biology, indicating that the extraction of occupational risk factors for medical personnel is closely related to medicine and biology.

3.3Analysis of co-citation references

In the quantitative research of bibliometrics, co-citation research is one of the most commonly used research methods, especially the References Co-citation Analysis [11, 12]. To further show the risk factors of infectious diseases for HWs, we used CiteSpace to conduct an article co-citation network analysis on the references cited in the search data set. We imported the 15,853 references from the 619 articles into CiteSpace. The nodes of the network were selected under the criteria that the top 10%of the most cited or occurred items were kept in each slice with the maximum number of selected items not more than 100 per slice.

There are a total of 544 nodes and 1040 edges in the citation network. Subsequently, we performed a cluster analysis on the resulting network. CiteSpace uses an internal spectral clustering algorithm to identify clusters, which is based on the eigenvectors of the Laplacian matrix derived from the original network. By integrating network visualisation, spectrum clustering, automatic clustering and text summarisation, the task of analysis and visualisation is simplified [13]. We used the CiteSpace clustering and cluster definition algorithm to obtain 117 clustering results, retaining the 11 main clusters, and obtaining the clustering results and the detailed information of each cluster label.

Figure 1 shows the clustering results of the co-citation network; the distribution of each cluster is relatively scattered. Among them, the research coverage about HWs is the widest, and the research is relatively scattered. The research on hand hygiene is relatively concentrated and has a large time span. Personal protective equipment (PPE) and hospital staff intersect the most with other clusters. To further understand the specific information of each cluster, we exported the detailed information results of the clusters.

Fig. 1

Co-citation network cluster view.

Table 3 lists the specific information of each cluster label, sorted according to the number of members in the cluster from large to small, namely hand hygiene compliance, hospital staff, personal protective equipment, COVID-19 pneumonia, rescue strategy, hand hygiene, SARS-CoV-2 infection, Ebola virus disease, current treatment, and isolation facilities. Considering that the smaller clusters are not representative, to avoid redundancy and overlap of the visual graphics, we filtered the clusters with fewer than 10 members among the 117 clusters. In each cluster, the indicator value “Silhouette”, which measures the clustering result, is close to 1, which means that each cluster has consistency within the other and that the results are convincing. In addition, Table 3 also shows the average publication year of the references contained in each cluster. There are more article clusters concentrated in 2019-2020.

To understand the time course of the research on risk factors of infectious diseases for HWs more intuitively and clearly, we derived a timeline view of the clusters from the calculation results of CiteSpace.

Based on Fig. 1 and Fig. 2, it can be concluded that the concerns about hand hygiene (#0 hand hygiene compliance and #5 hand hygiene) originated very early, judging by its citation in 2004. In addition, it involves the most research and is the largest cluster. Maintaining hand hygiene is considered to be the most important factor in preventing HWs from being infected. The transmission of infectious diseases from patients to HWs is usually because their hands are contaminated by contact with the virus [14].

Fig. 2

Co-citation network timeline view.

Research on PPE (#2 personal protective equipment) also became intensive after 2014. The infection of HWs with SARS-CoV-2 is probably caused, inter alia, by a lack of PPE or its incorrect use [15]. Having sufficient PPE and the correct use of PPE can effectively reduce HWs’ risk of being infected with SARS-CoV-2 [16, 17].

Research on isolation facilities (#10 isolation facilities) indicates that this is also an important risk factor. It began to receive widespread attention after SARS in 2003. Inadequate isolation facilities will increase the risk of HWs contracting infectious diseases. Placement of infectious patients in isolation facilities can effectively prevent HWs from contracting the virus [18].

In addition, the cluster (#1 hospital staff) for HWs also has a high research prevalence. As the main force in the fight against COVID-19, HWs play an important role in the prevention and treatment of infectious diseases. Since HWs are in close contact with infected patients at work, if they cannot be fully protected, it will make them more susceptible to infectious viruses, which can contribute to paralysis of medical systems [3]. In addition, through the clustering results, it was be found that the main types of viruses infecting HWs are viruses such as SARS-CoV-2, Ebola, and Middle East respiratory syndrome coronavirus (MERS-CoV).

On this basis, we also derived the 10 references with the strongest citation bursts through CiteSpace, as shown in Fig. 3. Co-citation references analyses the frequency of articles being cited by other articles at the same time, which helps to determine which articles have a significant impact in the research field and finds the key or core articles in the research field.

Fig. 3

Top 10 references with the strongest citation bursts.

From Fig. 3, we can get the top ten articles cited in the research and the time period when they were frequently cited. This information can reflect the hotspots in this research field to a certain extent. In addition, the high-frequency citations of these top 10 references are also closely related to the formation of clusters in CiteSpace, and each cluster has a corresponding relationship with each strongest citation burst.

3.4Study findings

Our initial keyword search obtained 628 articles that met the search criteria. After 9 articles with duplicate entries were eliminated through the first screening stage, 619 articles were obtained for bibliometric analysis. Then, the title of each article was preliminarily screened, and the articles related to infectious diseases and HWs were selected. A total of 66 articles met the requirements. The abstracts and full text of these articles were briefly browsed, and 24 articles that were not related to the risk factors of infectious diseases among HWs were excluded. After reading the remaining 42 articles, we excluded 16 articles that did not involve the study of occupational risk factors of HWs, and finally identified 26 articles for the extraction of risk factors.

We carefully read the full texts of the 26 articles used for information extraction and extracted relevant information from each article to obtain Table 4. It includes 10 reviews [19–28], one case control study [29], six retrospective cohort studies [3, 14, 30–33], two cross-sectional studies [34, 35], two observational cohort studies [4, 36], two prospective cohort studies [16, 37], two reports [38, 39], and one interview [40]. One article was published in 2013, two articles were published in 2019, and 23 articles were published in 2020. The types of viruses and diseases included in the studies are COVID-19, MERS-CoV, SARS, latent tuberculosis, bloodborne illnesses, and respiratory infections.

The information extracted in Table 4 specifically includes first author, country, publication year, title, study design, sample size, population, disease profile, and risk factor. To ensure the consistency, only one reviewer extracted the risk factors in the articles. The reviewer first carefully read the 26 articles and then summarised the risk factors in each article, classified and organized them, and created related narratives.

According to the extraction results, the occupational risk factors for HWs to contract the virus can be divided into nine categories. They are exposure to infected patients and the virus, lack of PPE, suboptimal hand hygiene, lack of infection control training, work overload, mental stress, cross-infection, unsafe disposal of medical waste, and inadequate disinfection.

The lack of PPE and exposure to infected patients and the virus are the most important risk factors for HWs. A total of 19 risk factors extracted from the research include these two factors. According to the results of the studies, both surgical and N95 masks can effectively prevent HWs from contracting COVID-19, MERS-CoV, SARS, and other respiratory infectious diseases, and the use of N95 can effectively reduce HWs’ risk of being infected with SARS-CoV-2 [25, 41, 42]. In Denmark, HWs lacking PPE will have an increased risk of contracting the virus [4]. Among HWs infected by the virus in China, those who use N95 protective masks are mostly infected mildly and asymptomatically [30]. An Italian study showed, among the 10 HW patients with severe COVID-19, nine were infected due to contact with infected colleagues without PPE [36].

Suboptimal hand hygiene is also one of the main risk factors. Hand hygiene is considered to be an effective way to prevent respiratory infections such as COVID-19, MERS-CoV, and SARS [25, 43, 44]. Maintaining hand hygiene is a well-known key measure to reduce cross-infection in medical facilities. Saudi Arabia physicians have suggested that after contact with contaminated medical equipment, the surrounding environment, and biological tissues, hands should be thoroughly cleaned with alcohol and soap [22].

Lack of training or poor knowledge is also one of the reasons why HWs contract the virus. The survey results of two studies in Africa showed that among HWs, 21%–32%of the respondents indicated that they lacked training or had poor knowledge about the latest measures to prevent infectious diseases [45, 46]. This will increase their risk of contracting infectious diseases. In the early stage of COVID-19 in China, the main reason for the large number of infections among HWs was lack of information about the virus and lack of training on the latest infection control methods [21]. Regular infection control training helps to reduce the risk of infection, and assists HWs to have the latest knowledge about virus control and prevention methods, so as to protect them from virus infection [34, 47].

Work overload and mental stress are also considered risk factors for HWs to contract infectious diseases. Research showed that HWs in China generally work longer hours, with an average working time of more than 54 hours per week. As the daily working hours of HWs increase, their likelihood of contracting COVID-19 will increase [14]. Work overload means that HWs lack the necessary rest, and at the same time exposes them to patients and viruses for a longer time, which increases the risk of infection [19, 39, 48]. In addition, longer working hours and a larger workload lead to mental stress in HWs. They are under greater psychological pressure when caring for infected persons. A study in China analysed the mental health status of 1,257 HWs caring for COVID-19 patients and found that 50.4%, 44.6%, 34.0%, and 71.5%of them had depression, anxiety, insomnia, and distress, respectively [49]. During the COVID-19 pandemic, HWs in Taiwan devote more time to coping with a high work load under conditions of tremendous psychological pressure and with insufficient resources, which will further increase their anxiety[39]. In Taiwan, HWs’ anxiety was further impacted by growing stigmatisation and loss of trust by their own communities during SARS [50].

In addition, the risk factors for HWs to contract infectious diseases also include cross-infection, unsafe disposal of medical waste, and inadequate disinfection. HWs are on the front line of fighting the virus, and they are prone to cross-infection of patients-to-HWs and HWs-to-HWs in their working environment, especially when the virus information is not clear at the initial stage [31, 51, 52]. According to Chinese radiologic technologists, HWs in radiology departments are at greater risk of cross-infection during COVID-19, and how to avoid cross-infection in this environment is particularly important [20]. The treatment of medical waste should be implemented in accordance with “Medical Waste Management and Regulations”. The waste of patients with COVID-19 should be regarded as infectious medical waste, and special sealed packaging should be used to properly dispose of it to prevent cross-infection [20]. A study in Taiwan showed routine disinfection of the daily working environment of HWs and the transitional environment in contact with patients can reduce the possibility of cross-infection for HWs [39].