Analysis and mapping of global research publications on shift work (2012–2021)

Background The main objective of the study was to identify research trends, collaboration patterns, and the most impactful publications in the field of shift work. Methods Documents published in scientific journals indexed in the Scopus database on shift work were retrieved and analyzed using the title/abstract search methodology. The study period was from January 1st, 2012, to December 31st, 2021. Results Two thousand three hundred twenty-eight documents were retrieved with an h-index of 71 and an average number of 4.5 authors per document. The cumulative number of publications showed a linear growth pattern, while that of citations showed an exponential pattern. The most frequent author keywords, excluding keywords related to shift work, were sleep, fatigue, and nursing. The average annual growth rate was 4.3, and the average doubling time was 3.2. No significant correlation was found between the number of publications and national income among prolific countries. Cross-country research collaboration was weak while the degree of author-author collaboration was relatively high. The Chronobiology International journal was the most prolific, while Harvard University was the most active institution in the field of shift work. Conclusions Given the volume and the negative health impact of shift work, better human resource management is needed to create safer and healthier working schedules.

Furthermore, a study showed that night Shift workers are at high risk of COVID-19 infection [15] which means that shift workers could be defined as a high-risk group for COVID-infections. The occupational adverse health effects of shift work are one side of the problem. The other side is the potential risk of injury, accidents, and errors due to fatigue, cognitive impairment, and the inability of shift workers to concentrate on performing critical tasks [16][17][18].
Shift work, especially the night shift, causes circadian misalignment. Circadian misalignment occurs when the inner physiological activities are at a different time than the actual time of the day. Circadian misalignment impairs normal biological processes like insulin sensitivity, immunity, blood pressure, and cardiac autonomic control, leading to different cardiovascular and metabolic health problems [5]. Randomized clinical trials showed that circadian misalignment can cause an increase in blood pressure and inflammatory markers like C-reactive protein, interleukin-6, and tumor necrotic factor, each of which can be an independent risk factor for cardiovascular diseases. To minimize the health risks of shift work, regulations and preventive measures were implemented by governments and manufacturers [19,20].
In September 2019, the 24th International Symposium on Shiftwork and Working Time ("Shiftwork2019") was held in the USA by a group of 189 shiftwork and working time professionals [21]. The discussions in the symposium focused on (1) the impacts of shiftwork on worklife balance, sleep, performance, health, and safety across a wider range of domains and (2) on research practice and innovations on shift work to improve the health and safety of shift workers. The adverse health effects of shift work and the risk of occupational accidents and poor performance attracted the attention of scholars from various scientific disciplines. Therefore, a large volume of literature has been published on shift work [3]. The large volume of literature on the topic requires the implementation of a bibliometric analysis of published literature to give scholars and health experts a snapshot of what has been published on the topic, specifically, research hotspots and future research agendas. The conventional bibliometric analysis uses statistical and mathematical tools to analyze scientific research volume, trends, key players, and citation patterns [22]. In modern bibliometric analysis, research topics, network collaborative ties, and the timeline of important topics are mapped and visualized. Bibliometric studies became attractive after the emergence of scientific academic databases such as Scopus and Web of Science since these databases provide researchers and academics with detailed information about scientific publications, including authorship, affiliation, keywords, citations, journals, and year of publication.
Analysis and mapping of published scientific literature on shift work will stimulate policymakers to develop new working schedules that cause the least occupational adverse health effects to employees and maximum benefits to employers. Identifying the volume, the quality of scientific publications, and hot topics on shift work increases the understanding of scholars and the public about the occupational health concerns of shift workers. Bibliometric studies also allow for comparisons between different countries, institutions, or authors. No studies were published to shed light on bibliometric indicators and the mapping of literature on shift work. However, bibliometric studies on circadian rhythm, in general, were published [23,24].
The current study aimed to give a holistic bibliometric analysis of the scientific literature on shift work across all occupations and professions to identify (1) volume and the annual number of publications produced globally, (2) countries that have produced the most research, (3) journals and authors publishing the most research, (4) extent of research collaboration, (5) most frequent author keywords and terms used in the scientific literature, and (6) the top cited publications that have the greatest scientific impact in the field.

Method
In the current study, scientific literature on shift work was obtained from Scopus using an extensive list of potential keywords and phrases. Scopus is an academic database, a product of Elsevier, with more than 24 thousand indexed journals in various scientific fields [25]. Scopus provides certain functions that allow for the assessment of research growth and trends. It also allows for the export of data to other programs for the mapping of the literature. Most studies that assessed research growth and patterns, used either Scopus or Web of Science to retrieve the global scientific literature. The fact that Scopus is more inclusive than the Web of Science favors the use of Scopus, which is also available for free to many scholars in low-income countries through the Hinari initiative. A comparative study indicated that about 99.11% of the journals indexed in the Web of Science are also indexed in Scopus [26].

Search strategy
The keywords used in the research strategy to retrieve relevant documents were shown in Table 1. The search strategy was developed based on review articles related to shift work [5,[27][28][29][30][31]. Different combinations of keywords were used and implemented in the title or title/ abstract search. The title/abstract search was implemented using certain restrictions to minimize false-positive results since the title/abstract search might retrieve irrelevant documents. Asterisks and quotation marks were used in writing the keywords to sharpen and widen the search. Different Boolean operators were used in the search query in Scopus. The study period was from January 1st, 2012, to December 31st, 2021. The one-decade period was chosen to allow for better bibliometric analysis and to increase the accuracy of the results since old data might not be relevant and not available in databases [26]. The research strategy included documents published in peer-reviewed scientific journals. Therefore, books and book chapters were not included. Of the retrieved documents, letters, notes, editorials, and errata were excluded.

Validation of the search strategy
Two colleagues in the field of biomedical sciences volunteered to check the validity of the search strategy. The check for validity consisted of two approaches. In the first approach, the colleagues were asked to confirm the absence of false-positive articles by reviewing 50 articles randomly selected from an Endnote file sent to the reviewers. The research strategy was enhanced by the feedback from the volunteers. In the second approach, the volunteering colleagues were asked to compare the number of publications of the top active authors with the actual number of articles for each scholar by investigating his or her Scopus profile. The results obtained from the two methods were compared by correlation testing to determine significance and the correlation coefficient.
This approach was used to confirm the absence of falsenegative results. The approach was adopted from previously published bibliometric studies [32].

Data export and data analysis
The documents obtained from the research strategy were exported to the Microsoft Excel program. The exported information included: (1) general characteristics of the retrieved documents, including the type of documents, encountered languages, and type of access. (2) subject areas of the journals publishing the retrieved articles. (3) most frequent author keywords presented in a network visualization map. In the network visualization map of author keyword co-occurrences, items are presented as nodes, and the larger the node size, the higher the frequency of the item in the retrieved documents. The distance between items represents relatedness. (4) annual growth of publications. Data on the annual number of publications were used to calculate the annual growth rate (AGR), average annual growth rate (AAGR), compound annual growth rate (CAGR), and doubling time (DT). The annual growth rate (AGR), defined as the percentage change in the number of publications for one year, was calculated based on the following equation: Table 1 Keywords used in the research strategy on shift work using the Scopus database for the period from 2012 -to 2021 Step strategy Keywords and limitations 1 Title search (title ("night shift* work*" or "night-shift* work*" or ("night job*" and shift*) or "late night work*" or "nightshift work*" or "rotating shift" or "late evening shift*" or "late evening work*" or ("night work*" and shift*) or ("evening work*" and shift) or "rotating shift*" or rotating-shift* or "alternating work*" or alternating-work* or ("non day work*" and shift*) or ("non-day-work*" and shift*) or "shift work*")) or ((title (night and work* and shift*) or title (night and job and shift*) or title (shift* and work*) or title (shift* and *hour* and work*) or title (*shift*) or title (*work* and sleep* and hour*)) and title-abs ("night shift*" or "shift work*" or "rotating shift*" or "shiftwork*" or "work shift" or "alternating shift")) 5524 2 Title/Abstract search with restrictions (TITLE-ABS ("night shift*" OR "shift work*" OR "rotating shift*" OR "shiftwork*" OR "shift work" OR "work shift" OR (alternating AND shift) OR "rotating shift")) AND TITLE (schedul* OR "occupational safety" OR "circadian synchr*" OR "working time arrangement*" OR "work" OR "working" OR shift OR night OR (sleep AND deprivation) OR "circadian disruption" OR "circadian misalignment" OR (circadian AND dysregulation) OR (circadian AND disruption) OR (circadian AND *alignment) OR (circadian AND disturbance*)) 7368 2 The average annual growth rate (AAGR) is the average change in the value of a measurement over the study period. The compound annual growth rate (CAGR) provides a constant rate of return over the study period [33]. Its formula is as follows: where "n" is the number of years.
The growth analysis was also presented as the "relative growth rate" (RGR), which was defined as the increase in the number of publications per unit of time.
The RGR was calculated based on the following equation [34]: where log e W 1 : log of the initial number of articles; log e W 2 : log of the final number of articles after a specific period of interval; and T2-T1: the unit difference between the initial and final times.
The RGR can be presented in a different format called "doubling time" (DT), defined as the time required for the number of publications to double in number in one year and was calculated based on the following equation [34]: (5) Top 10 active countries are presented in a table. (6) Cross-country research collaboration of countries with a minimum contribution of 10 publications presented as a network visualization map using the free online VOSviewer program [35]. The node size represents the relative number of collaborating countries while the thickness of connecting lines represents the strength of collaboration (i.e., joint research publications). (7) The top 10 active authors and author-author degree of collaboration. Prolific authors and authorauthor degree of collaboration. Author details were exported from Scopus to Microsoft Excel, where authorship analysis was carried out. Analysis in Microsoft Excel included the number of singleauthored, two-authored, three-authored, and multiauthored (joint) articles. Analysis of overall collaboration in the field was calculated using the following equation [36]: where N m = number of multi-authored papers and N s = number of single-authored papers.
The retrieved articles were published in 22 different languages, mainly English (n = 2100, 90.2%). The remaining documents were written in non-English but have bilingual abstracts (English and non-English). The presence of bilingual abstracts is a condition imposed by Scopus on all non-English journals that are indexed in Scopus database. The most common non-English languages encountered in the retrieved documents were German (n = 45, 1.9%), followed by French (n = 34, 1.5%), Russian (n = 34, 1.5%), Spanish (n = 26, 1.1%), The dominance of English language is partially because the majority of Scopus-indexed journals are English journals with fewer percentage of non-English journals.

Subject areas of the retrieved articles
Shift work is of concern to several scientific fields. The Scopus database has categorized the retrieved articles into 26 subject areas. Table 2 presents the top 10 subject areas on shift work research. The "medicine" subject area has the highest number of publications (n = 1559, 70.0%), followed by nursing (n = 315, 13.5%), biochemistry/ molecular biology/genetics (n = 293, and social sciences (n = 245, 10.5%). Because certain journals may be categorized in more than one field, there was an overlap in the subject areas of "shift work" research and the total number was greater than the retrieved number of articles.

Keyword co-occurrence analysis
Keyword co-occurrence analysis was carried out to identify research hotspots and future research directions of the academic field. In this study, a keyword co-occurrence visualization map was created by VOSviewer program. The top 50 author keywords were mapped. The most frequent author keywords in the dataset, excluding keywords related to shift-and night work, were sleep, nursing, fatigue, circadian rhythm, and circadian disruption ( Fig. 1)

Growth of publications and citations
The publication trend in the field is demonstrated in Table 3. The mean number of publications during the study period was approximately 233 documents per year. The highest number of published documents was 323 (13.9%) achieved in 2020. Growth analysis indicated that the highest AGR during the study period was 23.9%. The AGR had three negative values, indicating certain fluctuations during the study period. The RGR declined from 0.6 in 2013 to 0.2 in 2021. The DT increased from 1.2 in 2012 to 3.5 in 2021. Figure 2 presents the cumulative number of publications and the cumulative number

Spatial distribution of publications
The retrieved articles were published by authors from 87 different countries/territories. The country with the most publications was the US (n = 504, 21.6%), followed by Australia (n = 178, 7.6%), and the UK (n = 140, 6.0%). Table 4 lists the top 10 active countries (core countries) in the field. The list included four English-speaking   Table 5).

Authorship analysis, author-author degree of collaboration, and prolific authors
In total, 10,516 author names appeared on the retrieved documents, giving an average of 4.5 author names per document. There were 290 (12.5%) single-authored publications. The number of multi-authored publications was mainly those with five or more authors (n = 731, 31.4%). Table 6 shows the authorship pattern and author-author  degree of collaboration. The degree of collaboration among authors was 87.5%. The degree of author-author collaboration showed an increasing pattern, with approximately 84% during the early times of the study period and reaching approximately 93% toward the end of the study period. Table 7 shows the top 10 active authors. Bjorvatn, B. (Norway) was the most prolific author with 36 (1.5%) publications, followed by Pallesen, S. (Norway) with 33 (1.4%) publications. The list of active authors included four researchers from Norway, two from Finland, and two from the US. Figure 4 is a network visualization of collaborative ties among authors who published at least 10 documents in the dataset (n = 51). Fifty-one authors met the criteria. However, six of them did not fit into any research group and therefore were not shown on the map. The map shows 45 authors distributed into seven clusters, five of them included five or more researchers contributing to the development of the field. The largest collaboration network represents a research group composed of 13 scholars affiliated with institutions in the US and Canada. Table 8 lists the top 10 institutions/organizations on "shift work" research. ‫أ‬Harvard University (n = 97, 4.2%) was the most prolific institution in the field followed by Brigham and Women's Hospital (U.S) (n = 86, 3.7%), and the Finnish Institute Of Occupational Health (Työterveyslaitos) (n = 43, 1.8%) and the University of Bergen (n = 43, 1.8%).

Top 10 active journals in publishing documents on shift work
The retrieved articles were disseminated through 991 scientific journals.

Citation analysis
The retrieved documents received 32,301 citations with an H-index of 71. Table 10 shows the top 10 cited documents in the field of shift work [3,[38][39][40][41][42][43][44][45]. Six articles were review articles. The top-cited articles investigated the impact of shift work on health, including diabetes mellitus and cancer. One of the top-cited documents investigated the impact of shift work on nursing job satisfaction and burnout.

Discussion
The increasing prevalence of shift work across most professions throughout the world and the adverse health and social effects of shift work led to the accumulation of a large volume of scientific literature on the topic. In the current study, the author analyzed and assessed the scientific literature on shift work to give an overview of the evolution, growth, and key players in the field.
Most of the retrieved documents were published in journals categorized in the subject area of "medicine". However, there was a good share of publications in the "nursing" subject area. Most nurses worldwide work night shifts to cover patient care across 24 hours, leading to health and social problems for nurses and risks of poor performance and errors [5,16,46,47]. Nurses opt to work the night shifts for various financial and family reasons, but they are mostly unaware of the chronic negative health implications of working night shifts [48][49][50]. Shift work negatively affects the quality of life of shift workers, especially women, due to insufficient time for marital and child care [51,52]. Shift work negatively affects the individual's mental health, including psychological distress, anxiety, and depression [31,53,54].
The current study showed a positive value for AAGR. This increase could be due to (1) an increase in the prevalence of shift work across different societies and  professions, (2) the natural increase in the number of scholars and global research productivity in general, (3) the emergence of many specialized journals in the field of occupational health and sleep medicine, and (4) the appearance of studies linking shift work to serious health consequences such as cancer and cardiometabolic disorders. Keyword analysis and the top-cited articles indicated that various negative health consequences, sleep deprivation, fatigue, nursing, and circadian disruption were major hot topics in the field. Shift work is associated with short-and long-term health problems [55]. For example, insomnia, fatigue, and sleep disturbances are related to the acute effects of shift work while potential cardiometabolic and cancer health problems are related to the chronic effects of shift work [3,9,10,53,56,57]. A systematic review suggested that the increased health risks in shift workers may be due to the desynchronization of the circadian rhythm, that alters the normal regulation of physiological functions in the body [58].
Journals in the core list were mainly in the field of occupational health and sleep medicine. Journals in the field of nursing or mental health were under-represented. Shift workers constitute a good percentage in developed countries. For example, in Europe, 21% of the workforce is engaged in some type of shift work [59]. This relatively large proportion of the population requires special attention, and further research is needed to raise awareness and develop safer working schedules for workers across professions with 24-hour working systems. The presence of several occupational/work health journals and leading sleep journals emphasizes that shift work, especially, the night shift is unhealthy and may be dangerous.
The countries in the active list were mainly highincome and industrialized countries, including the US, Australia, and the UK. This is not surprising since highincome countries have well-established infrastructure and resources for scientific research as evident in previously published bibliometric studies [60,61]. Australia ranked second in this field despite its poor contribution to the field as measured one decade ago [62]. One potential reason for this finding is the increased health and medical research investment in Australia which led to a noticeable increase in PubMed publications from Australia [63]. A bibliometric analysis of the top 100 cited articles on sleep medicine found that two-thirds of the articles were published by authors in the US followed by Canada, the UK, Germany, and Australia [64]. Another bibliometric study of the top 100 cited chronotype research papers showed that Australia was in the sixth rank at the global level [24]. These studies showed that Australian researchers have active role in research related to sleep, biological clock, and circadian rhythm, which might explain their rank on the shift work research.
Certain developing countries such as Iran, China, and Brazil were also present in the core list. The emergence of these developing countries in the top active list is due to rapid industrialization and the need for shift working schedules in these countries. Furthermore, the increasing number of Scopus-indexed journals published by institutions and research centers in developing countries increased the visibility of research output from these countries. There has been a noticeable increase in the number of journals in the fields of public and occupational health, as well as sleep medicine, from various countries and world regions in the past two decades. Currently there are more than 100 health related journals published from Iran and indexed in Scopus.
In 1982, a study by Mahathevani, R. indicated that in developing countries, such as Malaysia and Singapore, there was a total lack of documentation on the different shift systems and that scientific literature on the effects and consequences of shift work was absent [65]. The author of the study as well as consequent studies indicated that shift work was heavily influenced by gender, social factors, type of industry, cultural values, and religious norms of the countries [66]. The limited contribution of certain countries and regions is mostly due to limited research resources and capacities rather than the absence of the problem. In a survey study of 1400 shift workers across 20 countries and all industries, 90% of the shift workers feel that they contribute to the success of their organization, 50% consider themselves to be essential workers, and 69% feel concerned about their job security [67].
The core countries included Canada, which ranked 8th on the list. However, publications from Canada were highly impactful. Cohort studies on the cost and health injuries due to shift work in Canada drew the attention of researchers and policy workers across the world. A cohort study conducted in Canada on 30,000 Canadians between the ages of 16 to 65, showed that from 1996 to 2006, the number of worker injuries decreased by 27.9% while the rate of injury among night shift workers remained stable [68]. From 1996 to 2006, the Canadian Labour force increased by 21.7%, with almost half of this growth in non-regular shift work. In 2006, there were 2.7 million lost time injury compensation claims awarded in Canada with approximately 107,000 claims among men and 200,000 claims from women, secondary to the high risk of injury associated with shift work. A study on the estimate of shift work indicated that in 2011, there were 1.8 million (12% of the working population) Canadians exposed to the night shiftwork [69].
Author-author collaboration and cross-country collaboration maps indicated the presence of noticeable collaboration networks at the author level but not at the cross-country level. Most of the collaboration networks were between researchers in countries that share cultural, geographical, or linguistic factors. This explains the existence of countries such as Iran and Poland at the periphery of the map with a limited number of links with other countries. The opposite applies to countries such as the US, the UK, Canada, and Australia. The inadequate international research collaboration in the field of shift work between researchers in developed and developing countries is a barrier to the development of this field. Shift work is a global phenomenon, and international collaboration among different countries is needed to strengthen research in this field and increase the visibility of shift worker-related work and health problems.

Strength and limitations
Several bibliometric studies were published to analyze and map the literature on sleep, circadian disruption, melatonin, and chronotype research [23,24,64,70]. However, there was one scientometric article published on shift work in 2011 in the German language [62]. The main strengths of the current study were (1) the comprehensiveness of the research strategy without language or subject restrictions and (2) the comprehensiveness of the analysis of the retrieved data, such that the current study can be considered a future reference for researchers in the field of shift work. However, the current study has a few limitations. The use of Scopus is a point of strength and a point of weakness. Scopus is a large academic database of all fields. However, the use of a single database makes the results less than perfect. The research strategy, despite validation, is not without error. Therefore, potential false-positive and negative results remain a possibility that needs to be considered by readers. The use of title search is expected to minimize any research errors.

Conclusions and recommendations
In the current study, scientific literature on shift-and night work was retrieved, analyzed, and interpreted using the bibliometric approach to explore the content, countries, institutions, authors, and journals involved in the field. The analysis revealed steep annual growth in publications over the last decade, with US authors and institutions dominating. Content analysis indicated the impact of shift work on sleep quality, risk of cardiometabolic diseases, and cancer. Of the professions most investigated by the retrieved literature was nursing. The current study is the first to conduct a bibliometric analysis of the literature on shift work, and it identified several hot topics that merit further investigation and research. Policymakers can build on the current study to implement human resources management that promotes better working shift schedules and a safer work environment.