Background

The World Health Organization (WHO) defines violence as "the intentional use of physical force or power, threatened or actual, against oneself, another person, or against a group or community, that either results in or has a high likelihood of resulting in injury, death, psychological harm, maldevelopment, or deprivation" [1]. Workplace violence (WPV) could be physical or psychological, including verbal violence, bullying/mobbing, racial harassment, and sexual violence [1]. Although the prevalence of sexual violence is lower than that of other types of violence [2,3,4], it should not be overlooked, as it impacts the health and quality of life of the harassed persons (hereafter survivors) negatively [5, 6]. According to the WHO, sexual violence is “any unwanted, unreciprocated, and unwelcome behavior of a sexual nature that is offensive to the person involved, and causes that person to feel threatened, humiliated, or embarrassed” [7]. In a systemic review and meta-analysis, Worke et al. [8] reported a prevalence of workplace sexual violence in all Ethiopian workplaces of 22%. In another review of patient violence against healthcare workers (HCWs) in psychiatric inpatient wards, the rate of sexual WPV was 9.5–37.2% [9]. Other reviews have reported varying rates of 0.3% in Taiwan [3], 12% in Ghana [10], and 73% in Turkey [5]. Another systematic review and meta-analysis of sexual WPV inflicted by patients and visitors reported a rate of 14.2% [2]. These variations may be due to different understandings of the meaning of sexual violence in different cultures and the availability of staff per population, noting that the lower the ratio, the heavier the workload, and the less time available for proper communication with the patients [2]. In Saudi Arabia, a conservative Arabic community, sexual harassment is a very sensitive issue. Reported rates of work violence in specific localities around Saudi Arabia ranged from 3% to 76.5% [4, 11,12,13]. None of these studies focused on sexual WPV and were conducted in certain cities in Saudi Arabia, in hospitals in the same city and same departments, such as the nursing or emergency department. The reluctance of victims to report incidents may be due to the fear of potential repercussions, such as damage to their professional or personal reputation, or the possibility of retaliation from the perpetrator [14]. Sexual attack can result in fear, safety concerns, injury, work leave [15], and diminished work quality [16]. Other effects include psychological disorders, such as anxiety, depression, posttraumatic stress, and/or eating disorders [6]. This low reporting rate can lead to underestimation of the problem and therefore imposes the need for stricter regulations and mechanisms to prevent the attacks and deal with their side effects.

Previous studies have focused on sexual WPV in high-risk environments and specialties in some cities in Saudi Arabia. To our knowledge, none of these studies covered the whole of Saudi Arabia and all specialties. In addition, few studies have focused on the association between WPV and independent risk factors, such as sociodemographic factors, working conditions, and factors from hospital violence reports.

Therefore, this study aimed to determine the prevalence of sexual WPV over a period of 12 months, circumstances related to events, consequences for attackers and survivors, target populations at all healthcare provider facilities in Saudi Arabia, and the most susceptible group of healthcare providers. We also identified the factors associated with WPV in healthcare facilities in Saudi Arabia.

Methods

Data collection

An analytical cross-sectional study was conducted between November 4, 2018, and July 1, 2019, among all healthcare providers registered in the Saudi Commission for Health Specialty (SCFHS) and who had been working for more than 1 year in the health sector (government or private) in Saudi Arabia as of May 2019. A non-probability convenient sampling technique was used; the desired sample size was determined based on a maximum variance assumption of 50% that the healthcare workers would report a positive experience of the types of abuse studied. The desired sample size required to detect the true proportion of individuals who had experienced any type of abuse studied with 95% confidence and a margin of error equal to 5% was deemed to be 384. All eligible participants (i.e., physicians, pharmacists, nurses, midwives, health specialists, healthcare technicians, and ambulance personnel) were invited to participate in the study. A total of 304,002 healthcare providers met the eligibility criteria. Students, interns, employees of the administrative department, healthcare providers not registered in the SCFHS, or providers with less than 1 year of work experience were excluded.

Data were collected using a modified self-administered questionnaire developed by the Joint Program on Workplace Violence in the Health Sectors of the WHO, International Labour Organization, International Council of Nurses, and Public Services International [17]. One of the authors (AH) translated the questionnaire into Arabic for staff who were not fluent in English. The questionnaire was then revised by the other two authors (FS and RS) who are both bilingual. Questions that did not apply to Saudi Arabia were omitted.

A pilot test was conducted for reliability and validity by distributing the questionnaire to five physicians, five dentists, five nurses, and five pharmacists, who were both Arabic and English speakers and had clinical experience in validating the Arabic translation to avoid misunderstandings; these practitioners were excluded from the main study.

The questionnaire included questions related to demographic data of the respondents, workplace characteristics, violent events in the previous 12 months, risk factors for WPV, personal opinions, perceptions, attitudes, experiences, and WPV-related recommendations. The questionnaire had a total of 88 questions divided into five sections: personal and workplace data, physical workplace violence, verbal abuse, bullying/mobbing, and sexual harassment.

The questionnaires were e-mailed to the study population by the researchers. To increase the response rate, the researchers sent reminder emails to the participants after 2 weeks.

Data analysis

Data were entered into SPSS IBM (Version 22). Descriptive statistics (frequency and table) were used to describe the basic features of the data. Continuous variables are expressed as mean and standard deviation (SD), whereas categorical variables are expressed as frequencies and percentages. Multiple response dichotomy analysis was used to describe the items measured with dichotomies (“tick all that apply to you” questions). The Kolmogorov–Smirnov test of normality and histograms were used to assess the statistical normality assumption of metric variables. Levene's test of homogeneity of variance was used to assess the statistical homogeneity of variance assumption. The chi-square test of independence was used to explore the associations between the categorical variables, while an adjusted likelihood ratio-chi-squared test was used when the expected count assumption of the chi-squared test was violated. An independent samples t-test was used to assess the mean differences of continuous variables across the levels of categorically binary measured variables.

A multivariate binary logistic regression analysis was conducted to assess the combined and individual associations between relevant predictors of exposure of the HCWs to recent physical violence at the workplace. Associations between the measured predictor variables and their outcomes are expressed as odds ratios (ORs) with 95% confidence intervals (CIs). A p-value below 0.05 was considered statistically significant.

Ethical approval

This study was conducted according to the guidelines of the Declaration of Helsinki. Approval was obtained from the institutional review board of King Saud University College of Medicine (approval number: E-18–3391) before the study was started. Written informed consent for participation, publication, and confidentiality was obtained from the study participants at the beginning of the survey.

Results

Demographic characteristics

A total of 304,002 HCWs were identified from the SCFHS database; only 7,398 (male, 51.3%; female, 48.7%) responded to the questionnaire. The participants’ mean age was 40 ± 8.62 years, and 60% were of non-Saudi origin. Of the participants, nurses, midwives, and healthcare specialists accounted for 38.1%, physicians for 30.91%, healthcare technicians and ambulance technicians for 25.54%, and pharmacists for 5.43%. Most of the participants were employed full-time (89.86%) in the public/government sector (72.47%). Their work settings were as follows in ascending order: ambulatory, specialized units, general medicine, emergency, intensive care, technical services, management, operating room, general surgery, psychiatric, and support services (Table 1).

Table 1 Descriptive analysis of healthcare workers’ sociodemographic and professional characteristics. N = 7398

Experience of sexual workplace violence

Only 3.9% of HCWs had experienced a sexual violence incident at their workplace in the last 12 months, with most of the sexual harassments coming from patients (29.5%) or a staff member (27.6%). Most of the survivors pretended that it had never happened (43.3%) immediately after the act, while 36.2% asked the offender to stop, and 28.4% had taken no action against the offender (Table 2).

Table 2 Descriptive analysis of healthcare workers’ perceptions and experience of sexual workplace violence

Consequences of sexual violence

As shown in Table 3, participants reported a moderate level of disturbance due to their distressing memories, with a self-rated score of 3.4 out of 5 bothering points. Additionally, the participants reported a relatively high level of hyper-alertness related to their experiences of sexual harassment, with a score of 3.93 out of 5 bothering points. In Table 4, 11.2% of the survivors believed an action was taken to investigate the event further by mainly the managers (86.7%). However, 60% of those whose events were investigated reported that a verbal warning was issued to the offenders. The overall satisfaction with the corrective and investigative actions taken to handle the sexual harassment event were between dissatisfied to slightly satisfied (mean satisfaction = 2.16 out of 5). The primary reasons for not reporting the sexual harassment were fear of the negative consequences, thought of reporting being pointless or useless, and shame.

Table 3 Bothered about sexual violence
Table 4 Consequences of sexual violence

Experience of sexual attacks and their sociodemographic and professional factors

Female HCWs had a higher rate of harassment than male HCWs. The age of the HCWs was significantly associated with exposure to sexual harassment at the workplace in the last 12 months (p < 0.001), as a higher proportion of the survivors were in the 30–39 and 20–29 years age groups than in the > 40 years age group. Non-Saudi HCWs were less sexually harassed in the last 12 months than Saudi HCWs (p = 0.003). In addition, physicians comprised the least proportion of the survivors (p < 0.001), while nurses comprised the greatest proportion. Furthermore, the consultant HCWs had a lower rate of sexual violence in the last year compared to the seniors and juniors (p < 0.001). In addition, the proportion of survivors in the working sector did not correlate significantly with exposure to sexual violence, indicating that the HCWs in different sectors may have a nearly equal rate (Table 5).

Table 5 Bivariate analysis of the association between healthcare workers’ experience of sexual workplace violence and sociodemographic/professional factors

Experience of sexual attacks and their working conditions

HCWs who worked in shifts, especially those working the night shift (18:00 to 07:00), were significantly more exposed to sexual violence at the workplace (p < 0.050) (Table 6). In addition, direct physical contact with the patients was a significant predictor of sexual violence among the HCWs (p = 0.001). Moreover, the sex of the patients with whom the HCWs had been working was not a significant predictor of sexual violence (p = 0.155).

Table 6 Bivariate analysis of the association between healthcare workers’ experience of sexual attack at the workplace with their working condition factors

Experience of sexual attacks and characteristics from hospital violence reporting guidelines

Violence-related worry was more common in survivors (mean score, 3.5/5 points using a Likert scale, SD = 1.21) than in non-survivors (mean score, 2.82; SD = 1.33) (p < 0.001). Furthermore, HCWs working in institutions with policies and guidelines for dealing with work-related violence had a lower rate of WPV than those in institutions with no such policies and guidelines (p = 0.006) (Table 7). Encouragement from work to report WPV of any type was a significant predictor of lower rates of sexual WPV in the last year (p < 0.001); HCWs in violence-intolerant work environments had a significantly lower exposure to WPV than those in workplaces without violence intolerance policies. Furthermore, encouragement from managers and employers was a significant predictor of lower exposure to sexual violence among the HCWs (p < 0.001).

Table 7 Bivariate analysis of the association between healthcare workers’ experience of sexual workplace violence and hospital violence reporting guidelines

Relationship between healthcare workers’ experience (in years) and exposure to sexual violence

Female HCWs had more significant exposure to sexual violence than male HCWs (Fig. 1). HCWs with 6–10 years of experience were the most susceptible group in both sexes. However, a decrease in exposure to sexual violence was observed with an increase in the HCWs’ experience (in years), regardless of sex.

Fig. 1
figure 1

Association between healthcare workers’ experience with risk of sexual workplace violence by sex

Multivariate logistic binary regression analysis results

Female HCWs had a more significant exposure to sexual violence in the past year than male HCWs (p = 0.002). There was no association between age of HCWs and sexual violence exposure (p = 0.227). However, non-Saudi HCWs had a significantly lower (33.3% times less) predicted rate of sexual violence than Saudi HCWs (p = 0.005). In addition, nurses had the greatest likelihood of being exposed to sexual WPV (52.8% times higher, p = 0.002). There was a significant negative association between HCW’s experience and risk of sexual exposure in the last year. HCWs with 6–10 years of experience were 3.696 times more (269.6% times more) exposed to sexual violence compared to those with ≥ 21 years of experience (p < 0.001). People working in the private sector had a greater risk (1.47 times more) of sexual violence than those working in other sectors (p = 0.016). The presence of a dedicated guideline/procedure for reporting/handling WPV was not significantly associated with exposure to sexual violence (p = 0.649), but encouragement by institution leaders and administrators to report all cases of sexual violence was a significant predictor of a reduced risk of WPV (38.4% times less) (p = 0.001). Moreover, HCWs caring for elderly patients during most of their work time were 2.51 times (151% times more) more exposed to sexual WPV than those caring for mainly non-elderly patients (p < 0.001) (Table 8).

Table 8 Multivariate logistic binary regression analysis of the predictors of exposure to recent sexual workplace violence

Discussion

To the best of our knowledge, this is the first study in Saudi Arabia to estimate the prevalence of sexual WPV in the healthcare sector. A low rate of sexual violence was observed in the present study, as only 3.9% of the participants were exposed. This study showed a significant association between exposure to sexual violence and being a female HCW (p = 0.002). The odds of being exposed to sexual violence were 1.5 higher (p = 0.002) among nurses. Sexual violence is a significant but not well-documented problem, as no study measured it in all cities in Saudi Arabia according to the authors’ knowledge, which this study discussed.

Most of the studies conducted nationally have mainly focused on a particular city, hospital, or specialty. Al Anazi et al. [18] reported no sexual violence case, possibly because the study was conducted in a small conservative city with a population quite familiar with each other, which has the potential of a negative social impact. The low reporting rates in the study by El-Gilany et al. [19] was due to sex separation in primary healthcare centers. Alharbi et al. [13] reported that almost 75% of their participants experienced sexual violence; this may be due to the different definition of sexual violence used in their study. In addition, most of their participants were female. In a cross-sectional study [11] conducted in Riyadh city among nurses, low sexual violence rates were observed. However, this study was conducted exclusively in Riyadh, and this finding cannot be extrapolated to the entire Saudi population. Moreover, most of the participants were female nurses (78.6%).

Most countries have a high prevalence of sexual violence [3, 20,21,22]. A cross-sectional study conducted in Iran and involving HCWs reported a sexual violence rate of 4.7% [23], which was consistent with our findings. Like Saudi Arabia, Iran is a conservative community. This could explain the low reporting rate due to the sensitivity of the subject and the lower focus on investigating the prevalence and causes. A systematic review was conducted to estimate the prevalence of sexual violence among native HCWs in high-income countries between 2001 and 2019. The prevalence of sexual violence (both harassment and abuse) among HCWs was 6% [24]. However, those countries are less conservative than second-world countries and have better reporting access and rules to prevent and deal with such events. In contrast, a quantitative review was conducted to estimate the rate of different violence types among nurses worldwide [20]. They had seven-fold higher rates of sexual violence than those in our findings (25%). This might be due to under-reporting of sexual harassment in Saudi Arabia, which means more efforts should be directed toward sexual violence in our region. Another cross-sectional study conducted in Ethiopia reported a higher rate [9]; the authors explained this by the unavailability of a sufficient and well-defined system of identification and control of such incidents, as well as a lack of concern about HCWs’ exposure to sexual violence. We observed that most of the perpetrators were patients/clients and not staff members, which is consistent with other findings [3, 13, 25]. In a study conducted in Macau by Cheung et al. [26], all but four of the survivors were harassed by patients and their relatives; the relatively small number of HCWs in Macau can explain this observation. Contrastingly, Celik et al. [27] found that most perpetrators were staff members. However, they involved only nurses, who answer to physicians, and could have placed them in a vulnerable position due to the power differences. Khoshknab et al. [23] reported that most perpetrators were relatives of patients/clients, followed by patients. Their study was conducted in teaching hospitals in which patients are usually surrounded by students and supervisors, giving the relatives a bigger chance for violence. The physical, psychological, and economical pressures on patients and their family members can account for the several sexual violence cases [28].

Most of the survivors in this study tried to pretend that the incident never happened and some of them told the attacker to stop his/her behavior or took no action. In another study [23], most did nothing or told the person to stop. This implies a very serious outcome—the incident can be repeated. Over three-quarter of the participants said that no investigations were conducted to explore the causes of the incident, which is more than the rates reported by Chen et al. [3] and Khoshknab et al. [23]. Apprehension of possible outcomes and feeling of no ensuing action were reasons for not reporting. A study in India involving women [29] revealed that not reporting may have been due to community standards and beliefs, as responsibility would entirely be shifted to the women’s behavior or attitude or the act would be considered normal. Ignorance about one’s rights may also explain under-reporting, as many are afraid to lose their jobs, especially those in the private sector or with temporary jobs. A study in China [30] reported that when the perpetrator was a co-worker, the survivor was unlikely to report the incident. This could be because the survivor does not want to be stigmatized in the workplace or is afraid of the potential outcomes, especially if the attacker has a higher position. Repeated exposure to sexual violence can make the HCW tolerate the act and consider it normal in their daily work [31]; this can also explain non-reporting. Conversely, nurses in Turkey [27] with a low educational level would rather not report sexual harassment, which may be due to their low knowledge of their rights or because they have a lower working position. Song et al. [32] found lack of knowledge of how and what incidents to report, lesser attention to the healthcare providers compared to the patients, and previous experience of no action taken by the authorities after reporting as the main reasons for not reporting. Most survivors were harassed during their night shifts between 18:00 to 07:00, which contrasts the findings of Khoshknab et al. [23] who reported that most incidents happen during morning shifts. The difference in work hours can explain the discrepancy.

In this study, female workers were more exposed to sexual WPV, which is consistent with some findings [22, 33,34,35] and not consistent with others [30, 36]. El-Gilany et al. [19] and Wang et al. [25] found no sex differences between survivors, while Alharbi et al. [13] observed a non-significant difference. In contrast to female workers, male workers may normalize sexual violence in the workplace because they perceive some situations as more sexually oriented than their female counterparts [37]; this may account for the low reporting rates among male survivors. Torre et al. [38] also found no sex differences. However, most of the participants were young (20–24 years), indicating less experience on how to act in such an event. Consistent with our findings, most of the survivors in the study by Fujita et al. [22] were nurses, explained by the high need for direct physical contact and interaction. Yenealem et al. [34] reported that almost all survivors had little work experience (1–5 years) and that 75% of them had procedures for reporting violence in their workplaces. This study found that juniors and HCWs with 6–10 years of experience are more vulnerable to sexual violence, implying that little experience could translate to lack of skills to manage such incidents [34]. In another study conducted in Addis Ababa, most of the nurses who were survivors were single and young, implying their lack of experience in handling such situations and their reverence for higher healthcare providers in their society [39].

The consequences of sexual violence include psychological stress, shame, depression, sleep disturbances, impaired practice, and unhealthy and uneasy relationships with patients [28, 36, 40]. Moreover, the survivor may refer several patients to other colleagues and ask for unnecessary investigations to get rid of the aggressor, which may subsequently lead to a greater cost [27]. Another study conducted in Iran [41] found that some survivors lost their jobs because of absenteeism following violence-related trauma. Some survivors even quit their jobs or prevented their children from working in their field or having a relationship with someone of the same profession. Moreover, some survivors’ relationship with their spouse was negatively affected, as their spouse would blame them for what happened or starting being more suspicious. This may contribute to under-reporting in future and family divisions. Zeighami et al. [42] conducted a qualitative study and proposed some strategies to prevent sexual violence by interviewing nurses with a prior experience. They found that portraying a strict attitude with the perpetrator such as being inactive or behaving ignorantly of the bad behavior, having a professional relationship and not talking or making jokes on private matters, and wearing an unattractive uniform so not to tempt others would stop him/her from continuing. In addition, having the healthcare provider care for the same sex or having a staff member with more experience in the same shift, providing more protective measures for HCWs on night shifts, and changing the workplace for HCWs with a prior experience, are good preventive measures. Nonetheless, education and training on sexual violence should be provided early in schools, colleges, and workplaces. Further, having a zero-tolerance policy by taking immediate legal actions should be promoted. Longitudinal studies are needed to explore the reasons for sexual violence and implement solutions accordingly. More awareness through educational programs and the media for HCWs, patients, and their relatives is important. In addition, a more encouraging environment to report every violent incident with strict consequences for the attackers should be implemented. More importantly, new regulations (e.g., more staff members, shorter waiting times, and more support, such as prevention programs) are necessary.

Limitations

The main limitation of this study is its use of a retrospective self-report questionnaire, which might cause recall bias. In addition, the subject’s sensitive nature may have prevented some workers from participating, resulting in low reporting rates and reporting bias. In addition, although this study had a sufficiently large number of participants, the results cannot be generalized to the entire population.

The strength of this study is that the participants were all HCWs from government or private institutions in Saudi Arabia, unlike previous studies that focused mainly on the emergency departments and nurses in specific cities.

Conclusion

The prevalence of sexual violence is low; however, it remains a risk faced by HCWs, especially those working night shifts and having direct physical contact with patients. The prevalence was highest among nurses, midwives, and healthcare specialists and lowest among physicians. To explore the causes of sexual violence and to implement solutions accordingly, further studies, especially longitudinal, are needed. Educational programs for HCWs, patients, and their relatives are required. Furthermore, increasing awareness using the media is important. The underreporting of cases may skew the magnitude of the problem; thus, a more encouraging environment to report every violence incident with strict consequences for the perpetrators should be implemented. More importantly, new regulations (e.g., more staff members, shorter waiting time, and more support such as prevention programs) are necessary.