Work Schedule and Client Characteristics Associated With Workplace Violence Experience Among Nurses and Midwives in Sub-Saharan Africa Mazen El Ghaziri, PhD, MPH, BSN, RN Shijun Zhu, PhD Jane Lipscomb, PhD, RN, FAAN Barbara A. Smith, PhD, RN, FAAN Violence against health care workers perpetrated by clients and/or their friends and family (Type II) is a growing problem that can severely impact health care delivery. We examined the prevalence of Type II workplace violence among nurses and midwives in sub-Saharan Africa and its association with work status, schedule, and client characteristics. Nurses and midwives (n 5 712) completed an anonymous survey while attending nursing meetings. Generalized estimating equation models, accounting for clustering within residing countries, were employed. Participants who were exposed to risky client characteristics (aOR 5 1.39–1.78, p , .001), and those who worked more than 40 hours a week were more likely to have experienced Type II workplace violence (aOR 5 1.72–2.15, p , .05). Findings will inform policy and organization level interventions needed to minimize nurses’ and midwives’ exposure to Type II workplace violence by identifying risky clients and addressing long work hours. (Journal of the Association of Nurses in AIDS Care, 25, S79-S89) Copyright Ó 2014 Association of Nurses in AIDS Care Key words: Africa, client characteristics, midwives, nurses, work characteristics, work status, workplace violence
Violence against health care workers is a growing problem worldwide and can severely impact health care delivery. Workplace violence is defined as ‘‘violent
acts, including physical assaults and threats of assault, directed toward persons at work or on duty’’ (Centers for Disease Control and Prevention/National Institute for Occupational Safety and Health, 1996, p.1). Violence can be classified into four types: Type I (criminal intent), Type II (customer/client), Type III (worker on worker), and Type IV (personal relationship) (University of Iowa Injury Prevention Research Center [IPRC], 2001). In sub-Saharan Africa, Type II workplace violence among nurses and midwives, the focus of this paper, is defined as that which occurs when the person who commits the act of workplace violence is either the recipient or object of service (current or former client, patient, customer, criminal suspect, or prisoner, etc.) provided in the workplace by a health care worker or victim (IPRC, 2001). Workplace violence is a common concern for both developed and developing countries. The economic burden of workplace violence associated with fatal Mazen El Ghaziri, PhD, MPH, BSN, RN, is a PostDoctoral Fellow, Center for the Promotion of Health in the New England Workplace (CPH-NEW), University of Connecticut Health Center, Farmington, Connecticut, USA. Shijun Zhu, PhD, is a Biostatistician, School of Nursing–Office of Research, University of Maryland, Baltimore, Maryland, USA. Jane Lipscomb, PhD, RN, FAAN, is Professor, School of Nursing, University of Maryland, Baltimore, Maryland, USA. Barbara A. Smith, PhD, RN, FAAN, is Professor and Associate Dean for Research at Michigan State University, College of Nursing, East Lansing, Michigan, USA.
JOURNAL OF THE ASSOCIATION OF NURSES IN AIDS CARE, Vol. 25, No. 1S, January/February 2014, S79-S89 http://dx.doi.org/10.1016/j.jana.2013.07.002 Copyright Ó 2014 Association of Nurses in AIDS Care
S80 JANAC Vol. 25, No. 1S, January/February 2014
or nonfatal assaults affects organizations as well as victims, costing billions of dollars in lost wages, medical costs, support costs, and lawsuits (Schmidtke, 2011). Moreover, workplace violence has tremendous psychological and physiological impact on the victims, their families, co-workers, and the organization as a whole, and elicits concerns about the organization’s commitment to safety on the job (Beech & Leather, 2006). It is worth noting that in many places in sub-Saharan Africa, 50% or more of hospitalized patients are infected with HIV (Madhava, Burgess, & Druker, 2002), which adds another layer of risk within the work environment for nurses and midwives. Despite the magnitude of the problem of workplace violence, research exploring the causes and evaluating interventions to reduce workplace violence is still limited. By contrast, numerous researchers representing various disciplines have described various conceptualizations of the problem, with a focus on both personal and workplace factors leading to violent behaviors and injury. Barling (1996) conceptualized workplace violence and included personal factors such as type A behavior, alcohol consumption, and history of aggression as predictors of workplace violence. Smith-Pittman and McKoy (1999) stated that every violent state involves ‘‘at least four elements: a perpetrator, causative factors, environment conducive to violence, and targets’’ (p. 6). Occupational health researchers, on the other hand, conceptualized workplace violence using a work organization framework (Centers for Disease Control and Prevention/ National Institute for Occupational Safety and Health, 2002; McPhaul & Lipscomb, 2004; McPhaul, London, & Lipscomb, 2013) that is inclusive of a comprehensive health and safety program approach (U.S. Department of Labor, & Occupational Safety and Health Administration, 2004). The literature provides limited data about the magnitude of Type II workplace violence in subSaharan Africa as well as other middle- and lowincome, developing countries. The purpose of this paper was to assess the prevalence of Type II workplace violence in a sample of nurses and midwives primarily from three sub-Saharan African countries: Nigeria, Tanzania, and Kenya.
Prevalence of Workplace Violence Globally and in Sub-Saharan Africa In addition to having 25% of the world’s disease burden, sub-Saharan countries have a severe shortage of health care workers (Dovlo, 2007). Nurses constitute 45% to 60% of the health care workforce in subSaharan Africa, with high nurse-to-physician ratios, and one of the lowest nurse-to-patient ratios in the world (Dovlo, 2007). The nursing shortage in subSaharan countries can be attributed to a number of factors, including: economics, politics, education, the HIV pandemic, work organization, and bureaucracy, all of which potentially affect the work environment of nurses and midwives by causing them to work long hours in critically understaffed facilities (Dovlo, 2007). Consequences of these shortages include an increasingly stressful work environment, placing the nurses at risk for violence and stressrelated diseases, ultimately compromising the quality of care. In 2002, a report on workplace violence in Australia, Brazil, Bulgaria, Lebanon, Portugal, South Africa, and Thailand, sponsored by the International Labor Office, International Council of Nurses, World Health Organization, and Public Services International, acknowledged the extent and severity of the problem of workplace violence in the health care setting (Di Martino, 2002). The report indicated that a majority of health care workers in each country had experienced at least one incident of physical or psychological violence in the previous year (67.2% in Australia, 46.7% in Brazil, 75.8% in Bulgaria, 61% in South Africa, 54% in Thailand, and, in Portugal, 60% in a health center and 37% in a hospital; Di Martino, 2002). In the United States, a high percentage of workplace incidents and deaths have been attributed to violence within health care settings (Centers for Disease Control and Prevention/National Institute for Occupational Safety and Health, 2002; Lipscomb, Silverstein, Slavin, Cody, & Jenkins, 2002; Peek-Asa, Cubbin, & Hubbell, 2002; PeekAsa et al., 2009; U.S. Department of Labor, Bureau of Labor Statistics [BLS], 2006). In each year from 1993 to 1999, 1.7 million incidents of violence occurred in the workplace (Duhart, 2001). The BLS
El Ghaziri et al. / Workplace Violence Among Nurses and Midwives in Sub-Saharan Africa
(2006) reported that 45% of all nonfatal assaults leading to lost workdays against workers took place in the health care sector, where patients were the perpetrators in 50% of the assaults. In 2005, the rate of nonfatal assaults among all health care workers combined was 8.8/10,000 cases (BLS, 2006). Mental health professionals had an annual incidence rate of 68.0/1,000 workers, and nurses had an annual incidence rate of 22.0/1,000 workers (BLS, 2006). Data from the United Kingdom have indicated that one third of nurses had been either abused or assaulted at work (Harulow, 2000); in Ireland, 60% of emergency department nurses and attendants had been physically assaulted at least once (Rose, 1997). Canadian researchers reported that 80% of nurses had experienced some form of violence at work (Cruikshank, 1995), and, in Australia, nursing as a profession was found to have the highest rate of exposure to violence, exceeding that perpetrated against prison and police officers (Perrone, 1999). Despite the scarcity of literature on workplace violence in Africa, injury patterns in Africa seem to be similar to those in developed countries (Forjouh, Zwi, & Mock, 1998), where reports from some sub-Saharan countries revealed that injury accounted for a considerable proportion of all deaths and morbidities (Forjouh et al., 1998). In a study investigating the prevalence and pattern of occupational health hazards in the obstetrics and gynecology unit of a hospital in Nigeria (n 5 78), the following were reported: work-related stress (83.3%), needle-stick injuries (75.6%), bloodstains on skin (73.1%), sleep disturbance (42.3%), skin reactions (37.2%), and hepatitis (8.9%). It is worth noting that almost one fourth of the staff (24.3%) had been assaulted by patients (Orji, Fasubaa, Onwudiegwu, Dare, & Ogunniyi, 2002). Another Nigerian study examined the prevalence of workplace violence (James, Isa, & Oud, 2011) and found that verbal aggression was the most prevalent type of violence experienced by nurses working in psychiatric facilities (67.2% occasionally/sometimes and 32.8% often/frequently). Among the 73 nursing staff, male nurses were more likely to have experienced physical violence and aggressive ‘‘spitting’’ behaviors, while nurses with more than 10 years of experience were more likely to have experienced verbal and humiliating aggressive behaviors.
S81
Among Southern Nigerian dental health professionals (n 5 175), the prevalence of violence was 31.9%, and violence was associated with long waiting times (27.3%), cancellation of appointments (13.6%), outcomes of treatment (11.4%), alcohol intoxication (9.1%), psychiatric mental health issues (6.8%), patient bills (4.5%), and others (27.3%). Types of violence included nonphysical violence in the form of loud shouting (50.0%), threats (22.7%), sexual harassment (6.8%), and swearing (2.3%), with bullying and hitting constituting the remaining 18.2%. The perpetrators of the violence were mainly patients (54.5%) and their relatives/friends (18.2%). The consequence of the violence resulted in reports of fear (18.2%), impaired job performance (15.9%), psychological problems (13.6%), and lost time (9.1%) (Azodo, Ezeja, & Ehikhamenor, 2011). Our paper describes the prevalence of Type II workplace violence in a sample of nurses and midwives practicing in sub-Saharan African countries. In addition, the paper identifies risk factors that, if addressed through prevention activities, might reduce the risk of violence toward health care workers in sub-Saharan Africa.
Methods Study Sample Data are from a sample of 756 health care workers who completed a questionnaire during the African Midwives Research Network biennial conference in Nairobi, Kenya (2007); Dar Es Salaam, Tanzania (2009); and nursing and midwifery meetings held throughout Nigeria from 2008 through 2010. The questionnaire included items about work schedule; work organization; patient/client factors such as alcohol use, mental status, and presence of a weapon; worker exposure to blood and other bodily fluids; knowledge and use of universal precautions; HIVrelated stigma; and demographics of nurses and midwives. This paper focuses on work organization and schedule, client and patient characteristics, and demographics of the nurses and midwives. Slight variations were noted across countries in some survey questions, which necessitated recoding and variable computation to ensure uniformity across the dataset.
S82 JANAC Vol. 25, No. 1S, January/February 2014
A total of 756 health care workers in attendance at the three meetings responded to the questionnaire. Out of the total sample, 722 were nurses and nurse midwives who provided direct clinical care to patients in sub-Saharan African countries; 10 resided outside of sub-Saharan countries (i.e., Australia, Sweden, the United Kingdom, and the United States). Therefore, 712 nurses and midwives meeting our eligibly criteria were included in the analyses. The University of Maryland–Baltimore institutional review board approved the project as exempt, as did the ethics committees in Kenya (Kenyan Medical Research Institute), Nigeria (National Health Research Ethics Committee), and Tanzania (National Institute for Medical Research). Measures Type II workplace violence was assessed by examining nurses’ and midwives’ previous 12-month exposures to patients, family members, or patients’ friends who had: (a) yelled or sworn at them; (b) threatened to assault them; or (c) physically hurt them (outcome variables), and five client characteristics as depicted in Table 1. The five items described as client characteristics assessed whether the exposure to violence involved a patient, family member, or patient’s friend who had mental illness, was intoxicated, had guns or visible weapons, or had a history of assault. Table 1 also documents whether the respondent often dealt with difficult patients, family members, or patients’ friends. The five client-characteristics items were analyzed individually and then combined into an index for Table 1.
subsequent analysis. A total index score was derived by summing the number of questions answered as yes. Scores ranged between 0 and 5, with a higher score representing more risk. The mean 6 standard deviation index score was 2.94 6 1.6. The five-item index demonstrated good internal consistency, with a Cronbach’s alpha of 0.74. Data Analysis Descriptive analyses were performed to describe demographic variables, work status, work schedule, and clients’ characteristics. Bivariate analyses (c2 tests) were performed to assess the relationship between outcome and independent variables. Because of the clustering effects of nurses within residing country (e.g., the interclass correlation 5 0.0928 for outcome of physically hurt/assaulted, data not shown), generalized estimating equation (GEE) logistic regression models, accounting for clustering effects, were performed to assess the associations between each aspect of workplace violence and independent variables (work status, work schedule, and client characteristics) separately, after controlling for the covariates. Covariates that were significant or marginally significant on bivariate analysis were included in the multivariate GEE logistic regression models. Multicollinearity of the predictors was checked using variance inflation factor. The significance level was set at 0.05, and two-tailed 95% confidence intervals (CI) were obtained for all analyses. All analyses were performed using SPSS 19.0 and STATA 12 software.
Descriptive Statistics of Workplace Violence Experience and Client Characteristics (N 5 712)a Parameter
Client risk characteristics I often deal with difficult patients/family/friends In the past 12 months I have worked with patients, family members, patients’ friends who had history of assault In the past 12 months I have worked with patients, family members, patients’ friends who had mental Illness In the past 12 months I have worked with patients, family members, patients’ friends who were intoxicated In the past 12 months I have worked with patients, family members, patients’ friends who had guns or weapons visible Workplace violence experience In the past 12 months I have worked with patients, family members, patients’ friends who yelled or swore at me In the past 12 months I have worked with patients, family members, patients’ friends who threatened to assault me In the past 12 months I have worked with patients, family members, patients’ friends who physically hurt/assaulted me a. Numbers of respondents may not sum to total due to missing data.
n
%
501 444 414 352 163
78.0 69.9 65.4 55.6 25.9
263 224 125
40.8 34.7 19.9
El Ghaziri et al. / Workplace Violence Among Nurses and Midwives in Sub-Saharan Africa
Results The sample included 712 nurses and midwives who were present at one of the three meetings and who completed the questionnaire. Participants resided in more than 12 sub-Saharan African countries, with the majority from Kenya (n 5 254), Nigeria (n 5 177), and Tanzania (n 5 169). A minority number of participants came from Uganda (n 5 31), Zimbabwe (n 5 6), Zambia (n 5 4), Ethiopia (n 5 2), Namibia (n 5 2), Central African Republic (n 5 1), Eritrea (n 5 1), Malawi (n 5 1), Togo (n 5 1), and other sub-Saharan African countries (n 5 63, unspecified). Six hundred nurses and midwives (84.3%) in the sample resided in Kenya (35.7%), Nigeria (24.9%), or Tanzania (23.7%), the three sites where the meetings occurred, which allowed us to account for the clustering within residing country. Table 1 presents data from the entire sample (n 5 712) and shows that 40.8% reported having been called names or been yelled or sworn at by patients, family members, or patients’ friends during the previous 12 months, while 34.7% reported being threatened to be physically hurt or assaulted (hit, slapped, punched, kicked, strangled); 19.9% were actually physically hurt or assaulted. Seventy-eight percent often dealt with difficult patients, family members, and patients’ friends, while 69.9% had worked with clients who had a history of assault. The participants also reported that they had worked with clients who had mental illness (65.4%), were intoxicated (55.6%), or had visible guns or weapons (25.9%). Table 2 presents the sample’s distribution and the bivariate associations between the demographics and work organization variables, and each of the three outcome variables. The mean age of the participants was 42 6 9 years. The majority of the participants were female (79.7%), held certificates and diplomas rather than the bachelor’s degree (75.3%), and worked in urban areas (73.1%). Approximately 37% reported direct patient care for 50% to 80% of their time, while 41.2% spent more than 80% of their time in direct patient care. Almost 36% reported working more than 40 hours per week. Thirty percent worked ‘‘off-hours’’ more than once a week, and more than half were required to visit patients in their homes.
S83
The number of working hours per week was associated with all three measures of workplace violence. That is, a higher percentage of nurses or midwives who worked 40 hours or more per week reported that they had been yelled or sworn at (51.2%), threatened with assault (44.9%), or physically hurt/assaulted (27.6%) by patients, patients’ family members, or patients’ friends in the previous 12 months, compared with those who worked 40 hours or less per week (all, p , .001; Table 2). Similarly, age was significantly associated with participants having worked in the previous 12 months with patients, family members, or patients’ friends who had yelled or sworn at them and threatened to assault them. For example, more than half of nurses or midwives (51.1%) in the age range of 22–35 years reported that they had been yelled or sworn at by patients, patients’ family members, or patients’ friends in the previous 12 months, compared with 36.1% and 40.3% of participants in the 36–45 and 46 or older age groups, respectively (p , .001). About 40% of the nurses or midwives ages 22–35 years reported having been threatened with assault, while 37% of participants between 36 and 45 years of age and 28.9% of those ages 46 years or older reported having been threatened with assault in the previous 12 months (p , .001). Furthermore, percentage of time spent in direct patient care was associated with being actually physically hurt or assaulted (p , .001). Of nurses and midwives with greater than 80% of time spent in direct patient care, approximately one third (28%) reported having been actually physically hurt or assaulted by patients, family members, or patients’ friends, compared with 17% of those who worked 50%–80% of their time in direct patient care, and 10.9% of those who worked less than 50% of their time in direct patient care. In addition, country of residence was associated with participants’ having worked in the previous 12 months with patients, family members, or patients’ friends who had yelled or sworn at them, threatened to assault them, and actually physically hurt or assaulted them (all, p , .05). Kenya had the highest prevalences of participants who had been yelled or sworn at (47.2%) and threatened with physical hurt or assault (42.5%), while Nigeria had the
Descriptive Statistics of the Sample in Relation to Workplace Violence (N 5 712)a
Parameter Gender Male Female Age group 22–35 years 36–45 years $46 years Education Certificate or other Baccalaureate degree Master’s or doctoral Work location Urban Rural Both Country Kenya Nigeria Tanzania Other sub-Saharan African country Percentage of time spent in direct patient care ,50 50–80 .80 Hours per week 0–40 .40 Required to visit patients at their homes Yes No Frequency of working off-hours Zero or once a week More than once a week
N
%
Yelled or Swore at Me Threatened to Assault Me Yes, n (%) No, n (%) p Value Yes, n (%) No, n (%) p Value .275
144 20.3 564 79.7
46 (36.8) 210 (42.2)
79 (63.2) 288 (57.8)
.604 48 (36.9) 170 (34.5)
82 (63.1) 323 (65.5)
.009b 201 29 225 32.5 266 38.4
94 (51.1) 73 (36.1) 89 (40.3)
90 (48.9) 129 (63.9) 132 (59.7)
387 75.3 70 13.3 60 11.4
136 (38.5) 22 (36.7) 13 (31)
217 (61.5) 38 (63.3) 29 (69)
516 73.1 172 24.4 18 2.5
181 (39.7) 65 (41.9) 5 (45.5)
275 (60.3) 90 (58.1) 6 (54.5)
75 (40.3) 76 (37.4) 63 (28.9)
111 (59.7) 127 (62.6) 155 (71.1)
118 (33.1) 25 (40.3) 11 (28.2)
238 (66.9) 37 (59.7) 28 (71.8)
145 (32.2) 62 (39) 6 (50)
306 (67.8) 97 (61) 6 (50)
122 (52.8) 111 (84.8) 85 (57) 52 (57.1)
144 22.2 237 36.6 267 41.2
46 (37.7) 101 (47.2) 97 (39.3)
76 (62.3) 113 (52.8) 150 (60.7)
416 64.3 231 35.7
140 (36.9) 107 (51.2)
239 (63.1) 102 (48.8)
366 54.9 301 45.1
148 (44.7) 103 (37.6)
183 (55.3) 171 (62.4)
457 70 196 30
163 (39.4) 85 (47.8)
251 (60.6) 93 (52.2)
139 (77.7) 154 (78.6) 179 (82.9)
80 (23.1) 14 (23.7) 3 (7.7)
266 (76.9) 45 (76.3) 36 (92.3)
87 (19.6) 28 (18.5) 2 (15.4)
356 (80.4) 123 (81.5) 11 (84.6)
.375
.081
.898
,.001 b
.020b 99 (42.5) 50 (31.4) 42 (28) 33 (36.3)
134 (57.5) 109 (68.6) 108 (72) 58 (63.7)
38 (31.1) 78 (36.3) 93 (38.1)
84 (68.9) 137 (63.7) 151 (61.9)
111 (29.4) 93 (44.9)
267 (70.6) 114 (55.1)
127 (38) 85 (31.4)
207 (62) 186 (68.6)
139 (33.8) 71 (39.2)
272 (66.2) 110 (60.8)
.132
20 (8.8) 51 (32.1) 32 (22) 18 (21.2)
207 (91.2) 108 (67.9) 114 (78) 67 (78.8)
13 (10.9) 36 (17) 66 (28)
106 (89.1) 176 (83) 170 (72)
56 (15) 55 (27.6)
317 (85) 144 (72.4)
71 (21.8) 44 (16.7)
255 (78.2) 219 (83.3)
76 (18.9) 40 (22.9)
326 (81.1) 135 (77.9)
,.001b
.421
,.001 b
.001b
.077
,.001b
.088
.058
a. Numbers (n) may not sum to total due to missing data. b. p , .05.
40 (22.3) 42 (21.4) 37 (17.1)
.15
.015b 109 (47.2) 50 (38.2) 64 (43) 39 (42.9)
96 (75.6) 390 (81.3)
.408
.843
35.67 24.86 23.74 15.73
.156 31 (24.4) 90 (18.7)
.041b
.624
254 177 169 112
Physically Hurt/Assaulted Me Yes, n (%) No, n (%) p Value
.124
.205
.276
S84 JANAC Vol. 25, No. 1S, January/February 2014
Table 2.
El Ghaziri et al. / Workplace Violence Among Nurses and Midwives in Sub-Saharan Africa
highest prevalence of being actually physically hurt or assaulted (32.1%). Odds ratios based on the GEE logistic regression models are presented in Table 3. The client risk characteristics index was significantly associated with each of the three outcome variables of interest after adjusting for covariates. A unit increase in the client risk characteristics index was associated with a greater likelihood of experiencing violence (adjusted odds ratio [aOR] 5 1.78, 95% CI [1.52 to 2.10], p , .001 for being yelled or sworn at; aOR 5 1.59, 95% CI [1.33 to 1.90], p , .001 for being threatened with assault; aOR 5 1.39, 95% CI [1.13 to 1.70], p 5 .002 for being actually physically hurt or assaulted). Compared with nurses who worked less than 40 hours a week, nurses who worked more than 40 hours a week were more likely to have experienced violence (aOR 5 2.1, 95% CI [1.29 to 3.4], p 5 .003 for being yelled or sworn at; aOR 5 2.15, 95% CI [1.34 to 3.48], p 5 .001 for being threatened; aOR 5 1.72, 95% CI [1.02 to Table 3.
S85
2.84], p 5 .042 for being actually physically hurt or assaulted). No multicollinearity was found of the predictor variables (variance inflation factor ,10). Other variables, including age, being required to visit patients’ homes, percentage of time spent in direct patient care, and frequency of working off-hours, were not significantly related to any aspect of having experienced violence in adjusted GEE logistic regression models.
Discussion Exposure to Type II workplace violence among sub-Saharan African nurses and midwives included being yelled or sworn at, threatened with assault, and actually being physically assaulted. Age was associated with exposure to being yelled or sworn at and with threats of assault. The average number of hours worked per week was significantly associated with all of the outcomes, and percentage of
Odds Ratios Estimates of Workplace Violence in Relation to Independent Variables Based on GEE Logistic Regression (N 5 712)a Yelled or Swore at Me Parameter
Age group 22–35 years (Reference) 36–45 years $46 years Required to visit patients at their homes No (Reference) Yes Percentage of time spent in direct patient care ,50 (Reference) 50–80 .80 Hours per week 0–40 (Reference) .40 Frequency of working off-hours Zero or once a week (Reference) More than once a week Client risk characteristics index
Threatened to Assault Me
Physically Hurt/Assaulted Me
aOR
95% CI
p Value
aOR
95% CI
p Value
aOR
95% CI
p Value
0.78 0.95
0.45–1.35 0.55–1.66
.382 .868
1.7 1.16
0.97–2.99 0.64–2.10
.063 .625
1.21 0.98
.66–2.20 0.51–1.85
.537 .939
1.08
0.68–1.72
.735
1.06
0.66–1.71
.784
1.28
0.76–2.17
.352
1.02 0.69
0.56–1.86 0.39–1.21
.947 .198
0.79 1.22
0.42–1.51 0.65–2.32
.489 .528
1.23 1.86
0.58–2.59 0.87–3.97
.594 .109
2.1
1.29–3.4
.003b
2.15
1.35–3.48
.001b
1.72
1.02–2.84
.042b
0.96 1.78
0.58–1.57 1.52–2.1
.861 ,.001b
1.12 1.59
0.67–1.89 1.33–1.9
.655 ,.001b
0.96 1.39
0.57–1.76 1.13–1.70
.999 .002b
Note. aOR 5 adjusted odds ratio (adjusted for all the other variables listed in the table); CI 5 confidence interval; GEE 5 generalized estimating equation. a. For country, 12 countries and 1 unspecified country were treated as clusters due to intraclass correlation within countries. b. p , .05.
S86 JANAC Vol. 25, No. 1S, January/February 2014
time spent in direct care was significantly associated only with having been actually physically assaulted. When the GEE model was employed, the average number of work hours per week and the risk index for client characteristics were significantly associated with the three workplace violence outcomes. We found that the rate of exposure to being yelled or sworn at in this population (40.8%) was within the range of verbal abuse (27%–67%) experienced by health care respondents in the World Health Organization country report (Di Martino, 2002), with 49.5% exposure to verbal abuse noted among the respondents in South Africa (Di Martino, 2002; Steinman, 2003). Nigerian nurses in psychiatric facilities reported a similar frequency of verbal aggression (James et al., 2011). By comparison, our findings were slightly less than the prevalence of loud shouting experienced by the dental health professionals (50%) in southern Nigeria (Azodo et al., 2011). The rate of exposure to physical violence (19.9%) was higher than the range experienced by midwives and nurses in South Africa (12%–14.1%; Di Martino, 2002; Steinman, 2003) but comparable to the bullying and hitting (18.2%) experienced by the dental health professionals in Southern Nigeria (Azodo et al., 2011). By contrast, 24.3% of nurses in an obstetrics and gynecology department in Nigeria had been assaulted by patients (Orji et al., 2002). Work organization and occupational risks associated with workplace violence in the literature include the number of hours worked, time pressures, and shift work (Campbell et al., 2011; Whittington, Shuttleworth, & Hill, 1996). In our sample, we found that nurses and midwives who worked more than 40 h/week were at higher risk for Type II workplace violence. Whereas the number of hours associated with co-worker violence among European nurses was 35 or more hours per week, where this association was not noted for exposure to Type II workplace violence (Camerino, Estryn-Behar, Conway, Van Der Heijden, & Hasselhorn, 2008). Perpetrator/client characteristics associated with violence have been shown to include male gender; a diagnosis of borderline personality disorder, schizophrenia, or organic mood disorder; substance abuse issues; or having a history of violence (Little, 1999). In our sample, the client characteristics associ-
ated with violence included patients, family, and/or friends who were difficult or intoxicated, had a mental illness, possessed visible weapons or guns, and had histories of assault. Despite the prevailing nursing shortage in sub-Saharan Africa, nurses and midwives reporting more than 80% of time spent in direct patient care were not at increased risk of Type II workplace violence as was illustrated in other studies (Whittington et al., 1996). The majority of our sample was female (79.7%), and gender was not associated with workplace violence. The association between age and exposure to Type II workplace violence was not evident among the sub-Saharan African nurses and midwives, which is inconsistent with literature that has shown that being younger and less experienced (Nolan, Soares, Dallender, Thomsen, & Arnetz, 2001) put nurses at higher risk and frequency of being exposed to Type II workplace violence (Camerino et al., 2008). A nurse’s level of education was not associated with any of the Type II workplace violence outcomes, which is inconsistent with other literature that reflected that nurses having an associate’s degree compared to other degrees increased the risk for exposure to workplace violence (Nachreiner et al., 2007). This variation may be attributed to limited variability due to the fact that most of the nurses and midwives involved in direct patient care were holding certificates or diplomas rather than bachelor’s degrees (75.3%), which limited variability in our study. Limitations Study limitations include the following. The crosssectional design of the study limited our ability to infer causality. Self-reported data may include information bias. Convenience samples limit the generalizability of results. Another design-related limitation was the slight variations noted across countries in terms of several survey questions, which necessitated recoding, and variable computation to ensure uniformity across the dataset. The strengths of the study reside in the fact that it was one of the first to examine questions related to workplace violence in this population. The study involved a multinational sample (12 specified and other sub-Saharan countries) with a high-risk
El Ghaziri et al. / Workplace Violence Among Nurses and Midwives in Sub-Saharan Africa
population of nurses and midwives. Moreover, it used the GEE as a method for determining the factors associated with exposure to workplace violence. The alarming nature of this global problem has prompted organizations such as the World Health Organization, International Council of Nurses, Occupational Safety and Health Administration, and the International Labor Office to develop prevention guidelines and position statements. Despite the complexity and severity of the problem, several initiatives have demonstrated the feasibility and effectiveness of guidelines for prevention, reporting, and training in developed countries such as the United States (Lipscomb et al., 2002; Lipscomb et al., 2006; McPhaul & Lipscomb, 2004). It is crucial to develop interventions that are informed by direct care staff experience and that are culturally appropriate to the setting and population. The use of specially trained security staff for highrisk situations, training in techniques for defusing aggression, the use of metal detectors at entrances, and specialized security police protections when violent patients or their families are on-premises are recommended, but they must be customized to meet the social and economic conditions in the respective countries and regions. Other interventions, including postincident assault services (medical care, mandatory reporting, immediate discharge or removal of perpetrator, access to counseling, and being able to leave work following an incident; May & Grubbs, 2002; McPhaul & Lipscomb 2004), should be explored while ensuring management commitment and employee involvement throughout the entire process (Lipscomb et al., 2006; McPhaul & Lipscomb, 2004).
Conclusions We can infer that long work hours (40 hours or more per week) and certain client characteristics are associated with experiencing Type II workplace violence. Further research is needed to assess additional work-related stressors that might be associated with Type II workplace violence. These findings support the need to pursue policy and organization level initiatives to adjust work hours to minimize the risk for Type II workplace violence. Our finding
S87
of a significant association between client characteristics and workplace violence indicates the need for a thorough assessment of such characteristics upon admission and the development of strategies to increase security or other interventions when encountering high-risk patient visits in order to maintain staff safety.
Key Considerations Violence against health care workers perpetrated by clients and/or their friends and families (Type II) is a growing problem that can severely impact health care delivery. Policy and organizational level interventions are needed to minimize nurses’ and midwives’ exposures to Type II workplace violence by identifying risky clients and addressing long work hours. Policy and organizational level initiatives are needed to adjust work hours to minimize the risk for experiencing Type II workplace violence. A thorough assessment of identified client risk characteristics should be completed upon admission. Strategies should be developed to increase security when encountering high-risk patient visits in order to maintain staff safety.
Disclosures The authors report no real or perceived vested interests that relate to this article that could be construed as a conflict of interest.
References Azodo, C. C., Ezeja, E. B., & Ehikhamenor, E. E. (2011). Occupational violence against dental professionals in southern Nigeria. African Health Sciences, 11(3), 486-492. Barling, J. (1996). The prediction, experience, and consequences of workplace violence. In G. VandenBos, & E. Bulatao (Eds.), Violence on the job: Identifying risks and developing
S88 JANAC Vol. 25, No. 1S, January/February 2014 solutions (pp. 29-49). Washington, DC: American Psychological Association. Beech, B., & Leather, P. (2006). Workplace violence in the health care sector: A review of staff training and integration of training evaluation models. Aggression and Violent Behavior, 11, 27-43. http://dx.doi.org/10.1016/j.avb.2005. 05.004 Camerino, D., Estryn-Behar, M., Conway, P. M., Van Der Heijden, B. I. J. M., & Hasselhorn, H.-M. (2008). Workrelated factors and violence among nursing staff in the European NEXT study: A longitudinal cohort study. International Journal of Nursing Studies, 45(1), 35-50. Campbell, J., Messing, J. T., Kub, J., Agnew, J., Fitzgerald, S., Fowler, B., . Bolyard, R. (2011). Workplace violence: Prevalence and risk factors in the Safe at Work Study. Journal of Occupational & Environmental Medicine, 53(1), 82-89. http://dx.doi.org/10.1097/JOM.0b013e3182028d55 Centers for Disease Control and Prevention/Department of Health and Human Services, National Institute for Occupational Safety and Health (NOISH). (1996). Violence in the workplace—Current Intelligence Bulletin 57. NOISH Publications and Products. Retrieved from http://www.cdc.gov/ niosh/docs/96-100/ Centers for Disease Control and Prevention/Department of Health and Human Services, National Institute for Occupational Safety and Health (NOISH). (2002). Violence: Occupational hazards in hospitals. Publication No. 2002-101. NOISH Publications and Products. Retrieved from http:// www.cdc.gov/niosh/docs/2002-101/ Cruikshank, C. (1995). Nurse abuse: A study of workplace violence in Nova Scotia. Nurse to Nurse, 6(2), 12. Di Martino, V. (2002). Workplace violence in the health sector: Country case studies (Brazil, Bulgaria, Lebanon, Portugal, South Africa, Thailand and an additional Australian study): Synthesis Report. ILO, ICN, WHO and PSI Joint Program on Workplace Violence. Retrieved from http://www.who. int/violence_injury_prevention/injury/en/WVsynthesisreport. pdf Dovlo, D. (2007). Migration of nurses from Sub-Saharan Africa: A review of issues and challenges. Health Services Research, 42(3), 1373-1388. Duhart, D. T. (2001). Violence in the workplace, 1993-99. U.S. Department of Justice, Office of Justice Programs, Bureau of Justice Statistics, National Crime Victimization Survey. Retrieved from http://bjs.gov/content/pub/pdf/vw99.pdf Forjouh, S. N., Zwi, A. B., & Mock, C. M. (1998). Injury control in Africa: Getting governments to do more. Tropical Medicine and International Health, 3(5), 349-356. http://dx.doi. org/10.1046/j.1365-3156.1998.00240.x Harulow, S. (2000). Ending the silence on violence. Australian Nursing Journal, 7(10), 26-29. James, B. O., Isa, E. W., & Oud, N. (2011). Patient aggression in psychiatric services: The experience of a sample of nurses at two psychiatric facilities in Nigeria. African Journal of Psychiatry (Johannesburg), 14(2), 130-133. http://dx.doi.org/ 10.4314/ajpsy.v14i2.4
Lipscomb, J., McPhaul, K., Rosen, J., Brown, J. G., Choi, M., Soeken, K., . Porter, P. (2006). Violence prevention in the mental health setting: The New York State experience. Canadian Journal of Nursing Research, 38(4), 96-117. Lipscomb, J., Silverstein, B., Slavin, T. J., Cody, E., & Jenkins, L. (2002). Perspectives on legal strategies to prevent workplace violence. Journal of Law, Medicine and Ethics, 30(Suppl. 3), 166-172. Little, L. (1999). Risk factors for assaults on nursing staff: Childhood abuse and education level. Journal of Nursing Administration, 29(12), 22-29. Madhava, V., Burgess, C., & Druker, E. (2002). Epidemiology of chronic hepatitis C virus infection in Sub Saharan Africa. The Lancet Infectious Diseases, 2(5), 293-302. May, D., & Grubbs, L. (2002). The extent, nature, and precipitating factors of nurse assault among three groups of registered nurses in a regional medical center. Journal of Emergency Nursing, 28(1), 11-17. http://dx.doi.org/10.1067/ men.2002.121835 McPhaul, K. M., & Lipscomb, J. A. (2004). Workplace violence in health care: Recognized but not regulated. Online Journal of Issues in Nursing. Retrieved from www.nursingworld.org/ MainMenuCategories/ANAMarketplace/ANAPeriodicals/ OJIN/TableofContents/Volume92004/No3Sept04/Violence inHealthCare.aspx McPhaul, K. M., London, M., & Lipscomb, J. A. (2013). A framework for translating workplace violence intervention research into evidence based programs. Online Journal of Issues in Nursing, 18(1), 4. Nachreiner, N. M., Hansen, H. E., Okano, A., Gerberich, S. G., Ryan, A. D., McGovern, M. P., . Watt, G. D. (2007). Difference in work-related violence by nurse license type. Journal of Professional Nursing, 23(5), 290-300. http://dx.doi.org/10. 1016/j.profnurs.2007.01.015 Nolan, P., Soares, J., Dallender, J., Thomsen, S., & Arnetz, B. (2001). A comparative study of the experiences of violence of English and Swedish mental health nurses. International Journal of Nursing Studies, 38(4), 419-426. Orji, E. O., Fasubaa, O. B., Onwudiegwu, U., Dare, F. O., & Ogunniyi, S. O. (2002). Occupational health hazards among health care workers in an obstetrics and gynecology unit of a Nigerian teaching hospital. Journal of Obstetrics and Gynecology, 22(1), 75-78. Peek-Asa, C., Casteel, C., Allareddy, V., Nocera, M., Goldmacher, S., O’Hagen, E., . Harrison, R. (2009). Workplace violence prevention programs in psychiatric units and facilities. Archives of Psychiatric Nursing, 23(2), 166-176. http://dx.doi.org/10.1016/j.apnu.2008.05.008 Peek-Asa, C., Cubbin, L., & Hubbell, K. (2002). Violent events and security programs in California emergency departments before and after the 1993 hospital security act. Journal of Emergency Nursing, 28(5), 420-426. http://dx.doi.org/10. 1067/men.2002.127567 Perrone, S. (1999). Violence in the workplace. Australian Institute of Criminology, Research and Public Policy Series, No. 22. Retrieved from http://www.aic.gov.au/documents/4/
El Ghaziri et al. / Workplace Violence Among Nurses and Midwives in Sub-Saharan Africa 1/0/%7B410FD288-AD82-4E4E-A6FE-C7C30D283528% 7Dfull_report.pdf. Rose, M. (1997). A survey of violence toward nursing staff in one large Irish accident and emergency department. Journal of Emergency Nursing, 23(3), 214-219. Schmidtke, R. (2011). Workplace violence: Identification and prevention. Journal of Global Health Care Systems. Retrieved from http://jghcs.info/index.php/j/article/view/22/ pdf Smith-Pittman, M. H., & McKoy, Y. D. (1999). Workplace violence in healthcare environments. Nursing Forum, 34(3), 5-13. http://dx.doi.org/10.1111/j.1744-6198.1999.tb00988.x Steinman, S. (2003). Workplace violence in the health sector— Country case study: South Africa. Retrieved from http://www. who.int/violence_injury_prevention/violence/interpersonal/en/ WVcountrystudysouthafrica.pdf University of Iowa Injury Prevention Research Center. (2001). Workplace violence—A report to the nation. University of Iowa Injury Prevention Research Center Web Site. Retrieved from http://www.public-health.uiowa.edu/iprc/resources/work place-violence-report.pdf U.S. Department of Labor, Bureau of Labor Statistics. (2006). Survey of workplace violence prevention, 2005. Retrieved from http://www.bls.gov/iif/oshwc/osnr0026.pdf U.S. Department of Labor, Occupational Safety and Health Administration. (2004). Guidelines for preventing workplace violence for healthcare & social service workers. Retrieved from http://www.osha.gov/Publications/OSHA3148/osha3148. html Whittington, R., Shuttleworth, S., & Hill, L. (1996). Violence to staff in a general hospital setting. Journal of Advanced Nursing, 24(2), 326-333. http://dx.doi.org/10.1046/j.13652648.1996.18114.x
S89
Violence against health care workers is a growing problem worldwide and can severely impact health care delivery. Workplace violence is defined as ‘‘violent
acts, including physical assaults and threats of assault, directed toward persons at work or on duty’’ (Centers for Disease Control and Prevention/National Institute for Occupational Safety and Health, 1996, p.1). Violence can be classified into four types: Type I (criminal intent), Type II (customer/client), Type III (worker on worker), and Type IV (personal relationship) (University of Iowa Injury Prevention Research Center [IPRC], 2001). In sub-Saharan Africa, Type II workplace violence among nurses and midwives, the focus of this paper, is defined as that which occurs when the person who commits the act of workplace violence is either the recipient or object of service (current or former client, patient, customer, criminal suspect, or prisoner, etc.) provided in the workplace by a health care worker or victim (IPRC, 2001). Workplace violence is a common concern for both developed and developing countries. The economic burden of workplace violence associated with fatal Mazen El Ghaziri, PhD, MPH, BSN, RN, is a PostDoctoral Fellow, Center for the Promotion of Health in the New England Workplace (CPH-NEW), University of Connecticut Health Center, Farmington, Connecticut, USA. Shijun Zhu, PhD, is a Biostatistician, School of Nursing–Office of Research, University of Maryland, Baltimore, Maryland, USA. Jane Lipscomb, PhD, RN, FAAN, is Professor, School of Nursing, University of Maryland, Baltimore, Maryland, USA. Barbara A. Smith, PhD, RN, FAAN, is Professor and Associate Dean for Research at Michigan State University, College of Nursing, East Lansing, Michigan, USA.
JOURNAL OF THE ASSOCIATION OF NURSES IN AIDS CARE, Vol. 25, No. 1S, January/February 2014, S79-S89 http://dx.doi.org/10.1016/j.jana.2013.07.002 Copyright Ó 2014 Association of Nurses in AIDS Care
S80 JANAC Vol. 25, No. 1S, January/February 2014
or nonfatal assaults affects organizations as well as victims, costing billions of dollars in lost wages, medical costs, support costs, and lawsuits (Schmidtke, 2011). Moreover, workplace violence has tremendous psychological and physiological impact on the victims, their families, co-workers, and the organization as a whole, and elicits concerns about the organization’s commitment to safety on the job (Beech & Leather, 2006). It is worth noting that in many places in sub-Saharan Africa, 50% or more of hospitalized patients are infected with HIV (Madhava, Burgess, & Druker, 2002), which adds another layer of risk within the work environment for nurses and midwives. Despite the magnitude of the problem of workplace violence, research exploring the causes and evaluating interventions to reduce workplace violence is still limited. By contrast, numerous researchers representing various disciplines have described various conceptualizations of the problem, with a focus on both personal and workplace factors leading to violent behaviors and injury. Barling (1996) conceptualized workplace violence and included personal factors such as type A behavior, alcohol consumption, and history of aggression as predictors of workplace violence. Smith-Pittman and McKoy (1999) stated that every violent state involves ‘‘at least four elements: a perpetrator, causative factors, environment conducive to violence, and targets’’ (p. 6). Occupational health researchers, on the other hand, conceptualized workplace violence using a work organization framework (Centers for Disease Control and Prevention/ National Institute for Occupational Safety and Health, 2002; McPhaul & Lipscomb, 2004; McPhaul, London, & Lipscomb, 2013) that is inclusive of a comprehensive health and safety program approach (U.S. Department of Labor, & Occupational Safety and Health Administration, 2004). The literature provides limited data about the magnitude of Type II workplace violence in subSaharan Africa as well as other middle- and lowincome, developing countries. The purpose of this paper was to assess the prevalence of Type II workplace violence in a sample of nurses and midwives primarily from three sub-Saharan African countries: Nigeria, Tanzania, and Kenya.
Prevalence of Workplace Violence Globally and in Sub-Saharan Africa In addition to having 25% of the world’s disease burden, sub-Saharan countries have a severe shortage of health care workers (Dovlo, 2007). Nurses constitute 45% to 60% of the health care workforce in subSaharan Africa, with high nurse-to-physician ratios, and one of the lowest nurse-to-patient ratios in the world (Dovlo, 2007). The nursing shortage in subSaharan countries can be attributed to a number of factors, including: economics, politics, education, the HIV pandemic, work organization, and bureaucracy, all of which potentially affect the work environment of nurses and midwives by causing them to work long hours in critically understaffed facilities (Dovlo, 2007). Consequences of these shortages include an increasingly stressful work environment, placing the nurses at risk for violence and stressrelated diseases, ultimately compromising the quality of care. In 2002, a report on workplace violence in Australia, Brazil, Bulgaria, Lebanon, Portugal, South Africa, and Thailand, sponsored by the International Labor Office, International Council of Nurses, World Health Organization, and Public Services International, acknowledged the extent and severity of the problem of workplace violence in the health care setting (Di Martino, 2002). The report indicated that a majority of health care workers in each country had experienced at least one incident of physical or psychological violence in the previous year (67.2% in Australia, 46.7% in Brazil, 75.8% in Bulgaria, 61% in South Africa, 54% in Thailand, and, in Portugal, 60% in a health center and 37% in a hospital; Di Martino, 2002). In the United States, a high percentage of workplace incidents and deaths have been attributed to violence within health care settings (Centers for Disease Control and Prevention/National Institute for Occupational Safety and Health, 2002; Lipscomb, Silverstein, Slavin, Cody, & Jenkins, 2002; Peek-Asa, Cubbin, & Hubbell, 2002; PeekAsa et al., 2009; U.S. Department of Labor, Bureau of Labor Statistics [BLS], 2006). In each year from 1993 to 1999, 1.7 million incidents of violence occurred in the workplace (Duhart, 2001). The BLS
El Ghaziri et al. / Workplace Violence Among Nurses and Midwives in Sub-Saharan Africa
(2006) reported that 45% of all nonfatal assaults leading to lost workdays against workers took place in the health care sector, where patients were the perpetrators in 50% of the assaults. In 2005, the rate of nonfatal assaults among all health care workers combined was 8.8/10,000 cases (BLS, 2006). Mental health professionals had an annual incidence rate of 68.0/1,000 workers, and nurses had an annual incidence rate of 22.0/1,000 workers (BLS, 2006). Data from the United Kingdom have indicated that one third of nurses had been either abused or assaulted at work (Harulow, 2000); in Ireland, 60% of emergency department nurses and attendants had been physically assaulted at least once (Rose, 1997). Canadian researchers reported that 80% of nurses had experienced some form of violence at work (Cruikshank, 1995), and, in Australia, nursing as a profession was found to have the highest rate of exposure to violence, exceeding that perpetrated against prison and police officers (Perrone, 1999). Despite the scarcity of literature on workplace violence in Africa, injury patterns in Africa seem to be similar to those in developed countries (Forjouh, Zwi, & Mock, 1998), where reports from some sub-Saharan countries revealed that injury accounted for a considerable proportion of all deaths and morbidities (Forjouh et al., 1998). In a study investigating the prevalence and pattern of occupational health hazards in the obstetrics and gynecology unit of a hospital in Nigeria (n 5 78), the following were reported: work-related stress (83.3%), needle-stick injuries (75.6%), bloodstains on skin (73.1%), sleep disturbance (42.3%), skin reactions (37.2%), and hepatitis (8.9%). It is worth noting that almost one fourth of the staff (24.3%) had been assaulted by patients (Orji, Fasubaa, Onwudiegwu, Dare, & Ogunniyi, 2002). Another Nigerian study examined the prevalence of workplace violence (James, Isa, & Oud, 2011) and found that verbal aggression was the most prevalent type of violence experienced by nurses working in psychiatric facilities (67.2% occasionally/sometimes and 32.8% often/frequently). Among the 73 nursing staff, male nurses were more likely to have experienced physical violence and aggressive ‘‘spitting’’ behaviors, while nurses with more than 10 years of experience were more likely to have experienced verbal and humiliating aggressive behaviors.
S81
Among Southern Nigerian dental health professionals (n 5 175), the prevalence of violence was 31.9%, and violence was associated with long waiting times (27.3%), cancellation of appointments (13.6%), outcomes of treatment (11.4%), alcohol intoxication (9.1%), psychiatric mental health issues (6.8%), patient bills (4.5%), and others (27.3%). Types of violence included nonphysical violence in the form of loud shouting (50.0%), threats (22.7%), sexual harassment (6.8%), and swearing (2.3%), with bullying and hitting constituting the remaining 18.2%. The perpetrators of the violence were mainly patients (54.5%) and their relatives/friends (18.2%). The consequence of the violence resulted in reports of fear (18.2%), impaired job performance (15.9%), psychological problems (13.6%), and lost time (9.1%) (Azodo, Ezeja, & Ehikhamenor, 2011). Our paper describes the prevalence of Type II workplace violence in a sample of nurses and midwives practicing in sub-Saharan African countries. In addition, the paper identifies risk factors that, if addressed through prevention activities, might reduce the risk of violence toward health care workers in sub-Saharan Africa.
Methods Study Sample Data are from a sample of 756 health care workers who completed a questionnaire during the African Midwives Research Network biennial conference in Nairobi, Kenya (2007); Dar Es Salaam, Tanzania (2009); and nursing and midwifery meetings held throughout Nigeria from 2008 through 2010. The questionnaire included items about work schedule; work organization; patient/client factors such as alcohol use, mental status, and presence of a weapon; worker exposure to blood and other bodily fluids; knowledge and use of universal precautions; HIVrelated stigma; and demographics of nurses and midwives. This paper focuses on work organization and schedule, client and patient characteristics, and demographics of the nurses and midwives. Slight variations were noted across countries in some survey questions, which necessitated recoding and variable computation to ensure uniformity across the dataset.
S82 JANAC Vol. 25, No. 1S, January/February 2014
A total of 756 health care workers in attendance at the three meetings responded to the questionnaire. Out of the total sample, 722 were nurses and nurse midwives who provided direct clinical care to patients in sub-Saharan African countries; 10 resided outside of sub-Saharan countries (i.e., Australia, Sweden, the United Kingdom, and the United States). Therefore, 712 nurses and midwives meeting our eligibly criteria were included in the analyses. The University of Maryland–Baltimore institutional review board approved the project as exempt, as did the ethics committees in Kenya (Kenyan Medical Research Institute), Nigeria (National Health Research Ethics Committee), and Tanzania (National Institute for Medical Research). Measures Type II workplace violence was assessed by examining nurses’ and midwives’ previous 12-month exposures to patients, family members, or patients’ friends who had: (a) yelled or sworn at them; (b) threatened to assault them; or (c) physically hurt them (outcome variables), and five client characteristics as depicted in Table 1. The five items described as client characteristics assessed whether the exposure to violence involved a patient, family member, or patient’s friend who had mental illness, was intoxicated, had guns or visible weapons, or had a history of assault. Table 1 also documents whether the respondent often dealt with difficult patients, family members, or patients’ friends. The five client-characteristics items were analyzed individually and then combined into an index for Table 1.
subsequent analysis. A total index score was derived by summing the number of questions answered as yes. Scores ranged between 0 and 5, with a higher score representing more risk. The mean 6 standard deviation index score was 2.94 6 1.6. The five-item index demonstrated good internal consistency, with a Cronbach’s alpha of 0.74. Data Analysis Descriptive analyses were performed to describe demographic variables, work status, work schedule, and clients’ characteristics. Bivariate analyses (c2 tests) were performed to assess the relationship between outcome and independent variables. Because of the clustering effects of nurses within residing country (e.g., the interclass correlation 5 0.0928 for outcome of physically hurt/assaulted, data not shown), generalized estimating equation (GEE) logistic regression models, accounting for clustering effects, were performed to assess the associations between each aspect of workplace violence and independent variables (work status, work schedule, and client characteristics) separately, after controlling for the covariates. Covariates that were significant or marginally significant on bivariate analysis were included in the multivariate GEE logistic regression models. Multicollinearity of the predictors was checked using variance inflation factor. The significance level was set at 0.05, and two-tailed 95% confidence intervals (CI) were obtained for all analyses. All analyses were performed using SPSS 19.0 and STATA 12 software.
Descriptive Statistics of Workplace Violence Experience and Client Characteristics (N 5 712)a Parameter
Client risk characteristics I often deal with difficult patients/family/friends In the past 12 months I have worked with patients, family members, patients’ friends who had history of assault In the past 12 months I have worked with patients, family members, patients’ friends who had mental Illness In the past 12 months I have worked with patients, family members, patients’ friends who were intoxicated In the past 12 months I have worked with patients, family members, patients’ friends who had guns or weapons visible Workplace violence experience In the past 12 months I have worked with patients, family members, patients’ friends who yelled or swore at me In the past 12 months I have worked with patients, family members, patients’ friends who threatened to assault me In the past 12 months I have worked with patients, family members, patients’ friends who physically hurt/assaulted me a. Numbers of respondents may not sum to total due to missing data.
n
%
501 444 414 352 163
78.0 69.9 65.4 55.6 25.9
263 224 125
40.8 34.7 19.9
El Ghaziri et al. / Workplace Violence Among Nurses and Midwives in Sub-Saharan Africa
Results The sample included 712 nurses and midwives who were present at one of the three meetings and who completed the questionnaire. Participants resided in more than 12 sub-Saharan African countries, with the majority from Kenya (n 5 254), Nigeria (n 5 177), and Tanzania (n 5 169). A minority number of participants came from Uganda (n 5 31), Zimbabwe (n 5 6), Zambia (n 5 4), Ethiopia (n 5 2), Namibia (n 5 2), Central African Republic (n 5 1), Eritrea (n 5 1), Malawi (n 5 1), Togo (n 5 1), and other sub-Saharan African countries (n 5 63, unspecified). Six hundred nurses and midwives (84.3%) in the sample resided in Kenya (35.7%), Nigeria (24.9%), or Tanzania (23.7%), the three sites where the meetings occurred, which allowed us to account for the clustering within residing country. Table 1 presents data from the entire sample (n 5 712) and shows that 40.8% reported having been called names or been yelled or sworn at by patients, family members, or patients’ friends during the previous 12 months, while 34.7% reported being threatened to be physically hurt or assaulted (hit, slapped, punched, kicked, strangled); 19.9% were actually physically hurt or assaulted. Seventy-eight percent often dealt with difficult patients, family members, and patients’ friends, while 69.9% had worked with clients who had a history of assault. The participants also reported that they had worked with clients who had mental illness (65.4%), were intoxicated (55.6%), or had visible guns or weapons (25.9%). Table 2 presents the sample’s distribution and the bivariate associations between the demographics and work organization variables, and each of the three outcome variables. The mean age of the participants was 42 6 9 years. The majority of the participants were female (79.7%), held certificates and diplomas rather than the bachelor’s degree (75.3%), and worked in urban areas (73.1%). Approximately 37% reported direct patient care for 50% to 80% of their time, while 41.2% spent more than 80% of their time in direct patient care. Almost 36% reported working more than 40 hours per week. Thirty percent worked ‘‘off-hours’’ more than once a week, and more than half were required to visit patients in their homes.
S83
The number of working hours per week was associated with all three measures of workplace violence. That is, a higher percentage of nurses or midwives who worked 40 hours or more per week reported that they had been yelled or sworn at (51.2%), threatened with assault (44.9%), or physically hurt/assaulted (27.6%) by patients, patients’ family members, or patients’ friends in the previous 12 months, compared with those who worked 40 hours or less per week (all, p , .001; Table 2). Similarly, age was significantly associated with participants having worked in the previous 12 months with patients, family members, or patients’ friends who had yelled or sworn at them and threatened to assault them. For example, more than half of nurses or midwives (51.1%) in the age range of 22–35 years reported that they had been yelled or sworn at by patients, patients’ family members, or patients’ friends in the previous 12 months, compared with 36.1% and 40.3% of participants in the 36–45 and 46 or older age groups, respectively (p , .001). About 40% of the nurses or midwives ages 22–35 years reported having been threatened with assault, while 37% of participants between 36 and 45 years of age and 28.9% of those ages 46 years or older reported having been threatened with assault in the previous 12 months (p , .001). Furthermore, percentage of time spent in direct patient care was associated with being actually physically hurt or assaulted (p , .001). Of nurses and midwives with greater than 80% of time spent in direct patient care, approximately one third (28%) reported having been actually physically hurt or assaulted by patients, family members, or patients’ friends, compared with 17% of those who worked 50%–80% of their time in direct patient care, and 10.9% of those who worked less than 50% of their time in direct patient care. In addition, country of residence was associated with participants’ having worked in the previous 12 months with patients, family members, or patients’ friends who had yelled or sworn at them, threatened to assault them, and actually physically hurt or assaulted them (all, p , .05). Kenya had the highest prevalences of participants who had been yelled or sworn at (47.2%) and threatened with physical hurt or assault (42.5%), while Nigeria had the
Descriptive Statistics of the Sample in Relation to Workplace Violence (N 5 712)a
Parameter Gender Male Female Age group 22–35 years 36–45 years $46 years Education Certificate or other Baccalaureate degree Master’s or doctoral Work location Urban Rural Both Country Kenya Nigeria Tanzania Other sub-Saharan African country Percentage of time spent in direct patient care ,50 50–80 .80 Hours per week 0–40 .40 Required to visit patients at their homes Yes No Frequency of working off-hours Zero or once a week More than once a week
N
%
Yelled or Swore at Me Threatened to Assault Me Yes, n (%) No, n (%) p Value Yes, n (%) No, n (%) p Value .275
144 20.3 564 79.7
46 (36.8) 210 (42.2)
79 (63.2) 288 (57.8)
.604 48 (36.9) 170 (34.5)
82 (63.1) 323 (65.5)
.009b 201 29 225 32.5 266 38.4
94 (51.1) 73 (36.1) 89 (40.3)
90 (48.9) 129 (63.9) 132 (59.7)
387 75.3 70 13.3 60 11.4
136 (38.5) 22 (36.7) 13 (31)
217 (61.5) 38 (63.3) 29 (69)
516 73.1 172 24.4 18 2.5
181 (39.7) 65 (41.9) 5 (45.5)
275 (60.3) 90 (58.1) 6 (54.5)
75 (40.3) 76 (37.4) 63 (28.9)
111 (59.7) 127 (62.6) 155 (71.1)
118 (33.1) 25 (40.3) 11 (28.2)
238 (66.9) 37 (59.7) 28 (71.8)
145 (32.2) 62 (39) 6 (50)
306 (67.8) 97 (61) 6 (50)
122 (52.8) 111 (84.8) 85 (57) 52 (57.1)
144 22.2 237 36.6 267 41.2
46 (37.7) 101 (47.2) 97 (39.3)
76 (62.3) 113 (52.8) 150 (60.7)
416 64.3 231 35.7
140 (36.9) 107 (51.2)
239 (63.1) 102 (48.8)
366 54.9 301 45.1
148 (44.7) 103 (37.6)
183 (55.3) 171 (62.4)
457 70 196 30
163 (39.4) 85 (47.8)
251 (60.6) 93 (52.2)
139 (77.7) 154 (78.6) 179 (82.9)
80 (23.1) 14 (23.7) 3 (7.7)
266 (76.9) 45 (76.3) 36 (92.3)
87 (19.6) 28 (18.5) 2 (15.4)
356 (80.4) 123 (81.5) 11 (84.6)
.375
.081
.898
,.001 b
.020b 99 (42.5) 50 (31.4) 42 (28) 33 (36.3)
134 (57.5) 109 (68.6) 108 (72) 58 (63.7)
38 (31.1) 78 (36.3) 93 (38.1)
84 (68.9) 137 (63.7) 151 (61.9)
111 (29.4) 93 (44.9)
267 (70.6) 114 (55.1)
127 (38) 85 (31.4)
207 (62) 186 (68.6)
139 (33.8) 71 (39.2)
272 (66.2) 110 (60.8)
.132
20 (8.8) 51 (32.1) 32 (22) 18 (21.2)
207 (91.2) 108 (67.9) 114 (78) 67 (78.8)
13 (10.9) 36 (17) 66 (28)
106 (89.1) 176 (83) 170 (72)
56 (15) 55 (27.6)
317 (85) 144 (72.4)
71 (21.8) 44 (16.7)
255 (78.2) 219 (83.3)
76 (18.9) 40 (22.9)
326 (81.1) 135 (77.9)
,.001b
.421
,.001 b
.001b
.077
,.001b
.088
.058
a. Numbers (n) may not sum to total due to missing data. b. p , .05.
40 (22.3) 42 (21.4) 37 (17.1)
.15
.015b 109 (47.2) 50 (38.2) 64 (43) 39 (42.9)
96 (75.6) 390 (81.3)
.408
.843
35.67 24.86 23.74 15.73
.156 31 (24.4) 90 (18.7)
.041b
.624
254 177 169 112
Physically Hurt/Assaulted Me Yes, n (%) No, n (%) p Value
.124
.205
.276
S84 JANAC Vol. 25, No. 1S, January/February 2014
Table 2.
El Ghaziri et al. / Workplace Violence Among Nurses and Midwives in Sub-Saharan Africa
highest prevalence of being actually physically hurt or assaulted (32.1%). Odds ratios based on the GEE logistic regression models are presented in Table 3. The client risk characteristics index was significantly associated with each of the three outcome variables of interest after adjusting for covariates. A unit increase in the client risk characteristics index was associated with a greater likelihood of experiencing violence (adjusted odds ratio [aOR] 5 1.78, 95% CI [1.52 to 2.10], p , .001 for being yelled or sworn at; aOR 5 1.59, 95% CI [1.33 to 1.90], p , .001 for being threatened with assault; aOR 5 1.39, 95% CI [1.13 to 1.70], p 5 .002 for being actually physically hurt or assaulted). Compared with nurses who worked less than 40 hours a week, nurses who worked more than 40 hours a week were more likely to have experienced violence (aOR 5 2.1, 95% CI [1.29 to 3.4], p 5 .003 for being yelled or sworn at; aOR 5 2.15, 95% CI [1.34 to 3.48], p 5 .001 for being threatened; aOR 5 1.72, 95% CI [1.02 to Table 3.
S85
2.84], p 5 .042 for being actually physically hurt or assaulted). No multicollinearity was found of the predictor variables (variance inflation factor ,10). Other variables, including age, being required to visit patients’ homes, percentage of time spent in direct patient care, and frequency of working off-hours, were not significantly related to any aspect of having experienced violence in adjusted GEE logistic regression models.
Discussion Exposure to Type II workplace violence among sub-Saharan African nurses and midwives included being yelled or sworn at, threatened with assault, and actually being physically assaulted. Age was associated with exposure to being yelled or sworn at and with threats of assault. The average number of hours worked per week was significantly associated with all of the outcomes, and percentage of
Odds Ratios Estimates of Workplace Violence in Relation to Independent Variables Based on GEE Logistic Regression (N 5 712)a Yelled or Swore at Me Parameter
Age group 22–35 years (Reference) 36–45 years $46 years Required to visit patients at their homes No (Reference) Yes Percentage of time spent in direct patient care ,50 (Reference) 50–80 .80 Hours per week 0–40 (Reference) .40 Frequency of working off-hours Zero or once a week (Reference) More than once a week Client risk characteristics index
Threatened to Assault Me
Physically Hurt/Assaulted Me
aOR
95% CI
p Value
aOR
95% CI
p Value
aOR
95% CI
p Value
0.78 0.95
0.45–1.35 0.55–1.66
.382 .868
1.7 1.16
0.97–2.99 0.64–2.10
.063 .625
1.21 0.98
.66–2.20 0.51–1.85
.537 .939
1.08
0.68–1.72
.735
1.06
0.66–1.71
.784
1.28
0.76–2.17
.352
1.02 0.69
0.56–1.86 0.39–1.21
.947 .198
0.79 1.22
0.42–1.51 0.65–2.32
.489 .528
1.23 1.86
0.58–2.59 0.87–3.97
.594 .109
2.1
1.29–3.4
.003b
2.15
1.35–3.48
.001b
1.72
1.02–2.84
.042b
0.96 1.78
0.58–1.57 1.52–2.1
.861 ,.001b
1.12 1.59
0.67–1.89 1.33–1.9
.655 ,.001b
0.96 1.39
0.57–1.76 1.13–1.70
.999 .002b
Note. aOR 5 adjusted odds ratio (adjusted for all the other variables listed in the table); CI 5 confidence interval; GEE 5 generalized estimating equation. a. For country, 12 countries and 1 unspecified country were treated as clusters due to intraclass correlation within countries. b. p , .05.
S86 JANAC Vol. 25, No. 1S, January/February 2014
time spent in direct care was significantly associated only with having been actually physically assaulted. When the GEE model was employed, the average number of work hours per week and the risk index for client characteristics were significantly associated with the three workplace violence outcomes. We found that the rate of exposure to being yelled or sworn at in this population (40.8%) was within the range of verbal abuse (27%–67%) experienced by health care respondents in the World Health Organization country report (Di Martino, 2002), with 49.5% exposure to verbal abuse noted among the respondents in South Africa (Di Martino, 2002; Steinman, 2003). Nigerian nurses in psychiatric facilities reported a similar frequency of verbal aggression (James et al., 2011). By comparison, our findings were slightly less than the prevalence of loud shouting experienced by the dental health professionals (50%) in southern Nigeria (Azodo et al., 2011). The rate of exposure to physical violence (19.9%) was higher than the range experienced by midwives and nurses in South Africa (12%–14.1%; Di Martino, 2002; Steinman, 2003) but comparable to the bullying and hitting (18.2%) experienced by the dental health professionals in Southern Nigeria (Azodo et al., 2011). By contrast, 24.3% of nurses in an obstetrics and gynecology department in Nigeria had been assaulted by patients (Orji et al., 2002). Work organization and occupational risks associated with workplace violence in the literature include the number of hours worked, time pressures, and shift work (Campbell et al., 2011; Whittington, Shuttleworth, & Hill, 1996). In our sample, we found that nurses and midwives who worked more than 40 h/week were at higher risk for Type II workplace violence. Whereas the number of hours associated with co-worker violence among European nurses was 35 or more hours per week, where this association was not noted for exposure to Type II workplace violence (Camerino, Estryn-Behar, Conway, Van Der Heijden, & Hasselhorn, 2008). Perpetrator/client characteristics associated with violence have been shown to include male gender; a diagnosis of borderline personality disorder, schizophrenia, or organic mood disorder; substance abuse issues; or having a history of violence (Little, 1999). In our sample, the client characteristics associ-
ated with violence included patients, family, and/or friends who were difficult or intoxicated, had a mental illness, possessed visible weapons or guns, and had histories of assault. Despite the prevailing nursing shortage in sub-Saharan Africa, nurses and midwives reporting more than 80% of time spent in direct patient care were not at increased risk of Type II workplace violence as was illustrated in other studies (Whittington et al., 1996). The majority of our sample was female (79.7%), and gender was not associated with workplace violence. The association between age and exposure to Type II workplace violence was not evident among the sub-Saharan African nurses and midwives, which is inconsistent with literature that has shown that being younger and less experienced (Nolan, Soares, Dallender, Thomsen, & Arnetz, 2001) put nurses at higher risk and frequency of being exposed to Type II workplace violence (Camerino et al., 2008). A nurse’s level of education was not associated with any of the Type II workplace violence outcomes, which is inconsistent with other literature that reflected that nurses having an associate’s degree compared to other degrees increased the risk for exposure to workplace violence (Nachreiner et al., 2007). This variation may be attributed to limited variability due to the fact that most of the nurses and midwives involved in direct patient care were holding certificates or diplomas rather than bachelor’s degrees (75.3%), which limited variability in our study. Limitations Study limitations include the following. The crosssectional design of the study limited our ability to infer causality. Self-reported data may include information bias. Convenience samples limit the generalizability of results. Another design-related limitation was the slight variations noted across countries in terms of several survey questions, which necessitated recoding, and variable computation to ensure uniformity across the dataset. The strengths of the study reside in the fact that it was one of the first to examine questions related to workplace violence in this population. The study involved a multinational sample (12 specified and other sub-Saharan countries) with a high-risk
El Ghaziri et al. / Workplace Violence Among Nurses and Midwives in Sub-Saharan Africa
population of nurses and midwives. Moreover, it used the GEE as a method for determining the factors associated with exposure to workplace violence. The alarming nature of this global problem has prompted organizations such as the World Health Organization, International Council of Nurses, Occupational Safety and Health Administration, and the International Labor Office to develop prevention guidelines and position statements. Despite the complexity and severity of the problem, several initiatives have demonstrated the feasibility and effectiveness of guidelines for prevention, reporting, and training in developed countries such as the United States (Lipscomb et al., 2002; Lipscomb et al., 2006; McPhaul & Lipscomb, 2004). It is crucial to develop interventions that are informed by direct care staff experience and that are culturally appropriate to the setting and population. The use of specially trained security staff for highrisk situations, training in techniques for defusing aggression, the use of metal detectors at entrances, and specialized security police protections when violent patients or their families are on-premises are recommended, but they must be customized to meet the social and economic conditions in the respective countries and regions. Other interventions, including postincident assault services (medical care, mandatory reporting, immediate discharge or removal of perpetrator, access to counseling, and being able to leave work following an incident; May & Grubbs, 2002; McPhaul & Lipscomb 2004), should be explored while ensuring management commitment and employee involvement throughout the entire process (Lipscomb et al., 2006; McPhaul & Lipscomb, 2004).
Conclusions We can infer that long work hours (40 hours or more per week) and certain client characteristics are associated with experiencing Type II workplace violence. Further research is needed to assess additional work-related stressors that might be associated with Type II workplace violence. These findings support the need to pursue policy and organization level initiatives to adjust work hours to minimize the risk for Type II workplace violence. Our finding
S87
of a significant association between client characteristics and workplace violence indicates the need for a thorough assessment of such characteristics upon admission and the development of strategies to increase security or other interventions when encountering high-risk patient visits in order to maintain staff safety.
Key Considerations Violence against health care workers perpetrated by clients and/or their friends and families (Type II) is a growing problem that can severely impact health care delivery. Policy and organizational level interventions are needed to minimize nurses’ and midwives’ exposures to Type II workplace violence by identifying risky clients and addressing long work hours. Policy and organizational level initiatives are needed to adjust work hours to minimize the risk for experiencing Type II workplace violence. A thorough assessment of identified client risk characteristics should be completed upon admission. Strategies should be developed to increase security when encountering high-risk patient visits in order to maintain staff safety.
Disclosures The authors report no real or perceived vested interests that relate to this article that could be construed as a conflict of interest.
References Azodo, C. C., Ezeja, E. B., & Ehikhamenor, E. E. (2011). Occupational violence against dental professionals in southern Nigeria. African Health Sciences, 11(3), 486-492. Barling, J. (1996). The prediction, experience, and consequences of workplace violence. In G. VandenBos, & E. Bulatao (Eds.), Violence on the job: Identifying risks and developing
S88 JANAC Vol. 25, No. 1S, January/February 2014 solutions (pp. 29-49). Washington, DC: American Psychological Association. Beech, B., & Leather, P. (2006). Workplace violence in the health care sector: A review of staff training and integration of training evaluation models. Aggression and Violent Behavior, 11, 27-43. http://dx.doi.org/10.1016/j.avb.2005. 05.004 Camerino, D., Estryn-Behar, M., Conway, P. M., Van Der Heijden, B. I. J. M., & Hasselhorn, H.-M. (2008). Workrelated factors and violence among nursing staff in the European NEXT study: A longitudinal cohort study. International Journal of Nursing Studies, 45(1), 35-50. Campbell, J., Messing, J. T., Kub, J., Agnew, J., Fitzgerald, S., Fowler, B., . Bolyard, R. (2011). Workplace violence: Prevalence and risk factors in the Safe at Work Study. Journal of Occupational & Environmental Medicine, 53(1), 82-89. http://dx.doi.org/10.1097/JOM.0b013e3182028d55 Centers for Disease Control and Prevention/Department of Health and Human Services, National Institute for Occupational Safety and Health (NOISH). (1996). Violence in the workplace—Current Intelligence Bulletin 57. NOISH Publications and Products. Retrieved from http://www.cdc.gov/ niosh/docs/96-100/ Centers for Disease Control and Prevention/Department of Health and Human Services, National Institute for Occupational Safety and Health (NOISH). (2002). Violence: Occupational hazards in hospitals. Publication No. 2002-101. NOISH Publications and Products. Retrieved from http:// www.cdc.gov/niosh/docs/2002-101/ Cruikshank, C. (1995). Nurse abuse: A study of workplace violence in Nova Scotia. Nurse to Nurse, 6(2), 12. Di Martino, V. (2002). Workplace violence in the health sector: Country case studies (Brazil, Bulgaria, Lebanon, Portugal, South Africa, Thailand and an additional Australian study): Synthesis Report. ILO, ICN, WHO and PSI Joint Program on Workplace Violence. Retrieved from http://www.who. int/violence_injury_prevention/injury/en/WVsynthesisreport. pdf Dovlo, D. (2007). Migration of nurses from Sub-Saharan Africa: A review of issues and challenges. Health Services Research, 42(3), 1373-1388. Duhart, D. T. (2001). Violence in the workplace, 1993-99. U.S. Department of Justice, Office of Justice Programs, Bureau of Justice Statistics, National Crime Victimization Survey. Retrieved from http://bjs.gov/content/pub/pdf/vw99.pdf Forjouh, S. N., Zwi, A. B., & Mock, C. M. (1998). Injury control in Africa: Getting governments to do more. Tropical Medicine and International Health, 3(5), 349-356. http://dx.doi. org/10.1046/j.1365-3156.1998.00240.x Harulow, S. (2000). Ending the silence on violence. Australian Nursing Journal, 7(10), 26-29. James, B. O., Isa, E. W., & Oud, N. (2011). Patient aggression in psychiatric services: The experience of a sample of nurses at two psychiatric facilities in Nigeria. African Journal of Psychiatry (Johannesburg), 14(2), 130-133. http://dx.doi.org/ 10.4314/ajpsy.v14i2.4
Lipscomb, J., McPhaul, K., Rosen, J., Brown, J. G., Choi, M., Soeken, K., . Porter, P. (2006). Violence prevention in the mental health setting: The New York State experience. Canadian Journal of Nursing Research, 38(4), 96-117. Lipscomb, J., Silverstein, B., Slavin, T. J., Cody, E., & Jenkins, L. (2002). Perspectives on legal strategies to prevent workplace violence. Journal of Law, Medicine and Ethics, 30(Suppl. 3), 166-172. Little, L. (1999). Risk factors for assaults on nursing staff: Childhood abuse and education level. Journal of Nursing Administration, 29(12), 22-29. Madhava, V., Burgess, C., & Druker, E. (2002). Epidemiology of chronic hepatitis C virus infection in Sub Saharan Africa. The Lancet Infectious Diseases, 2(5), 293-302. May, D., & Grubbs, L. (2002). The extent, nature, and precipitating factors of nurse assault among three groups of registered nurses in a regional medical center. Journal of Emergency Nursing, 28(1), 11-17. http://dx.doi.org/10.1067/ men.2002.121835 McPhaul, K. M., & Lipscomb, J. A. (2004). Workplace violence in health care: Recognized but not regulated. Online Journal of Issues in Nursing. Retrieved from www.nursingworld.org/ MainMenuCategories/ANAMarketplace/ANAPeriodicals/ OJIN/TableofContents/Volume92004/No3Sept04/Violence inHealthCare.aspx McPhaul, K. M., London, M., & Lipscomb, J. A. (2013). A framework for translating workplace violence intervention research into evidence based programs. Online Journal of Issues in Nursing, 18(1), 4. Nachreiner, N. M., Hansen, H. E., Okano, A., Gerberich, S. G., Ryan, A. D., McGovern, M. P., . Watt, G. D. (2007). Difference in work-related violence by nurse license type. Journal of Professional Nursing, 23(5), 290-300. http://dx.doi.org/10. 1016/j.profnurs.2007.01.015 Nolan, P., Soares, J., Dallender, J., Thomsen, S., & Arnetz, B. (2001). A comparative study of the experiences of violence of English and Swedish mental health nurses. International Journal of Nursing Studies, 38(4), 419-426. Orji, E. O., Fasubaa, O. B., Onwudiegwu, U., Dare, F. O., & Ogunniyi, S. O. (2002). Occupational health hazards among health care workers in an obstetrics and gynecology unit of a Nigerian teaching hospital. Journal of Obstetrics and Gynecology, 22(1), 75-78. Peek-Asa, C., Casteel, C., Allareddy, V., Nocera, M., Goldmacher, S., O’Hagen, E., . Harrison, R. (2009). Workplace violence prevention programs in psychiatric units and facilities. Archives of Psychiatric Nursing, 23(2), 166-176. http://dx.doi.org/10.1016/j.apnu.2008.05.008 Peek-Asa, C., Cubbin, L., & Hubbell, K. (2002). Violent events and security programs in California emergency departments before and after the 1993 hospital security act. Journal of Emergency Nursing, 28(5), 420-426. http://dx.doi.org/10. 1067/men.2002.127567 Perrone, S. (1999). Violence in the workplace. Australian Institute of Criminology, Research and Public Policy Series, No. 22. Retrieved from http://www.aic.gov.au/documents/4/
El Ghaziri et al. / Workplace Violence Among Nurses and Midwives in Sub-Saharan Africa 1/0/%7B410FD288-AD82-4E4E-A6FE-C7C30D283528% 7Dfull_report.pdf. Rose, M. (1997). A survey of violence toward nursing staff in one large Irish accident and emergency department. Journal of Emergency Nursing, 23(3), 214-219. Schmidtke, R. (2011). Workplace violence: Identification and prevention. Journal of Global Health Care Systems. Retrieved from http://jghcs.info/index.php/j/article/view/22/ pdf Smith-Pittman, M. H., & McKoy, Y. D. (1999). Workplace violence in healthcare environments. Nursing Forum, 34(3), 5-13. http://dx.doi.org/10.1111/j.1744-6198.1999.tb00988.x Steinman, S. (2003). Workplace violence in the health sector— Country case study: South Africa. Retrieved from http://www. who.int/violence_injury_prevention/violence/interpersonal/en/ WVcountrystudysouthafrica.pdf University of Iowa Injury Prevention Research Center. (2001). Workplace violence—A report to the nation. University of Iowa Injury Prevention Research Center Web Site. Retrieved from http://www.public-health.uiowa.edu/iprc/resources/work place-violence-report.pdf U.S. Department of Labor, Bureau of Labor Statistics. (2006). Survey of workplace violence prevention, 2005. Retrieved from http://www.bls.gov/iif/oshwc/osnr0026.pdf U.S. Department of Labor, Occupational Safety and Health Administration. (2004). Guidelines for preventing workplace violence for healthcare & social service workers. Retrieved from http://www.osha.gov/Publications/OSHA3148/osha3148. html Whittington, R., Shuttleworth, S., & Hill, L. (1996). Violence to staff in a general hospital setting. Journal of Advanced Nursing, 24(2), 326-333. http://dx.doi.org/10.1046/j.13652648.1996.18114.x
S89
