bs_bs_banner
International Journal of Mental Health Nursing (2021) 30 (Suppl. 1), 1342–1353 doi: 10.1111/inm.12878
ORIGINAL ARTICLE
Measures and narratives of the nature, causes and
consequences of violent assaults and risk percep-
tion of psychiatric hospitals in Ghana: Mental
Health workers’ perspectives
Roger A. Atinga,1 Lily Yarney,1 Kingsley Saa-Touh Mort,2 Joshua A. Gariba3 and
Joana Salifu Yendork4
1Department of Public Administration and Health Services Management, University of Ghana Business School,
Accra, 2Department of Social Work, School of Social Sciences, University of Ghana, Accra, 3Department of
Sociology, School of Social Sciences, University of Ghana, Accra, and 4Department of Psychology, School of Social
Sciences, University of Ghana, Accra, Ghana
ABSTRACT: Literature shows that mental healthcare workers in Ghana face incessant fear of
patient violence that compromise safe care delivery. However, the nature, scale, perceived causes,
and consequences of these assaults and how they shape risk perceptions have received limited
empirical attention, hence the need for this study. The study employed sequential explanatory
mixed methods where questionnaire administration preceded and informed the design of an
interview guide used for in-depth interviews with health workers in referral psychiatric hospitals.
Descriptive statistics and multivariate logistic regressions were used to analyze the quantitative
data. Qualitative data were transcribed and analyzed thematically. Findings showed that physical
and non-physical violent assaults and risk perceptions of the hospitals were statistically and
significantly associated with females (P < 0.01), nurses (P < 0.01), other clinical cadre (P < 0.01),
and those with low job tenure (P < 0.05). About 57% and 71% of the sample reported
experiencing physical and non-physical assaults, respectively. Major and minor injuries and
psychosocial problems were frequent sequelae following physical violent assaults. As a result, 80%
of the participants perceived the hospitals environment to be unsafe to provide care. Violent
assaults compromises safety and care delivery efforts suggesting the need for systematic
interventions to minimize mental healthcare workers exposure to patient violence.
KEY WORDS: mental health worker, patient violence, psychiatric hospital, safety, violent assault.
INTRODUCTION
Correspondence: Roger A. Atinga, Department of Public Admin- Mental health workers in Ghana are often constrained
istration and Health Services Management, University of Ghana
Business School, Legon, Accra, Ghana. Email: ayimbillah@yahoo.- in their quest to deliver quality health services to
com patients in contact with psychiatric hospitals. Reports
Declaration of Interest: None declared. show that mental health providers work in complex,
Roger A. Atinga,PhD.
Lily Yarney,PhD. insecure environments and with limited resources, and
Kingsley Saa-Touh Mort,PhD. these expose them to occupational hazards such as
Joshua A. Gariba,PhD.
Joana Salifu Yendork,PhD. patient violent assaults (Alhassan & Poku 2018; Roberts
Accepted April 19 2021. et al. 2014). However, perspectives of the nature, scale,
© 2021 John Wiley & Sons Australia, Ltd
VIOLENT ASSAULTS IN PSYCHIATRIC HOSPITALS 1343
causes and aftermath of patient violent assaults, and In Ghana, mental health care is delivered in psychi-
the specific cadre of mental health workers at risk of atric units of district and regional hospitals, but special-
assaults remains understudied, hence the need for this ist care is provided in three main referral psychiatric
study. Violent assault according to the World Health hospitals, all of which are located in two coastal
Organization (WHO) is ‘the intentional use of physical regions. This demonstrates weak referral pathways and
force or power, threatened or actual, against oneself, constrained access to quality mental health services by
another person, or against a group or community, that many populations in the country. The hospitals provide
either results in or has a high likelihood of resulting in general psychiatric care and specialized services for
injury, death, psychological harm, maldevelopment, or geriatric, pediatric, and forensic psychiatry (Roberts
deprivation’ (WHO 2002, p. 4). et al. 2014). The free admission for all mentally ill
Drawing on the WHO definition, violent assaults patients coupled with subsidized medications and
have been broadly categorized into physical assault examination in the hospitals has often produced over-
(inflicting harm or initiating unwanted contact on a crowding with prolonged patient stay and detention.
person) and non-physical assault (Kisa 2008). Patient Bed capacities in the hospitals are deficient to accom-
violence in mental health care has reached epidemic modate the monthly average of between 300 to 1200
proportions and keeps increasing due to infrastruc- patients seeking treatment for mental disorders (Jack
tural, staffing, and resources gaps at points of care, et al. 2013). As a result, some patients on admission
especially in low- and middle-income settings (Bruck- often sleep on concrete floors and locally prepared thin
ner et al. 2011; Fricchione et al. 2012). Globally, an mats until beds are available. Ghana has 0.1 and 7.7
estimated 60% of mental health workers are exposed psychiatrists and nurses, respectively, per 100 000 pop-
to violent assaults annually (d’Ettorre & Pellicani ulation working in the mental health sector (WHO
2017; Olashore et al. 2018). In all acts of patient vio- 2019), reflecting significant deficits of critical human
lence in mental health care, the most affected have resources in psychiatric hospitals. Clinical psychologists,
been nurses (El Ghaziri et al. 2014; Niu et al. 2019; occupational therapists, community psychiatric nurses,
Spector et al. 2014), novice clinicians (Antonius et al. and social workers working in the mental health sector
2010), and psychiatrists (Dhumad et al. 2007; Rueve are also woefully inadequate (Roberts et al. 2014).
& Welton 2008). There is also evidence of non- Ghana has a legal instrument setting up a Mental
clinical staff such as secretaries, receptionist, and Health Authority to strengthen governance, enforce
administrative staff being subjected to patient violence patient rights, safeguard health providers and ensure
(Anderson & West 2011). that mental health care is delivered optimally (Govern-
Violent assaults destabilize healthcare delivery in ment of Ghana 2012). However, infrastructural, struc-
several ways. Loss of working days due to injuries tural, and human resource deficits drive patient
related to hospitalization, low productivity, and effi- violence, undermining health providers’ safety and
ciency have been reported (Cashmore et al. 2012; adversely affecting retention of critical mental health
Roche et al. 2010). Health workers experiencing violent workforce (Jack et al. 2013). Health providers operate
assaults were found to report low job satisfaction, without adequate safety and protection from aggressive
increased occupational strains as well as minimal self- patients. Findings of a study in two psychiatric hospi-
confidence and quality of working life (Magnavita & tals showed that frontline staff complained about poor
Heponiemi 2011; Roche et al. 2010). Working amid safety which creates room for patient violence (Alhas-
fear of being abused verbally or physically decreases san & Poku 2018). Another study established that psy-
emotional stability and mental attachment to work, chiatric hospitals in Ghana are used as dumping
while flashbacks impair concentration during moments grounds for the mentally ill who are difficult to manage
of care (Camerino et al. 2008). In the long term, these at home (Fournier 2011) and that contributes to violent
consequences may produce incentives for sustained assaults. These suggest the need for explicit evidence
poor quality organizational behaviors. For instance, targeting patient violent assaults to understand why
health providers might simply devise default care they occur, the form they take and the consequences,
options for simplifying violent assaults, by engaging in those who are most exposed, and the implication for
selective treatment of patients with less aggressive health worker safety perceptions of mental healthcare
behaviors (Boafo & Hancock 2017), demonstrate poor environments.
care behaviors or strive less to achieve best therapeutic More importantly, context-based evidence on violent
outcomes of patients (Kindy et al. 2005). assaults in mental health care is crucial as
© 2021 John Wiley & Sons Australia, Ltd
1344 R. A. ATINGA ET AL.
generalization to other settings from the few empirical collect qualitative data to help explain and interpret
studies (Baby et al. 2014; Chen et al. 2010; McKinnon the phenomenon of violent assaults and safety of the
& Cross 2008; Shiao et al. 2010) can be problematic, hospitals. We iteratively examined this assumption
because health systems vary in structure, organization, drawing on sequential explanatory mixed method
processes, and delivery of mental health care. Also, vio- design to collect data between September and Decem-
lent assaults elicit varying interpretations, behaviors, ber 2018. Following the principles of this sequential
and responses from different mental health professional design type, data were collected in two distinct phases.
cadre (Lanctôt & Guay 2014). Thus, drawing on the Quantitative data collection preceded and informed the
mixed methods design, this study examined the nature, design and collection of qualitative data (Creswell &
perceived causes, experiences, and consequences of Clark 2011).
patient violent assaults among clinical and non-clinical
health workers in Ghana’s referral psychiatric hospitals. Quantitative process
How patient violence shapes staff perception about Sample respondents of the quantitative study com-
safety of the hospitals was also examined. With the pas- prised of nursing cadre (psychiatric, registered,
sage of the mental health law in Ghana, findings will enrolled, community, and public health nurses), other
guide formative decisions and actions towards building clinical cadre (health assistants, pharmacists, pharmacy
safety climate and culture that support the delivery of technicians, and mental health officers), and non-
optimal mental health care for effective therapeutic clinical staff (caterers, administrative managers and
outcomes. assistants, laundry staff, records managers, and techni-
cians). Given the numerical deficits of all cadre in the
hospitals, opportunistic sampling was employed to
MATERIALS AND METHODS
select respondents from the OPD, inpatient wards,
community psychiatry, occupational therapy, catering,
Study setting and design
administration, laundry, and other units. All the psychi-
The study was conducted in Ghana’s three main refer- atrists invited did not respond to the quantitative study.
ral pychiatric hospitals (Accra Psychiatric Hospital A questionnaire consisting of four parts was used to
(APH), Pantang Psychiatric Hospital (PPH), and Anka- collect data. The first part of the questionnaire cap-
ful Psychiatric Hospital (AnPH). The hospitals provide tured data on demographic characteristics such as age,
outpatient department (OPD) services and inpatient sex, education, profession, job tenure (number of years
care with a combined estimated 1322 beds. Only the in practice), and unit of work. Part two measured expo-
APH provides structured inpatient care for children sure to physical and non-physical violent assaults,
although the other two offer pediatric services when respectively, with 9 and 8 items derived from a review
necessary. All three hospitals have less than ten psychi- of related literature (Boafo & Hancock 2017; d’Ettorre
atrists who are supported by medical doctors (non- & Pellicani 2017; ILO/WHO/ICN/PSI 2003). Respon-
specialist in psychiatry), medical assistants, nurses, clin- dents were asked to indicate the extent to which staff
ical psychologists, occupational therapists, and other of the hospital in which they work are subjected to
essential cadre (Roberts et al. 2014). Across all the hos- physical harm by patients; violent abuse by patients;
pitals, inadequate operational resources including forced orders by patients; experience of sexual harass-
essential medical equipment and supplies tend to ment; and verbal abuse and threats of harm by
undermine the delivery of optimal clinical services. patients. Some examples of such questions were as fol-
Patients mostly present to the hospitals for OPD and lows: ‘Do staff of this hospital experience physical
inpatient care with conditions relating to mood disor- abuses by patients?’; ‘Have you ever witnessed an inci-
ders, stress and neurotic disorders, schizophrenia, sub- dent where a staff is being attacked by a patient?’; and
stance dependence, and mental retardation among ‘Would you say that staff of this hospital suffer from
others (Roberts et al. 2014). patients verbal abuses?’.
Based on a scrutiny of earlier literature, we hypothe- Part three used 7 items teased out from the works
sized that more than 50% of our sample will likely of Lanctôt and Guay (2014) and Pelto-Piri et al. (2019)
report experiencing some form of physical or non- to measure perceptions about safety of the hospitals.
physical violent assaults which in turn will shape per- Such questions were framed like ‘I have concerns
ceptions that the care environment is precarious. If this about the safety of this hospital’; ‘My coworkers com-
assumption was confirmed, then there was the need to plain about safety of this hospital’; and ‘Safety of the
© 2021 John Wiley & Sons Australia, Ltd
VIOLENT ASSAULTS IN PSYCHIATRIC HOSPITALS 1345
hospitals should be improved’. These questions were the independent variables (age, profession, job tenure,
placed on a five-point Likert agreement/disagreement unit of practice) with experience of physical violent
scale. The final part asked each respondent to indicate assault (model 1); experience of non-physical violent
whether within the last 12 months preceding this assault (model 2); and perceived safety of the hospital
study, they personally encountered physical violent (model 3). In models 1 and 2, we controlled for routine
assault and non-physical violent assault by a patient, contact with patients since it potentially confounds the
and the number of times they experienced these results. Coefficients (Coef.) and odds ratios (OR) of the
assaults. Respondents were also asked whether in the regression models are reported.
last 12 months, the work environment has been safe or
unsafe to provide care and a definite question: ‘Would Qualitative process
you say that it is risky working in this hospital?’ Analysis of the quantitative data revealed that about
Responses to these last set of questions were dichoto- 57% and 71% of the respondents reported ever experi-
mized as 1 = ‘yes’ if a respondent confirms the ques- encing some form of patient physical and non-physical
tion being asked and 0 = ‘no’ if otherwise. violent assault, respectively, while 80% perceived that
The questionnaire was pretested among six health they work in a risky environment. These respondents
workers for feedback and revised before being used for from the different professional cadre and across
data collection. The final questionnaire was self- departments/units were then recruited into the qualita-
administered to respondents to complete and submit to tive study to gain in-depth knowledge on the nature,
the data collectors assigned to the hospitals. However, perceived causes, and consequences of the violent
some respondents cited busy routines as an obstacle to assaults and why the mental healthcare environment
on-the-spot collection of the completed questionnaire. was perceived as unsafe. They were contacted via
For such respondents, the questionnaire was delivered phone calls, mails, and meetings at the hospitals and
in a sealed envelope for them to complete and return asked whether they would participate in a second
later. Respondents unable to complete the question- round of study to share experiences of the violent
naire within reasonable time were given text messaging assaults they encountered. Of the 195 successful con-
reminders once every 2 days. Respondents completed tacts initiated, 35 comprising some psychiatric nurses,
and delivered the questionnaire in an average of 5 health assistants, pharmacists, and administrative staff
working days. Of the 562 questionnaires administered, agreed to participate in the qualitative study. Busy rou-
501 (representing a response rate of 89.1%.) were fully tines at work, leave days, and lack of interest were
completed and returned (AnPH: 170; APH 162 and some of the reasons for which most of the contacted
PPH 169). participants did not participate.
Data were analyzed with the aid of SPSS version 24. Data were collected using an interview guide cover-
We started the analysis by using principal component ing topics such as causes, nature, and consequences of
analysis (PCA) with varimax rotation to condense the violent assaults, and perception of safety, unsafety,
items of the instrument into a few linearly uncorrelated security, and protection of staff against patient vio-
variables. In other words, we sought to reduce the lence. Face-to-face in-depth interviews were held with
large set of items of the questionnaire into a distinct participants outside working hours at places convenient
few by employing PCA. All the items were loaded into for them. Two graduate psychology students and the
the PCA, but 5, 3, and 2 items, respectively, of physical first author conducted the interviews which were
violent assault, non-physical violent assault, and safety recorded and lasted for an average of 30 minutes. All
constructs were poorly loaded and accordingly omitted. the participants who agreed to participate completed
The remaining items were strongly loaded with Cron- the interview process. The qualitative data were tran-
bach alpha (a) values above the recommended 0.75 scribed, cleaned, and analyzed for themes. Topics of
threshold (Field 2009) as follows: physical violent the interview guide were used as initial set of broad
assault (four items; a = 0.78), non-physical violent themes to guide coding. Sub-themes were subsequently
assault (5 items; a = 0.79), and safety of the hospital derived and appropriately nested into the broad themes
(five items; a = 0.81). by reading through the transcripts and assigning codes
Demographic characteristics of respondents were (Hsieh & Shannon 2005). After coding every eight suc-
analyzed descriptively. To determine the degree of cessive transcript, a qualitative researcher cross-
exposure to violent assaults, three multivariate regres- checked the codes for validity and their alignment with
sion models were computed to examine association of the themes. This was done until all the transcripts were
© 2021 John Wiley & Sons Australia, Ltd
1346 R. A. ATINGA ET AL.
coded. Qualitative data about the nature of physical RESULTS
and non-physical violent assaults, and injuries sustained
from the former were quantified using Microsoft Excel. Demographic characteristics
Because of the multiple mentions obtained, the sum of
Participants’ characteristics are shown in Table 1. Par-
the frequencies exceeded 100%.
ticipants in the qualitative study were slightly older
(Mean = 33.0 years) than the quantitative participants
Ethical approval (Mean = 29.9 years). Most of the participants for both
Ethical approval was obtained from the Ghana Health studies were females, nurses, and non-clinical staff.
Service. Participants were made aware that their partic- Results from Table 2 show that more females
ipation in the study was voluntary and that they could (58.6%) than males (41.4%) reported experiencing
withdraw from the study at any time or refuse to physical violent assaults (P < 0.01). Only about 20% of
answer any question during the interview process. To the respondents described the work environment as
ensure privacy and confidentiality, interviews were con- safe which also accounted for less than a quarter
ducted at locations devoid of any influence or interfer- (19.1%) of females being satisfied with safety of the
ence of other staff. The data were subsequently workplace. Overall, few males (23.9%) and females
passworded. (17.0%) perceived the workplace to be safe (P < 0.01)
TABLE 1 Respondents’ characteristics
Quantitative study (n = 501) Qualitative study (n = 34)
Characteristic Description n (%) n (%)
Sex Male 213 (42.5) 17 (48.6)
Female 288 (57.5) 18 (51.4)
Age (years) Mean (std. dev.) 29.9 (30.2) 33.0 (27.1)
Range 20–56 23–42
Job tenure (years) Mean (std. dev.) 9.0 (4.2) 5.1 (8.2)
Range 1–29 3–27
Education Secondary 132 (26.3) 6 (17.1)
Nursing/health training 216 (43.1) 20 (57.2)
Tertiary 153 (30.5) 9 (25.7)
Staffing category Psychiatric nurse 86 (17.2) 6 (17.1)
Other nursea 130 (26.0) 8 (22.9)
Other clinical cadreb 92 (18.4) 6 (17.1)
Non-clinical staffc 193 (38.4) 15 (42.9)
aRegistered nurse, enrolled nurse, community health nurse, and public health nurse.
bHealth assistant, pharmacist, pharmacy technician, mental health officer.
cCaterers, administrative managers and assistants, laundry staff, records managers, technicians.
TABLE 2 Descriptive statistics of violent assaults and safety of the workplace by gender
Male (n = 213) Female (n = 288)
Yes Yes
Variable Freq. % Freq. % P-value
Ever been physically assaulted by a patient 118 41.4 167 58.6 0.001
Ever been verbally assaulted by a patient 140 65.7 215 74.7 0.000
Ever been threatened with harm by a patient 171 80.3 250 86.8 0.002
Ever been sexually harassed 67 31.5 201 69.8 0.034
The work environment is safe 51 23.9 49 17.0 0.005
Satisfied with safety of the workplace 59 27.7 55 19.1 0.024
© 2021 John Wiley & Sons Australia, Ltd
VIOLENT ASSAULTS IN PSYCHIATRIC HOSPITALS 1347
which suggests that about 80% of the respondents per- between 1 and 5 years (OR: 3.493: P < 0.01). Respon-
ceived the hospital work environment as risky or dents aged 51 years and above and those with many
unsafe. About 59% and 63% of the participants years of work experience were less likely to perceive
reported experiencing physical and non-physical violent the clinical setting as unsafe.
assaults, respectively, ranging between 1 and five times
in the last 12 months (Fig. 1).
Nature and consequences of violent assaults
Participants revealed experiencing various forms of
Experience of physical and non-physical violent
physical and/or non-physical violent assaults. The most
assaults and risk perception
dominant forms of physical assaults experienced
Table 3 presents regression results of respondents’ included slapping (27%), punching (22%), hitting with
experience of physical and non-physical violent assaults an object (17%), and kicking (20%). These assaults pro-
and perceived risk of the hospital environment. Com- duced injuries like swelling, laceration, loss of sight,
pared with males, females were significantly more likely body pains, and emotional distress. Some of the non-
to experience physical violent assaults (OR = 2.957; physical violent assaults encountered were sexual
P < 0.01) and non-physical violent assaults harassment (11%), spitting on (14%), and threat to
(OR = 1.610; P < 0.01). All but non-clinical staff were harm (29%). Detail descriptions of the nature of the
statistically and significantly more likely to experience physical and non-physical violent assaults experienced
both physical and non-physical violent assaults. How- and their consequences, and quotes illustrating key
ever, compared to the other professional cadre, nurses experiences are shown in Table 4.
highly experienced physical violent assaults
(OR = 3.704; P < 0.01) and non-physical assaults
Perceived causes of violent assaults
(OR = 3.861; P < 0.01). Having many years of work
experience in the hospitals reduces the chances of The causes of violent assaults were attributed to several
experiencing physical violent assault. factors, notable among them were poor stock of essen-
Respondents’ experience of both types of assaults tial medicines, health workers’ approach to patient care
varied by departments. Staff working in the OPD and and management, interactions with patients, and staff
inpatient wards significantly experienced both types of relations and attitudes to patients. These are further
violent assaults compared to those in other clinical explained below.
departments who encounter only non-physical violent
assault. Although all staff category perceived the hospi- Poor stock of essential medicines
tals environment as unsafe, it was statistically and sig- Antipsychotic medications for managing psychotic dis-
nificantly high for nurses (OR = 4.396; P < 0.01), other orders were reported to run out frequently, which
clinical staff (OR = 2.983; P < 0.01), those aged 31- sometimes constrain sedation efforts of providers. This
40 years (OR: 3.395; P < 0.01), and whose tenure is was further compounded by the free admission system
and medications subsidy that contributed to rapid
70.0
62.8 depletion of essential drug supplies. As a result, some
58.8
60.0 staff complained about experiencing violence escalation
whenever medicines run out.
50.0
40.0 Patients become violent and attack us when medica-
tions run out, because during such times controlling or
30.0 25.3 23.5 calming down aggressive patients becomes difficult
(Nurse, APH)
20.0
11.4 Aggressive patients become too violent and attack staff10.5
10.0 5.4 when there are no drugs. (Nurse, PPH)
2.3
0.0
1 to 5 6 to 10 11 to 15 16 plus 1 to 5 6 to 10 11 to 15 16 plus
Approach to patient care and management
physical violent assault non-physical violent assault
Violent assaults were found to be associated with how
FIG. 1 Number of times respondents experienced physical and health providers managed and cared for patients. Using
non-physical violent assaults within the last 12 months to the study. some form of restraint to administer treatment,
© 2021 John Wiley & Sons Australia, Ltd
Percent 
1348 R. A. ATINGA ET AL.
© 2021 John Wiley & Sons Australia, Ltd
TABLE 3 Regression results of the experience of physical and non-physical violent assaults and perceived risk of the hospital environment
Physical violent assault Non-physical violent assault Perceived risk of the hospital environment
Variable Coef. OR CI (95%) Coef. OR CI (95%) Coef. OR CI (95%)
Gender (ref. male)
Female 1.144 2.957** 0.624–3.470 0.494 1.610** 0.391–1.951 0.274 1.315** 0.392–2.918
Age (ref 20–30 years)
31–40 1.683 5.382** 1.951–7.460 1.333 3.793* 1.613–5.478 1.222 3.395** 0.204–5.887
41–50 1.031 2.803 0.481–3.332 1.601 4.959* 1.778–9.595 0.368 1.444* 0.171–2.825
51+ 
0.915 1.498 0.480–2.011 1.126 3.082* 0.543–5.503 
0.455 1.576 0.119–1.971
Profession (ref. Psychiatric nurse)
Other nursea 1.351 3.704** 1.274–8.809 1.351 3.861** 1.410–10.569 1.926 4.396** 1.134–8.501
Other clinical cadreb 1.045 2.844** 1.025–5.893 0.911 2.826** 1.297–7.303 1.017 2.983** 0.737–4.980
Non-clinical staffc 
1.264 0.538 0.268–1.659 
0.486 0.615 0.414–2.608 0.619 1.844* 0.162–2.497
Job tenure (ref. < 1 year)
1–5 0.546 1.782* 0.145–4.217 1.969 3.040* 0.023–3.954 1.706 3.493** 1.051–3.719
6–10 
0.477 1.860* 0.093–2.280 1.183 2.306* 0.155 
2.706 1.311 2.269** 0.150–2.624
11+ 
0.689 0.902 0.098–2.561 1.665 0.189* 0.033–1.083 1.333 2.717 0.474–3.318
Unit (ref. inpatient)
OPD 1.803 2.233* 0.976–5.107 1.194 3.823* 0.305–4.621 1.732 3.481** 1.332–5.586
Inpatient ward 1.955 4.066* 2.452–9.364 1.429 5.651* 3.211–11.007 1.781 3.458** 0.125–3.908
Other clinical unitd 1.413 3.110 0.192–5.552 1.040 2.644* 0.109–6.845 1.663 2.940* 0.787–3.336
Support unite 1.223 1.238 0.910–5.805 0.018 1.889 0.135–2.861 0.747 1.474 0.542–1.686
v2(25) 107.82 82.76 55.29
Prob > v2 0.000 0.000 0.000
Pseudo R2 0.260 0.214 0.152
Log likelihood 578.774 541.813 530.582
aRegistered nurse, enroll nurse, community health nurse, and public health nurse.
bHealth assistant, pharmacist, pharmacy technician, mental health officer.
cCaterers, administrative managers, laundry staff, records managers, technicians.
dOccupational therapy, community psychiatry, rehabilitation, pharmacy.
eCatering, administration, laundry, records, and technical units.
**, *significant at 1% and 5%, respectively.
VIOLENT ASSAULTS IN PSYCHIATRIC HOSPITALS 1349
TABLE 4 Physical and non-physical violence experienced and the consequences
Nature of physi- Nature of non-
cal violent Per physical violent Per Selected quotes of non-
assault cent Selected quotes of physical assaults Consequence assaults cent physical assault
Slapped 26 ‘Slapping is common. I have been Laceration Screamed at 42 ‘Patient yelled at me for
slapped countless times’ wound looking at her’
Swelling on ‘Patient kept shouting at me’
the face
Punched 22 ‘Patient punched me at the back and Swelling at Insults 31 ‘some of the patients keep
when I turned, he slapped me again’ the back insulting me’
‘I have been receiving insults’
Kicked at the 20 ‘I was about leaving the ward when he Body pains Threat to harm 29 ‘Patient threatened that he
back kicked me to the ground’ Emotional doesn’t want to see me again’
distress ‘Patient threatened to beat
me anytime she sees seen’
Smashed with 19 ‘The patient smashed my face with Swelling on Spitting on 14 ‘Patient spat on my face’
food and/or food served’ the face
water ‘Water was smashed all over my body’ Temporary
loss of sight
Hit with an 17 ‘Patient threw a bench at me’ Scratch on Sexual harassment 11 ‘patient grabbed my breast
object ‘A stone was pelted at my back’ the body and fondled my buttocks’
‘Patient hit me with dustbin’ Shoulder
subluxation
Bone sprain
Chased with 13 ‘Patient chased me with a cutlass’ Body pains Death threat 9 ‘I have been receiving death
stabbing weapon ‘She chased me with a knife to kill’ from falls threats from a patient’
‘Patient always threaten to
kill me’
Strangled on the 10 ‘The patient pounced on me and tightly Broken Forced kissing 6 ‘Patient forcefully kissed me’
neck held my neck until I was rescued’ mouth ‘Patient trailed and kissed me
‘I was strangled one evening by a Traumatized at the neck’
patient’ Anxiety disor-
ders
Attempted rape 5 ‘He trailed me from the back, violently None
tore my dress in an attempt to rape’ reported
medication, or persuade a patient to respond to treat- and attitude to patients. Behaviors considered prob-
ment plan could trigger aggression and assaults. In lematic rather than therapeutic provoked aggressive
addition, verbal and non-verbal interaction such as cues outburst and assaults. A participant remarked that
were also reported as sources of assaults especially by ‘patients do not want to be seen or referred to as
patients with paranoid conditions. Similarly, unneces- mad, they are humans, they want to be respected’.
sary requests and denial of patients’ preferred care For this reason, staff could be assaulted if they
options prompted assaultive behaviors as noted by this exhibited behavior considered by a patient as disre-
participant. spectful, degrading, dehumanizing, or undermining
personal values. Moreover, despite awareness by pro-
He said he wanted a certain medication which he says
viders that each patient had a pathology that con-
was good for him, but I refused. As I was about leav-
ing, he trailed and hit me at the back. I had to run for tributes to deviation from the straight, some health
my life. (Psychiatric nurse, APH) workers still related to or used terms considered
offensive to patients. Such behaviors were flashpoints
Staff relations and attitude to patients for violent assault, as for example:
Participants explained that violent assaults are some- Staff have been told to behave professionally towards
times triggered by the nature of staff relations with, these patients. But some take things for granted by
© 2021 John Wiley & Sons Australia, Ltd
1350 R. A. ATINGA ET AL.
calling patients all sort of names. When you do that the respectively, are on the high side compared with rates
patient will attack you. (Administrative staff, AnPH) reported in other countries (Olashore et al. 2018;
Ukpong et al. 2011). The nature of physical and non-
Perceived risk of the workplace physical violent assaults discovered such as hitting with
an object, name calling, sexual harassment, spitting on,
Participants raised concerns about safety of the hospi- smashing with food and strangling, builds on, and com-
tals. They felt that although health providers in psychi- plements previous studies within mental health care con-
atric hospitals are more prone to violent assaults than text (d’Ettorre & Pellicani 2017; McKinnon & Cross
other clinical environments, limited consideration has 2008; Olashore et al. 2018). Major and minor injuries
been given to building a sustainable working environ- and psychosocial issues were frequent sequelae following
ment that support safe clinical practice and patient physical violent assault. These consequences combined
care. with the fear of working in an unsafe environment scaled
We suffer more at the hands of patients, that is the back health providers’ therapeutic efforts. This is seen as
true fact. But there is no good protection of staff. a threat to ongoing structural reforms aimed at creating
(Nurse, APH) patient-centred mental health care as well as attraction
‘Safe environment? I don’t think so. We are not safe and retention of critical skill cadre in Ghana’s referral
enough to do our work. (Health Assistant, PPH) psychiatric hospitals (Jack et al. 2013).
Frequent stock outs of essential medications and
There were others who described the practice envi-
health provider behavioral elements were common
ronment as risky because the occurrences of violent
antecedent triggers of patient violence. Of these, stock
assaults are unpredictable which makes it difficult to
outs of essential medications is the most disturbing
anticipate and plan mitigation strategies against them.
since biomedical treatment of mental disorders is phar-
A participant lamented that no matter how careful one
macotherapy dependent. Medication is preferred to
might be, the possibility of being assaulted is still high
coercive restraint in any form (Tishler et al. 2013).
due to unsafe and insecure environment for patient
Administering psychotropic medication, for example,
care.
has proven to be effective in mood stabilization and
Today may be calm, no violence, nothing, but tomor- minimal violence risk in patients with mental disorders
row can be something else where you have to run for (Fazel et al. 2014). Similarly, medication reduces unde-
your life. (Nurse, AnPH) sirable feeling of anguish and aggression in patients
(Xavier et al. 2014). In the hospitals studied, however,
Some participants were of the view that mental dis-
unavailability of needed medicines at all times often
orders are on the rise in Ghana, and this puts pressure
resulted in the use of force restraints and seclusion to
on resources, infrastructure, and health professionals in
control aggression. Despite the inclusion of a range of
referral facilities. Seclusion rooms, for example, were
psychotropic medications of varying therapeutic class to
inadequate relative to the number of patients with
the essential drug lists, most facilities are not consistent
aggressive behaviors, thus putting everyone at risk.
with stock of newer medications such as Olanzapine
Moreover, the fact that staff routinely experienced
for sedation (Roberts et al. 2014). This limits efforts by
some form of violent assaults was noted as a reason for
providers to effectively control aggressive patients.
which some participants described the workplace as
The behavioral precursors of violent assaults espe-
unsafe. This sarcastic statement by one of the partici-
cially those about stigmatizing remarks and rudeness
pants is worth highlighting:
have also been reported elsewhere (Hallett et al. 2014).
We are safe, why not. But some of us suffer at the Behavioral risk factors of assaults do arise because
hands of patients. Will you say we are safe? (Psychiatric patients with mental disorders also expect to be recog-
nurse, APH) nized and accorded with dignity, good therapeutic rela-
tions, and valued as humans (Gilburt et al. 2008).
DISCUSSION Thus, behaviors perceived as dehumanizing, downgrad-
ing, and disrespectful may provoke patient violence.
Both the quantitative and qualitative data revealed that The findings showed that almost every staff espe-
mental healthcare workers are exposed to a range of cially the clinical cadre experienced violent assaults in
physical and non-physical violent assaults. The physical one form or another. The descriptive statistics
and non-physical violent assault rates of 57% and 71%, (Table 2) showed that compared to males, females
© 2021 John Wiley & Sons Australia, Ltd
VIOLENT ASSAULTS IN PSYCHIATRIC HOSPITALS 1351
were more at risk of both assault types. This is evi- Across the hospitals, safety of the practice environ-
denced by the high proportions of females than males ment was compromised by the prevalence of patient
reporting physical and non-physical abuses. A possible violence (Blando et al. 2013). The perception of work-
explanation of this is variation in the degree of control ing in risky a environment and the feeling that individ-
and power that health providers have and exercise over uals are responsible for their own safety contrast with
patients when administering care (Oram et al. 2017). what pertains in other settings (Pelto-Piri et al. 2019;
This control tends to be high for males than females Privitera et al. 2005), where patients and mental health
who by their biological makeup are vulnerable to care providers alike felt being safe within psychiatric
assaults at many levels of social organizational struc- care delivery spaces. Nonetheless, poor safety of men-
tures (Islam et al. 2018). As women encountered more tal health care workers has been a global problem. A
violent assaults than men, it was not surprising that the review of violence and risk of psychiatric hospitals
practice environment was perceived as unsafe by a vast across countries highlighted insecurity and poor safety
majority of the female respondents. of mental health workers as a concern (d’Ettorre &
Corroborating earlier evidence in mental health care Pellicani 2017). Poor safety of mental health workers
(AbuAlRub et al. 2014; Baby et al. 2014; Spector et al. raises questions around the effectiveness of legal and
2014; Stevenson et al. 2015), the experiences of both policy instruments providing structures for care deliv-
physical and non-physical violent assaults were statisti- ery. In the Ghanaian context, for example, the mental
cally and significantly high among nurses. This finding health law (Act 846) guarantees mental patients in con-
could be attributed to nurses being mainly at the front- tact with facilities the right to the highest attainable
line of care including administering medication, coun- standard of mental health care (Government of Ghana
selling, restraining aggressive patients, and responding 2012). However, achieving this appears a distant goal if
to patient care demands. Scarcity of psychiatrists in the mental health providers are not equally guaranteed
hospitals further overburdened nurses with multiple stable and congenial practice spaces that supports safe
patient orientated roles. Navigating through routine care.
clinical complexities to improve outcomes probably
puts nurses more at risk of patient assaultive behaviors.
CONCLUSION
Nonetheless, the fact that physical and non-physical
assaults were statistically and significantly associated Findings of the study highlight the forms, causes, and
with other clinicians suggests widespread assaults consequences of patient violent assaults and the health
against clinicians in general (d’Ettorre & Pellicani workers more at risk of these assaults in Ghana’s psy-
2017). Non-clinical staff were not exposed to physical chiatric hospitals. The experience of violent assaults
violent assaults contrary to the literature (Anderson & was not only associated with clinicians because of their
West 2011), probably because of their distant role proximal role in patient care and management, but also
around patient care and management. among women and those with low tenure and age. This
There were statistically significant associations of age demonstrates the widespread nature of patient violent
and job tenure with the experiences of both physical assaults. Violent assaults occurrences were somewhat
and non-physical violent assaults (Antonius et al. 2010). linked to resource gaps pertaining to medication and
However, the negative regression coefficients indicate essential supplies that undermined effective clinical
that high job tenure and being older reduces the likeli- control of patients. It is not surprising, therefore, that
hood of experiencing physical violent assaults as most participants perceived the hospitals environments
opposed to related studies (Privitera et al. 2005; Soares to be unsafe for optimal health care delivery. Gener-
et al. 2000). Possibly longer service and increasing age ally, mechanisms of containing the forms of violent
increases awareness of, and the skill to predict and assaults are crucial as are strategies to improve safety,
control or avoid imminent violent behavior. In addition, resources, and confidence of health workers delivering
the ability to engage with patients using logical reason- care in psychiatric hospitals.
ing and more friendly discussion tactics have been
shown to be high among health providers who are
RELEVANCE FOR CLINICAL PRACTICE
older and with high job tenure (Hallett et al. 2014).
We think that such experiences might have helped the The findings suggest the need for timely supply and
more experienced health workers to better manage availability of essential medications in psychiatric hospi-
patients with aggressive behaviors. tals. Uninterrupted stocking of psychotropic medication
© 2021 John Wiley & Sons Australia, Ltd
1352 R. A. ATINGA ET AL.
and mood stabilizers, for example, enables clinicians to programmes and workplace aggression on nurse
manage high aggressive patients with schizophrenia, perceptions of safety. Journal of Nursing Management, 21
substance abuse, and dependence. The findings showed (3), 491–498.
that most health workers, especially the clinical cadre Boafo, I.M. & Hancock, P. (2017). Workplace violence
against nurses: A cross-sectional descriptive study of
and women, are more exposed to violent assaults. Ghanaian nurses. Sage Open, 7 (1), 1–9.
Accordingly, we recommend that they should be (re) Bruckner, T.A., Scheffler, R.M., Shen, G. et al. (2011). The
oriented on non-drug approaches to managing patients mental health workforce gap in low-and middle-income
with mental disorders. Such approaches include but countries: A needs-based approach. Bulletin of the World
not limited to how to engage patients with good com- Health Organization, 89, 184–194.
munication, well-mannered interaction, friendly postur- Camerino, D., Estryn-Behar, M., Conway, P.M., van Der
ing, calm demeanor, and maintaining a safe distance Heijden, B.I.J.M. & Hasselhorn, H.-M. (2008). Work-
related factors and violence among nursing staff in the
from aggressors. Building structures that ensure a safe
European NEXT study: A longitudinal cohort study.
and supportive working environment for mental health International Journal of Nursing Studies, 45 (1), 35–50.
workers should focus the attention of health decision Cashmore, A.W., Indig, D., Hampton, S.E., Hegney, D.G. &
makers. For instance, designing occupational health Jalaludin, B.B. (2012). Workplace violence in a large
and safety protocols with safety matrix for assessing risk correctional health service in New South Wales, Australia:
levels can greatly inform appropriate management deci- A retrospective review of incident management records.
sions and actions. This should be complemented with BMC Health Services Research, 12, 245.
Chen, W.-C., Hwu, H.-G., Lin, Y.-P. et al. (2010). Workplace
the provision of adequate security for dealing with inci-
violence from psychiatric patients. Journal of Occupational
dence of physical violent assaults. Safety and Health, 18, 163–176.
Creswell, J.W. & Clark, V.L.P. (2011). Designing and
ACKNOWLEDGEMENT Conducting Mixed Methods Research. , 2nd edn. New
York, NY: SAGE Publcations Inc.
We are grateful to all the staff of the hospitals who d’Ettorre, G. & Pellicani, V. (2017). Workplace violence
took time off their busy schedules to participate in this toward mental healthcare workers employed in psychiatric
study. We are also indebted to the data collectors and wards. Safety and Health at Work, 8 (4), 337–342.
Dhumad, S., Wijeratne, A. & Treasaden, I. (2007). Violence
all those who read and shaped the quality of this against psychiatrists by patients: Survey in a London
paper. mental health trust. Psychiatric Bulletin, 31 (10), 371–374.
El Ghaziri, M., Zhu, S., Lipscomb, J. & Smith, B.A. (2014).
Work Schedule and Client Characteristics Associated With
REFERENCES Workplace Violence Experience Among Nurses and
AbuAlRub, R.F., Khawaldeh, A. & Talal, A. (2014). Midwives in Sub-Saharan Africa. Journal of the Association
Workplace physical violence among hospital nurses and of Nurses in AIDS Care, 25(Suppl. 1), S79–S89.
physicians in underserved areas in Jordan. Journal of Fazel, S., Zetterqvist, J., Larsson, H., Langstr€om, N. &
Clinical Nursing, 23 (13–14), 1937–1947. Lichtenstein, P. (2014). Antipsychotics, mood stabilisers,
Alhassan, R.K. & Poku, K.A. (2018). Experiences of frontline and risk of violent crime. The Lancet, 384 (9949), 1206–
nursing staff on workplace safety and occupational health 1214.
hazards in two psychiatric hospitals in Ghana. BMC Public Field, A. (2009). Discovering Statistics Using SPSS, vol 3.
Health, 18 (1), 1–12. London: Sage.
Anderson, A. & West, S.G. (2011). Violence against mental Fournier, O.A. (2011). The status of mental health care in
health professionals: When the treater becomes the Ghana, West Africa and signs of progress in the greater
victim. Innov Clin Neurosci, 8 (3), 34–39. Accra region. Berkeley Undergraduate Journal, 24 (3), 8–
Antonius, D., Fuchs, L., Herbert, F. et al. (2010). Psychiatric 34.
assessment of aggressive patients: A violent attack on a Fricchione, G.L., Borba, C.P.C., Alem, A., Shibre, T.,
resident. American Journal of Psychiatry, 167 (3), 253– Carney, J.R. & Henderson, D.C. (2012). Capacity building
259. in global mental health: Professional training. Harvard
Baby, M., Glue, P. & Carlyle, D. (2014). ‘Violence is not part Review of Psychiatry, 20 (1), 47–57.
of our job’: A thematic analysis of psychiatric mental Gilburt, H., Rose, D. & Slade, M. (2008). The importance of
health nurses’ experiences of patient assaults from a New relationships in mental health care: A qualitative study of
Zealand Perspective. Issues in Mental Health Nursing, 35 service users’ experiences of psychiatric hospital admission
(9), 647–655. in the UK. BMC Health Services Research, 8 (1), 92.
Blando, J.D., O’Hagan, E., Casteel, C., Nocera, M.-A. & Government of Ghana (2012). Mental Health Act, 2012.
Peek-Asa, C. (2013). Impact of hospital security Accra: Assembly Press.
© 2021 John Wiley & Sons Australia, Ltd
VIOLENT ASSAULTS IN PSYCHIATRIC HOSPITALS 1353
Hallett, N., Huber, J.W. & Dickens, G.L. (2014). Violence inpatients in Sweden. International Journal of Mental
prevention in inpatient psychiatric settings: Systematic Health Systems, 13 (1), 23.
review of studies about the perceptions of care staff and Privitera, M., Weisman, R., Cerulli, C., Tu, X. & Groman, A.
patients. Aggression and Violent Behavior, 19 (5), 502– (2005). Violence toward mental health staff and safety in
514. the work environment. Occupational Medicine, 55 (6),
Hsieh, H.-F. & Shannon, S.E. (2005). Three approaches to 480–486.
qualitative content analysis. Qualitative Health Research, Roberts, M., Mogan, C. & Asare, J.B. (2014). An overview of
15 (9), 1277–1288. Ghana’s mental health system: Results from an assessment
ILO/WHO/ICN/PSI. (2003). Workplace Violence in the using the World Health Organization’s Assessment
Health Sector Country Case Studies Research Instruments: Instrument for Mental Health Systems (WHO-AIMS).
Survey Questionnaire. Geneva: International Labor International Journal of Mental Health Systems, 8 (1), 16.
Organization. Roche, M., Diers, D., Duffield, C. & Catling-Paull, C.
Islam, M.M., Jahan, N. & Hossain, M.D. (2018). Violence (2010). Violence toward nurses, the work environment,
against women and mental disorder: A qualitative study in and patient outcomes. Journal of Nursing Scholarship, 42
Bangladesh. Tropical Medicine and Health, 46 (1), 5. (1), 13–22.
Jack, H., Canavan, M., Ofori-Atta, A., Taylor, L. & Bradley, Rueve, M.E. & Welton, R.S. (2008). Violence and mental
E. (2013). Recruitment and retention of mental health illness. Psychiatry, 5 (5), 34.
workers in Ghana. PLoS One, 8 (2), e57940. Shiao, J.-C., Tseng, Y., Hsieh, Y.-T., Hou, J.-Y., Cheng, Y. &
Kindy, D., Petersen, S. & Parkhurst, D. (2005). Perilous Guo, Y.L. (2010). Assaults against nurses of general and
work: Nurses’ experiences in psychiatric units with high psychiatric hospitals in Taiwan. International Archives of
risks of assault. Archives of Psychiatric Nursing, 19 (4), Occupational and Environmental Health, 83 (7), 823–832.
169–175. Soares, J.J., Lawoko, S. & Nolan, P. (2000). The nature,
Kisa, S. (2008). Turkish nurses’ experiences of verbal abuse extent and determinants of violence against psychiatric
at work. Archives of Psychiatric Nursing, 22 (4), 200–207. personnel. Work & Stress, 14 (2), 105–120.
Lanctôt, N. & Guay, S. (2014). The aftermath of workplace Spector, P.E., Zhou, Z.E. & Che, X.X. (2014). Nurse
violence among healthcare workers: A systematic literature exposure to physical and nonphysical violence, bullying,
review of the consequences. Aggression and Violent and sexual harassment: A quantitative review.
Behavior, 19 (5), 492–501. International Journal of nursing Studies, 51 (1), 72–84.
Magnavita, N. & Heponiemi, T. (2011). Workplace violence Stevenson, K.N., Jack, S.M., O’Mara, L. & LeGris, J. (2015).
against nursing students and nurses: An Italian experience. Registered nurses’ experiences of patient violence on
Journal of Nursing Scholarship, 43 (2), 203–210. acute care psychiatric inpatient units: An interpretive
McKinnon, B. & Cross, W. (2008). Occupational violence descriptive study. BMC Nursing, 14 (1), 35.
and assault in mental health nursing: A scoping project for Tishler, C.L., Reiss, N.S. & Dundas, J. (2013). The
a Victorian Mental Health Service. International Journal assessment and management of the violent patient in
of Mental Health Nursing, 17 (1), 9–17. critical hospital settings. General Hospital Psychiatry, 35
Niu, S.-F., Kuo, S.-F., Tsai, H.-T., Kao, C.-C., Traynor, V. & (2), 181–185.
Chou, K.-R. (2019). Prevalence of workplace violent Ukpong, D., Abasiubong, F., Ekpo, A., Udofia, O. & Owoeye,
episodes experienced by nurses in acute psychiatric O. (2011). Violence against mental health staff in Nigeria:
settings. PLoS One, 14 (1), e0211183. Some lessons from two mental hospitals. Nigerian Journal
Olashore, A.A., Akanni, O.O. & Ogundipe, R.M. (2018). of Psychiatry, 1, 14–17.
Physical violence against health staff by mentally ill WHO (2002). The World Report on Violence and Health:
patients at a psychiatric hospital in Botswana. BMC Summary. Geneva: World Health Organization.
Health Services Research, 18 (1), 362. WHO (2019). Mental Health Human Resources Data by
Oram, S., Khalifeh, H. & Howard, L.M. (2017). Violence Country. Geneva: WHO.
against women and mental health. The Lancet Psychiatry, Xavier, M.D.S., Terra, M.G., Silva, C.T.D., Mostardeiro,
4 (2), 159–170. S.C.T.D.S., Silva, A.A.D. & Freitas, F.F.D. (2014). The
Pelto-Piri, V., Wallsten, T., Hylen, U., Nikban, I. & Kjellin, meaning of psychotropic drug use for individuals with
L. (2019). Feeling safe or unsafe in psychiatric inpatient mental disorders in outpatient monitoring. Escola Anna
care, a hospital-based qualitative interview study with Nery, 18 (2), 323–329.
© 2021 John Wiley & Sons Australia, Ltd