Violence towards personnel in out-of-hours primary care: A cross-sectional study

Abstract

Objective. To investigate (1) the prevalence of occupational violence in out-of-hours (OOH) primary care, (2) the perceived cause of violence, and (3) the associations between occupation, gender, age, years of work, and occupational violence. Design. A cross-sectional study using a self-administered postal questionnaire. Setting. Twenty Norwegian OOH primary care centres. Subjects. Physicians, nurses, and others with patient contact at OOH primary care centres, 536 responders (75% response rate). Main outcome measures. Verbal abuse, threats, physical abuse, sexual harassment. Results. In total, 78% had been verbally abused, 44% had been exposed to threats, 13% physically abused, and 9% sexually harassed during the last 12 months. Significantly more nurses were associated with verbal abuse (OR 3.85, 95% confidence interval 2.17–6.67) after adjusting for gender, age, and years in OOH primary care. Males had a higher risk for physical abuse (OR 2.36, CI 1.11–5.05) and higher age was associated with lower risk for sexual harassment (OR 0.28, CI 0.14–0.59), when adjusted for background variables. Drug influence and mental illness were the most frequently perceived causes for the last occurring episode of physical abuse, threats, and verbal abuse. Conclusion. This first study on occupational violence in Norwegian OOH primary care found that a substantial number of health care workers experience occupational violence from patients or visitors. The employer should take action to prevent occupational violence in OOH primary care.

Key Words::

• Cross-sectional studies
• general practice
• nurses
• out-of-hours
• physicians
• prevalence
• violence

This study describes the prevalence of occupational violence among health workers in Norwegian out-of-hours primary care.

• One in three has been exposed to physical abuse during their working career in out-of-hours primary care.

• Nurses experience more verbal abuse than the other occupational groups.

• The perceived main causes of occupational violence are drug influence and mental illness.

Introduction

Threats and violence are problems both on a social level and in public health settings. The occupations doctor and nurse are associated with an especially high risk of being exposed to violence at work [1]. Occupational violence is defined as “incidents where an employee is abused, sexually harassed, or assaulted in circumstances relating to their work, involving an explicit or implicit challenge to their safety, well-being or health” [2].

The prevalence of occupational violence among health care workers varies among professionals and health care settings [3–9] but is found to be high in emergency departments [7,10,11] and among general practitioners [4,5,12–15]. Patients in emergency settings, such as out-of hours (OOH) primary care, have acute illnesses or are in acute need of help, and some patients are intoxicated with drugs or alcohol. This may produce volatile situations for the health care workers [16].

The prevalence of occupational violence and perceived causes in Norwegian OOH primary care is not known. Knowledge about both the scope and possible causes is important to prevent occupational violence. The OOH primary care service is usually managed by the regular general practitioners’ surgeries during office hours and by municipality-maintained OOH duties by general practitioners (GPs) during evenings, nights and weekends, often based in local casualty clinics with nursing professionals employed [17].

The aims of this study were to investigate (1) the prevalence of occupational violence in OOH primary care in Norway, (2) the perceived causes of occupational violence, and (3) the associations between occupation, gender, age, working years, and occupational violence.

Material and methods

A questionnaire was sent to 20 of about 220 OOH primary care centres in Norway in an attempt to obtain a representative sample with regard to different population sizes, centres located in urban and in rural areas, and centres in all parts of Norway. Power analysis based on an estimate that 30% of the persons had experienced violence showed that a total of 482 participants would give a 95% CI of 0.26–0.34 with 80% power. Study-potential OOH primary care centres were initially invited by phone to participate. The study included all nurses, physicians, and other personnel with patient contact during the study period of three weeks (January–February 2009). All eligible participants received the information letter and questionnaire through their local leaders. After completion, the anonymously answered questionnaire was returned to the National Centre for Emergency Primary Health Care.

Our questionnaire was based on an Australian questionnaire regarding occupational violence among GPs [14] and contained sections for the core study domains (verbal abuse, threats, physical abuse, and sexual harassment) with questions regarding experiences during the working career in OOH primary care and the last 12 months. Each type of aggression was defined in the questionnaire (Table I). The worker was asked whether the incident occurred in direct or indirect patient contact, frequency of the incident, and perceived cause of the last episode. The following possible causes were given: mental illness, drug influence, anxiety, grief, pain, dissatisfaction with service, and other factors. Mental illness and drug influence were combined into one category in the analysis, as many responders had ticked both. Demographic data included gender, age, occupation, and working years in OOH primary care.

Table I. Definitions of violent behaviours.

Violent behaviour	Definition
Verbal abuse	A patient, their friend(s), or family member(s) swearing or using obscene gestures with the intent of offending you. This can include abuse over the phone
Threats	A patient, their friend(s), or family member(s), through word or gesture, threatens you (including attempts to pressure you into prescribing or completing a certificate)
Physical abuse	A patient, their friend(s), or family member(s) physically attacks you. This includes behaviours such as punching, slapping, kicking, or use of a weapon or other object with the intent of intimidating you or causing bodily harm
Sexual harassment	Any form of sexual propositions or unwelcome sexual attention from patients or their friend(s) or family member(s). It includes behaviours such as humiliating or offensive jokes and remarks with sexual overtones, suggestive looks or physical gestures, requests for inappropriate physical examinations, and touching or grabbing

Source: Adapted from Magin, University of Newcastle, Australia [15].

Analyses were performed using SPSS 16.0. Continuous data are presented as mean with standard deviation (SD). Categorical variables were analysed using chi-square. Multiple logistic regression analyses were performed to investigate associations between the independent variables occupation, gender, age, and years at work in OOH primary care and each of the dependent variables verbal abuse, threats, physical abuse, and sexual harassment. Odds ratios and their confidence intervals are reported. P-values <0.05 were considered statistically significant. The study was approved by the Norwegian Social Science Data Services.

Results

A total of 716 subjects were asked to participate, and 536 subjects answered the questionnaire (response rate 75%); 70% of the respondents were women, and 62% were nurses (Table II). Other personnel (nurse aides, medical secretaries) accounted for 3%. Mean age was 42 years (SD = 11, range 21–75). The mean years of working in OOH primary care was eight (median = 7).

Table II. Description of participants (n = 536).*

	n	%
Gender
Male	158	30
Female	377	70
Profession
Nurse	331	62
Physician	190	35
Other	15	3
Age
≤ 30 years	69	13
31–49	308	58
≥ 50	155	29
Years working in OOH primary care
< 10 years	381	72
11–19	86	16
> 20	62	12

Note: ^*Due to missing data some of the total numbers are less than 536.

A total of 32% reported that they had been exposed to physical abuse during their working career in OOH primary care; mostly men (p = 0.014), but with no difference between occupations. Some 65% had been threatened, with no differences between occupations; 15% reported that they had been sexually harassed during their work in OOH primary care, comprising significantly more nurses than other employees (p = 0.016).

Altogether, 78% of the respondents reported having been verbally abused during the last 12 months (Table III). More nurses than employees in other occupations had experienced verbal abuse in both direct patient contact and by telephone in the last 12 months (p <0.001). Nine per cent of the actual nurses reported that this occurred weekly, significantly more than for other occupations (p = 0.002). Forty-four per cent had experienced threats during the last 12 months. Eighty-four per cent of the yes-responders reported that such events happened less frequently than once a month, and 9% reported that such events happened on a monthly basis. Some 13% had been exposed to physical abuse during the last 12 months. Nine per cent had been sexually harassed during the last 12 months.

Table III. Prevalence of occupational violence in last 12 months by occupation.

	Nurse (n = 331)			Physician (n = 190)			Other (n = 15)			Total (n = 536)
	n^*	%	95% CI	n	%	95% CI	n	%	95% CI	n	%	95% CI
Verbal abuse	283/329	86	82–90	121/188	64	58–71	8/14	57	31–83	412/531	78	74–81
Threats	148/318	47	41–52	72/184	39	32–46	5/15	33	10–57	225/517	44	39–48
Physical abuse	38/324	12	8–15	23/188	12	8–17	6/15	40	15–65	67/527	13	10–16
Sexual harassment	30/316	10	6–13	16/182	9	5–13	0/15	0	0–0	46/510	9	7–12

Note: ^*Number of undesired behaviours/number of respondents.

The multiple logistic regression analysis showed that physicians were less at risk of verbal abuse (OR = 0.26, CI 0.15–0.46) than nurses when adjusted for gender, age, and years of working in OOH primary care (hence OR = 3.85 for nurses) (Table IV). Men were more at risk of physical abuse (OR = 2.36, CI 1.11–5.05). Higher age was associated with less sexual harassment (OR = 0.28, CI 0.14–0.59). There were no significant associations between studied variables and threats.

Table IV. Multiple logistic regression of the association between occupation, age, gender, working years in OOH primary care, and the dependent variables verbal abuse, threats, physical abuse, and sexual harassment during the last 12 months (n = 536).

	Verbal abuse		Threats		Physical abuse		Sexual harassment
Variable	OR	95% CI	OR	95% CI	OR	95% CI	OR	95% CI
Occupation
Nurse	1.0		1.0		1.0		1.0
Physician	0.26	0.15–0.46	0.66	0.40–1.08	0.56	0.26–1.20	1.19	0.53–2.68
Gender
Female	1.0		1.0		1.0		1.0
Male	1.23	0.68–2.18	1.14	0.68–1.89	2.36	1.11–5.05	0.51	0.20–1.26
Age
<= 40 years	1.0		1.0		1.0		1.0
> 40 years	0.77	0.46–1.28	0.69	0.45–1.04	0.87	0.46–1.64	0.28	0.14–0.59
Years working in OOH primary care
<= 5 years	1.0		1.0		1.0		1.0
> 5 years	1.48	(0.90–2.43)	1.31	0.87–1.97	0.69	0.37–1.27	2.34	0.16–4.73

Notes: OR = odds ratio; CI = confidence interval.

Drug influence and mental illness were the most frequently perceived causes for the last episode in respect of the factors physical abuse, threats, and verbal abuse (Table V). Dissatisfaction with service was reported in 25% of the incidents of verbal abuse. Relatively few persons perceived pain and anxiety as causes.

Table V. Frequency of perceived causes for the last episode of verbal abuse, threats, and physical abuse.

	Verbal abuse		Threats		Physical abuse
	n	%	n	%	n	%
Drug influence or mental illness	251	61	238	77	146	91
Anxiety	47	11	21	7	3	2
Dissatisfied with service	97	25	36	12	4	3
Pain	1	0	4	1	2	1
Other	13	3	9	3	5	3
Total	409	100	308	100	160	100

Discussion

This is the first study of occupational violence in Norwegian OOH primary care. These workers have a high risk of being verbally abused, threatened, or physically abused by patients. Being a nurse implies a higher risk of verbal abuse, whereas men are more at risk of physical abuse. The perceived causes of occupational violence are drug influence and mental illness.

The strength of this study is the high response rate (75%) compared with similar studies [3–6,8,18–20]. Participating OOH primary care centres were distributed among those of different sizes, and there was a broad geographical distribution. The distribution of nurses and physicians is supposed to reflect the distribution of workers at the centres. We therefore assume that the results are representative of Norwegian OOH primary care. A limitation of the study is the possible recall bias among the respondents, which may cause a skewed distribution of answers in the direction of underreporting adverse events that happened some time ago, and also because of a perception that assaults are part of the job [21].

Verbal abuse was the most common undesired behaviour, among both nurses and physicians (78% in the last 12 months). An Australian study among GPs found a considerably lower prevalence of verbal abuse in the last 12 months [14]. However, the same study found that occupational violence in general was more likely among GPs providing OOH care. Verbal abuse seems to be part of the job among workers (physicians, nurses, assistants) in accident and emergency departments [19], a job situation comparable to OOH primary care. In Spain, 82% of the health care workers in such departments had experienced insults in the previous 12 months [7], and high numbers are also found in Ireland and the United States [11,20]. Nurses experienced more verbal abuse in both direct patient contact and by telephone. Nurses to a greater extent operate the phone and work at the reception desk in the emergency room, which possibly makes them more exposed to verbal abuse. A study among Australian emergency department nurses found that verbal violence was most often reported in the triage area [18]. Nurses are possibly also more exposed to undesired behaviour as they have more hours on duty in OOH clinics due to the organization of OOH primary care.

Threats were also a frequent occurrence in our study (44% in the last year). An Australian study among GPs found that one in four had experienced threats in the last year [14]. On the other hand, 81% of the nursing staff in emergency rooms in Ireland had experienced threatening verbal aggression in the past month [11]. However, the absence of clear and commonly shared definitions of terms to define occupational violence, including verbal abuse, threats, physical abuse, and sexual harassment, as well as cultural differences in perceived undesired behaviour, makes it difficult to quantify and compare the prevalence of such events between studies [22].

Physical abuse was experienced by one in three workers during their career in OOH primary care, and by 13% in the last year. Our study showed a higher prevalence during career than GPs in Australia, where one in five had experienced physical abuse [14], but lower than US studies among emergency physicians [20] and hospital emergency nurses [21]. More men than women in our study reported physical abuse. There is contradictory evidence concerning gender as a risk factor for being physically abused by patient or visitors. While a study among Australian GPs did not find gender differences in physical abuse [13], a study among physicians in Japan found that female employees were more often exposed [6]. However, the Japanese study included all types of specializations among physicians, and is possibly not comparable to our study of OOH primary care. In terms of fear that violence may occur, Tolhurst et al. [13] found that more female physicians experienced fear of violence being enacted against them compared with male physicians.

A fairly high proportion of the respondents in our study were exposed to sexual harassment (15% over lifetime, 9% in the last 12 months), with no gender differences. This is in contrast to other studies, which have found a higher risk of sexual harassment among women [4,8,13,14]. However, in our study there was an age difference in sexual harassment, as the workers below the age of 40 had a higher risk of experiencing sexual harassment than the older workers.

Working years in OOH primary care was not associated with occupational violence. The findings from other researchers are contradictory. Some studies found that more years of experience as a physician were associated with less occupational violence [14,20]. Others found the opposite [22].

Drug influence and mental illness were the most frequently perceived causes of verbal abuse, threats, and physical abuse. Several studies confirm the association between occupational violence and drug use/mental illness [18,19,23,24]. An American study found a higher incidence of violence from people with mental illness, and this risk of violence seemed to be higher because they often displayed other traits associated with violence, such as substance abuse [25]. Studies among GPs in Australia also found that violence and aggression are associated with patient drug and alcohol intoxication and psychological disorders [13,14]. Very few responders in our study perceived anxiety or pain as the causes of adverse behaviours. In a Dutch study [26], anxiety, depression, and pain were found to be connected to half of the situations where patients acted aggressively or inappropriately in contact with a doctor's office. The differences might be attributable to cultural differences in explanatory mechanisms related to the same phenomenon. Expressed anger might, for example in a Norwegian culture, be explained as drug-induced behaviour or mental illness symptoms, while in another culture the same phenomenon would be explained as caused by pain.

A substantial number of health care personnel in Norwegian OOH primary care experience verbal abuse, threats, physical abuse, and sexual harassment. As such, our study contributes to the overall understanding of violence in OOH primary care. More knowledge is needed regarding the consequences of the occupational violence in OOH primary care. In any case, employers should have a zero-tolerance policy on violence towards health personnel and take action to prevent undesired behaviour in OOH primary care.

Acknowledgements

The study was initiated by Sandnes legevakt and funded by the National Centre for Emergency Primary Health Care, Uni Health, Uni Research, Bergen, Norway. The authors would like to thank all participating OOH primary Care Centres. Thanks are also offered to Hogne Sandvik for giving valuable comments on the manuscript.

Conflict of interest

None declared.