INTRODUCTION

Workplace violence (WPV) is defined by the World Health Organization (WHO) as “incidents where staff is abused, threatened, or assaulted in circumstances related to their work, including commuting to and from work, involving an explicit or implicit challenge to their safety, well-being, or health”.1 In the United States, the healthcare industry experiences high rates of injury due to WPV. Persons working in the healthcare industry are five times more likely to be injured due to WPV than workers in other service industries such as finance, food and beverage, and retail.2 Moreover, WPV against healthcare workers is increasing, with incidence rates of non-fatal workplace violence almost doubling from 6.4 to 10.4 per 10,000 workers from 2011 to 2018.3

Despite these distressing numbers, little is known about WPV in primary care. Studying WPV in primary care is important because primary care is the foundation of the US health care system and is a platform for continuous, person-centered, relationship-based care that considers the needs and preferences of the patient and their community.4 Primary care visits are the most common form of medical care delivered in an office setting. In 2008, there were approximately 956 million office visits with 51% which were in primary care settings.5 Moreover, primary care visits are intimate encounters designed to create a safe space for clinicians and patients to establish a confidential line of trust and communication. Tragically, primary care remains underfunded and understaffed as a specialty, receiving only 5% of all health care spending dollars.4 These challenges may increase safety risks for primary care clinicians. Primary care clinicians serve as gatekeepers for specialty care and to other psychosocial services. Primary care clinicians see more patients and are often the managers of most of a patient’s medications.6 This role as a gatekeeper could place clinicians at an increased risk of conflict when expectations are not met regarding referral to services or prescribing medications.

To summarize the existing literature, we conducted a narrative review of WPV in primary care settings in the US. Specifically, we sought to explore the following regarding WPV: (1) prevalence, (2) types, (3) mechanisms, (4) causes, and (5) effects.

METHODS

We conducted a narrative review of the literature related to workplace violence of primary care clinicians. For classifications of WPV in this review, we defined WPV as violence perpetrated by a customer or patient against an employee. We used the National Institute of Safety and Health definition of Type 2 WPV to develop our review criteria. Type 2 WPV is violence towards an employee by a client, family member, or visitor. Separately, we organized WPV into two categories: (1) physical, defined as pushing, spitting, property damage, punching, kicking, or sexual assault; and (2) non-physical defined as verbal insults, racist/bigoted comments, online threats/insults, verbal threats, bullying, sexual harassment, intimidation, and stalking.

To find articles, we conducted a search without any time parameters of PubMed and Ovid for articles on WPV against clinicians in adult primary care using the following MESH terms (“ambulatory care OR ambulatory care facilities” OR “outpatients” OR “family practice OR general practice OR primary care”) AND (“physician” OR “doctor” OR “general practitioner” OR “nurse practitioner” OR “physician assistant”) AND (“violence” OR “harassment, non-sexual, OR sexual harassment OR bullying OR “aggression” OR “crime” OR “crime victims” OR “abuse” OR “intimidation” OR “threat”) AND (“workplace violence” OR “occupational violence”). Additional studies were found through reference lists of included studies. Studies were selected for further analysis after reviewing the abstracts. Our inclusion criteria were (1) studies in English; (2) internal medicine, family medicine, or general practice clinicians (including physicians, nurse practitioners, and physician assistants) must be at least part of the study population; and (3) US-based setting. Our exclusion criteria were (1) exclusively inpatient settings, (2) exclusively overnight or settings where a clinician visits a patient’s home, and (3) exclusively pediatric populations in the practice. Our initial query yielded only one article about WPV in primary care settings in the US and five studies conducted in the US but not exclusively studying primary care clinicians. As such, we expanded our initial inclusion criteria to also include international studies of WPV.

To summarize article content, we used deductive thematic analysis to group articles into pre-defined areas of emphasis. Our pre-defined areas of emphasis included (1) prevalence of WPV, (2) types of WPV, (3) mechanisms of WPV, (4) causes for WPV, and (5) effects of WPV. One author (NT) reviewed all abstracts and articles and discussed findings with another author (HP), where consensus was reached on inclusion and exclusion of articles within themes. As study design and geographic area differed among included studies, no meta-analysis was conducted on findings, and instead data is presented narratively. In our analysis, we focused on findings that pertained to primary care clinicians (physicians, nurse practitioners, physician assistants) when possible; however, some studies did not conduct specific analyses on these populations.

RESULTS

Types of Surveys

After applying our expanded inclusion criteria, we selected 70 studies of which 60 were cross-sectional, two were repeated cross sectional, four were prospective studies, three were qualitative studies, and two were mixed methodology studies with both cross-sectional and qualitative components. Only five studies were conducted in the United States;7,8,9,10,11 two of which examined WPV among primary care clinicians. One of the studies involved only adult primary care clinicians.7 The other study Internal Medicine residents and assessed WPV in both the inpatient and outpatient settings.11 Two other US-based studies examined WPV clinicians in multiple specialties including physicians, nurses, and nurse’s assistants.8 One US-based study examined WPV experienced by women physicians in multiple specialties, but did not specify specific WPV incidence between specialties.9 The remainder, all international studies, were conducted in multiple countries including Barbados, Pakistan, Saudi Arabia, Japan, Turkey, Israel, Brazil, Belgium, Bulgaria, Lebanon, Portugal, South Africa, Thailand, Australia, Sweden, Serbia, New Zealand, China, Spain, Poland India, Italy, Canada, Congo, UK, Rwanda, Ireland, Croatia and Herzegovina, Germany, and Lithuania. Within studies, the sample sizes of clinicians studied ranged from 28 to 3514.12,13

Types of Violence

Definitions of WPV varied among included studies, reflecting that terminology is often inconsistently applied to episodes of WPV. Most studies were in agreement with the definition of physical violence because it requires physical contact. On the other hand, definitions of non-physical violence were more variable. Ten studies labeled non-physical violence as simply verbal abuse.12,14,15,16,17,18,19,20,21,22 The remaining studies described a wide variety of behaviors in their definitions of non-physical violence, including verbal insults, racist/bigoted comments, online threats/insults, verbal threats, bullying, sexual harassment, intimidation, and stalking.

Frequency of Violence

Thirty-three studies reported non-physical WPV to be more commonly experienced than physical WPV.9,14,20,21,23,24,25,26,27,28,29,30,31,32,33,34,35,36,37,38,39,40,41,42,43,44,45,46,47,48,49,50,51 The most common form of non-physical WPV reported was verbal abuse. The prevalence of verbal abuse within the prior 12 months ranged from 44 to 64% among outpatient clinicians.25,34,36,37,39 Some form of verbal abuse was noted in 60.7% of respondents in the one US-based primary care study.7 In contrast, the prevalence of physical violence within the prior 12 months in primary care settings was much less common, ranging from 0.5 to 15.9%.25,29,32,38,39

Since most of the studies were cross-sectional in design, there are limited data about trends in WPV in primary care settings. One prospective study and two repeated cross-sectional studies identified in the literature did not yield consistent results. A Spanish prospective study from 2010 to 2015 showed a statistically significant decrease in WPV from 20.2 to 15 per 10,000 clinicians over the 5-year period (p < 0.0001).46 One repeated cross-sectional study from Norway among randomly selected clinicians reported no significant difference in experiencing threats of violence from 1993 to 2014.19 Another repeated cross-sectional study from Italy that administered sequential surveys from 2005 to 2009 among outpatient healthcare staff found no change in physical violence and a decrease in non-physical violence over the 4-year time period.40 One US-based study noted an increase in incident rate from 4 to 6 per 10,000 worker-months between 2012 and 2014; however, this study did not specify trends in primary care settings.8

Gender

There are inconsistent conclusions about differences in WPV between men and women noted in the literature. Twelve studies found no difference in WPV between women and men. 19,22,24,25,28,29,32,34,36,42,46,47 However, seven studies found that women experienced more non-physical violence,13,18,35,39,43,48,49 and two studies that found women experienced more physical violence than men.17,39 One type of WPV, sexual harassment, was shown in ten studies to consistently affect women more than men.9,17,35,36,37,39,50,51,52,53 In contrast, five studies noted that male clinicians were more likely to experience non-physical WPV 12,14,32,51,54 while 14 studies noted that men were more likely to experience physical violence more than women. 12,31,33,35,36,44,48,49,50,51,54,55,56,57

Root Causes of WPV

When clinicians from international studies were asked what they thought contributed to incidents of WPV, there were some common themes that emerged as potential root causes. We categorize these into patient-level root causes, clinician-level root causes, clinical encounter specific root causes, and operational root causes (Table 1).

Table 1 Tyau WPV Review Table

Patient-level root causes can potentially escalate tensions and could precipitate a violent incident. Ten studies noted that patients’ substance dependence and/or psychiatric conditions led to violent incidents.7,14,24,25,26,33,46,51,58,59 Some of these conditions included anxiety and psychosis, but clinicians also noted that some patients had high levels of stress, had previous violent episodes, or had personality traits such that made them quicker to become angry.16,26 Two studies noted that violent incidents were results of reactions patients had when they received bad news such as a serious illness or death of a family member.25,58 Two studies mentioned that patients’ “lack of education” or “low educational level” contributed to episodes of WPV. 26,28 Separately, lower health literacy was mentioned as a potential root cause.25 While any of these are potential contributors, a systematic categorization of patient-level root causes has not yet been developed.

Clinician-level root causes also contributes to tensions that can lead to WPV. Two studies noted that clinicians felt that stress due to concerns about job security, poor working conditions, a lack of resources, lack of training, clinician error, and understaffing contribute to WPV.16,26 Other causes, such as clinician personality traits or communication style, could contribute to tensions as well.27,57,59

While individual patient or individual clinician root causes can contribute to WPV, clinical encounter specific factors can also lead to several unique root causes. Seven studies cited unmet expectations for care as the root cause of an incident.22,25,44,46,51,54,60 Six studies noted clinicians not meeting patient expectations, 27,35,42,46,51,54 including a physician’s refusal to complete a work letter or refusal to prescribe a particular medication. Seven studies noted a disagreement in treatment plan,15,27,35,42,44,45,61 and three studies noted an error in communication as a clinical encounter specific factor.22,25,26

Clinicians also reported that operational root causes can contribute to WPV. Ineffective workflows within an institution could lead to long workdays and understaffing, which contributes to increases in wait times.26,28 Additionally, staff shortages can lead to long waiting times, which has been listed by thirteen studies.15,18,25,27,28,32,35,42,46,54,58,59,61 Other operational problems include underdeveloped or inadequate safety policies or procedures for WPV. Clinicians noted in four studies that inadequate policies were a contributor to incidents of WPV. 22,25,26,59 Additionally, two studies also stated that clinicians felt that a lack of consequence for aggressors encouraged further WPV.25,26

Effects of Violence on Clinicians

Even though the vast majority of WPV is nonphysical, it has a significant effect on its victims. The effects can influence people psychologically and studies to date have found experiences of WPV are associated with symptoms of anxiety, depression, post-traumatic stress disorder, hyper vigilance, suspicion, burnout, distress, poor well-being, and work dissatisfaction in clinicians.22,36,54,62,63,64 Additionally, three studies have shown WPV leads to a reduction in work performance.25,26,64 Clinicians from two studies have also noted negative effects on their quality of life outside of work.15,22 Ultimately, these effects have been associated with increased absenteeism and increased turnover.41,65

Lack of institutional support exacerbates effects of WPV on primary care clinicians. Two studies note that that a lack of institutional support in response to WPV can leave clinicians feeling vulnerable and unsafe at work.30,47 Clinicians may also feel unsupported if they do not have the proper training to deal with WPV. They may feel frustrated because of a lack of de-escalation training, and because they may feel they cannot discharge violent patients from their practice.25,66 If there is no confidence in their employer for support, this may lead to decreased motivation to report concerns or incidents, exacerbating symptoms of burnout and work dissatisfaction.6,18,62. Several studies demonstrate that many clinicians do not report episodes of violence even when there are operational causes due to perceived indifference by their employer.29,54 Other researchers have concluded that empowering primary care clinicians to report may increase confidence and comfort, and may actually lead to lower incidence of WPV.23 Even if reporting does not result in lower incidence of WPV, more comprehensive data can spur employers to prioritize the problem.46

Conclusion

We found that data for WPV in primary care setting in the US was sparse, though the one study that we did find was consistent with the prevalence of WPV reported in international primary care studies, ranging from 10 to 75%.67 Non-physical violence, especially verbal abuse, was by far the most common type of violence experienced by primary care clinicians. While there were inconsistent findings with respect to how WPV may differ by gender, women were significantly more likely to be targets of sexual harassment. Primary care clinicians named numerous possible root causes for WPV, which could be categorized into patient level, clinician level, clinical encounter specific, and operational root causes. Even though most WPV is non-physical, there can be profound negative psychological and emotional effects on clinicians that can directly affect work performance and may lead to increased turnover.

This review highlights the need for increased investigation into WPV in the US. We found that the majority of international studies are small and cross-sectional, and depend on adequate reporting. Studies using longitudinal designs are scarce. Two existing longitudinal studies from Israel and Norway thus far have not shown increases in violence over time.15,19 Additional longitudinal studies from other countries including the United States are needed to better understand how WPV in primary care evolves over time. One significant challenge in data collection is the cultural belief within healthcare occupations that being verbally accosted, threatened, or even physically assaulted is “part of the job.” As a result of this attitude and because of a lack of confidence in safety policies or post incident support, the true extent of WPV remains under reported and therefore poorly understood.18,29,40 Allowing this problem to continue will only perpetuate the negative effects of WPV on clinicians. Allowing even verbal abuse to continue not only has profound psychological and emotional impact, but can have harsher downstream consequences since verbal assault is a risk factor for physical assault.68

More primary care–specific WPV data will allow for stronger national interventions to support clinicians and other clinical staff at any stage surrounding a violent incident. There are initiatives in other medical specialties, such as emergency medicine, to develop protective policies. In the hospital, advocacy from clinicians has led to new Joint Commission standards as of January 2022, which included guidance for employers aimed at reducing WPV in the hospital setting. These standards recommend processes for monitoring workplace conditions, educating clinicians, and creating a culture of safety. It has also led to proposed legislation such as the “Safety from Violence for Healthcare Employees (SAVE) Act,” which establishes federal penalties for violence against hospital-based healthcare workers. While some of these protections benefit clinicians from every specialty, no specific initiatives have been developed for outpatient settings. Additionally, more primary care–specific WPV data could assist with the creation of local office–based initiatives specifically protecting US primary care clinicians. Highlighting the unique challenges that the primary care setting presents would allow for specific protections and federal support for programs to address WPV. For example, there has been some data to suggest that implementing risk assessment can reduce short term WPV. In one US study conducted at a Veterans Affairs hospital, flagging violent patients reduced violent incidents by 91.6%.69 Another study noted a significant reduction in violent attacks by 34% when using a short-term risk assessment followed by preventive measures in a psychiatric hospital.70

Additionally, when considering WPV, the unique experiences of male and female primary care clinicians should be individually considered. Historically, women have consistently been shown to be targets of sexual harassment from both their colleagues and by patients.9,51,62 Women represent a significant proportion of clinicians, including in adult primary care related fields. In 2019, women were 38.7% of internal medicine, 41.3% of family medicine/general practice, and 53.1% of medicine/pediatrics (Association 2020) This highlights the urgent need for more rigorous research into the effects of WPV on different genders in the primary care setting.

This narrative review on primary care WPV has several limitations. We reported mainly on smaller cross-sectional studies, and the lack of prospective studies weakened our ability to analyze patterns of violence over time in the primary care setting. Additionally, due to the lack of uniform definitions and terminology of WPV, it is possible that our literature search missed additional studies. Another significant limitation is the lack of data from the United States. While it is reasonable to extrapolate some general assumptions about WPV in primary care based on international studies, it is also possible that there are unique features of primary care WPV in the US, such as systemic barriers to access of healthcare, that may be uniquely relevant. Future research in the United States needs to begin with describing the scope of WPV in primary care settings. Such studies should focus not only on describing the frequency and type of violence committed, but also conduct root cause analyses to identify common health care driven causes as well as patient driven causes. We excluded studies conducted in primarily pediatric settings from this current review, though future studies that incorporate pediatric settings may enhance our current understanding of primary care WPV.

In summary, we found that the data on primary care WPV in the USA is scarce. International studies do provide some insight into this problem. WPV can take many forms but can generally be either physical or non-physical WPV. Non-physical WPV is the most reported type of WPV, with most of it being verbal. While both men and women experience physical and non-physical forms of WPV, women are consistently more likely to experience sexual harassment, and therefore, more study on these gender differences is necessary. There are numerous potential causes of WPV in primary care including those from the patient, the clinician, clinical-encounter based causes, and operational causes. Irrespective of the form or the root cause of WPV, it leaves a detrimental impact. Finally, more, higher quality studies are needed in the United States to better inform both local and national interventions to address this problem.