Professional Documents
Culture Documents
Psychosomatics 2018:&:1 7 © 2018 Academy of Consultation-Liaison Psychiatry. Published by Elsevier Inc. All rights reserved.
TagedPLisa J. Rosenthal, M.D., FACLP, Ashley Byerly, M.B.A., M.P.H., Adrienne D. Taylor, M.D.,
Zoran Martinovich, Ph.D.
Background: Threatening and assaultive behaviors physical assault. Of those with any incident of physical or
against healthcare workers is a growing national con- verbal violence, 60.2% endorsed at least one posttrau-
cern.1,2,3 To assess the incidence and impact of aggres- matic symptom, 9.4% missed work, and 30.1% had
sion against healthcare workers, a safety and quality thoughts about leaving their job or career. The reported
improvement project was initiated in an academic, ter- impact was the same for physical or verbal incidents.
tiary care, and urban hospital. Methods: Through the Discussion: Physical and verbal abuse of healthcare
Northwestern Academy of Quality and Safety Initiatives workers is prevalent and has a significant impact on
program, an invitation to complete an online survey was employee engagement and posttraumatic spectrum symp-
sent to healthcare workers. The survey inquired about toms. These results are based on a cross-sectional survey
prevalence, location, and type of experience of physical at one institution and may have a significant selection and
or verbal abuse by patients or families. Other goals were: response bias. Conclusion: Assessment of both verbal and
1) worker knowledge and use of reporting systems physical aggression against healthcare workers should be
2) effect on healthcare worker engagement, and 3) report standard. Front line consulting psychiatrists and psychi-
of posttraumatic symptoms. Results: 34.4% of healthcare atric programs for employee wellness could assess and
workers reported any incident of verbal or physical vio- manage this impact.
lence in the proceeding 12 months, with 13.5% reporting (Psychosomatics 2018; &:1 7)
TagedPKey words: healthcare worker wellness, workplace violence, posttraumatic stress, physician engagement, security,
nursing.
Received February 6, 2018; revised April 27, 2018; accepted April 27, Send correspondence and reprint requests to Lisa J. Rosenthal, M.D.,
2018. From the Department of Psychiatry and Behavioral Sciences FACLP, Northwestern University, Feinberg School of Medicine, 446 E
(L.J.R., A.D.T., Z.M.), Northwestern University, Chicago, IL; Ontario St, 7th Floor, Chicago, IL 60611; e-mail: lrosenth@nm.org
Human Resources (A.B.), Northwestern University, Feinberg School © 2018 Academy of Consultation-Liaison Psychiatry. Published by
of Medicine, 446 E Ontario St, 7th Floor, Chicago, IL 60611 Elsevier Inc. All rights reserved.
TagedPact or threat of violence, ranging from verbal abuse to TagedPunfamiliarity with best reporting practices or available
physical assaults directed toward persons at work or on support services, and disinterest or cynicism about
duty.”8 As defined, violence includes events that do not reporting sequela such as counseling.
result in physical injury. However, these events are not
included in mandated reporting, thus there is an ongo- TAGEDH1MATERIALS AND METHODSTAGEDN
ing need to demonstrate the incidence and impact on
involved workers. Previous healthcare worker surveys TagedPThrough the Northwestern Academy of Quality and
have demonstrated a significant number of verbal Safety Initiatives (AQSI) program, a multidisciplinary
assaults on healthcare workers and have included some team was assembled by the primary author (LR),
impact data, but few were completed in the US with including members from medicine, psychiatry, emer-
wide survey of hospital environments, worker types, gency medicine, social work, nursing leadership, human
or inclusion of posttraumatic spectrum symptoms.9 resources, and security, as well as resident trainees from
Although the Joint Commission mandates action taken psychiatry and internal medicine. The AQSI selected
to assess and address violence, it is not clear that hospi- this proposed project for a year-long certificate program
tals are collecting data of incidents without physical that includes classes and leadership coaches. The
injury routinely; without this data it seems unlikely that quality initiatives focused on Northwestern Memorial
they are creating programs that truly meet the needs of Hospital, an 894-bed, tertiary care, urban, academic
the workforce. medical center. The hospital had approximately 86,300
TagedPConsulting psychiatrists in the hospital are often emergency room presentations and 69,400 observation
involved in clinical care and policy creation for agitated and inpatient admissions in 2016.
patients, and are frequent witnesses to violent incidents TagedPThe team reviewed existing reports of violence
and their aftermath. There is limited evidence base to against healthcare workers at the hospital and found
guide assessment of physicians and other healthcare three administratively separate systems. These included
workers following these events, or to guide support for a Risk Management driven online reporting system
wellbeing of affected staff and faculty. With further for safety or systems concerns involving patients, the
data, psychiatrists can be advocates to assess and Human Resources Employee Incident Report online
address the impact of aggressions against healthcare system focused on Occupational Safety and Health
workers. Consulting psychiatrists are also uniquely Association (OSHA) reportable events, and the security
posed to impact perceptions of “psychiatric patients,” department’s informal record of violent events on the
including improving screening for common etiologies of hospital campus that included statements about injured
aggression such as neurocognitive disorders. employees. Details included in these reports did not per-
TagedPIn our tertiary care, urban, academic medical center mit cross referencing. In 2016, a total of 117 reports
there was strong interest in addressing violence, but it were made in these three systems, 56 of which were
was immediately apparent that we first needed accurate through the appropriate channel of Human Resources.
measurement of physical and verbal assault directed These results suggest far less than 2% reported any inci-
against our workforce. Standard methods of reporting dent, and only 19 reports referenced verbal assault.
existed but demonstrated low incidence and did not Data was so limited that the AQSI team was unable to
assess for prevalence, location, or type of healthcare initiate interventions to address violence without fur-
worker involved, or resulting impact of incidents with- ther information.
out physical injury in a standardized fashion. TagedPInvestigation began to assess the prevalence, tar-
TagedPHealthcare workers may be reluctant to report vio- gets, locations, and impact of physical and verbal vio-
lence, as described in Phillips’ review on the epidemic of lence within the hospital. With the support from AQSI
violence in healthcare.1 Root cause analysis suggested and Hospital Administration, an IRB exempt online
that this was the case at our institution. Other potential survey was sent to all medical staff physicians, nurses,
barriers included multiple disconnected and laborious nurse assistants, advance practice providers, and
reporting systems, limited knowledge of reportable social workers. Reported populations on distribution
event criteria, healthcare worker reluctance to “com- e-mail lists were 2005 physicians, 445 nurse practi-
plain,” lack of education on posttraumatic symptoms, tioners or physician assistants, 2455 nurses, and 52
TagedPSocial Workers. Approximately 275 certified nurse sTagedP o, the relevant statistics for the contrasted full analytic
assistants received invitations; numbers are not exact sample are listed.
due to manager distribution methods. Other types of
healthcare workers such as respiratory therapists, physi- TAGEDH1RESULTSTAGEDN
cal and occupational therapists, radiology technicians,
and medical trainees were excluded. TagedPOf 948 respondents to the survey, 802 were eligible to
TagedPThe survey required at least 12 months employ- provide complete responses (Table 1). The vast majority
ment at the hospital and asked a series of 18 ques- of ineligible responses were due to duration of employ-
tions about physical or verbal assault and its impact. ment under twelve months. There was an approxi-
The goal was to determine prevalence, location, and mately 19% response rate to the survey, without the
type of healthcare worker experience of physical or exact number of direct daily care providers (mostly cer-
verbal abuse by patients or families. Other goals tified nursing assistants). Of the 802 eligible respond-
were: 1) worker knowledge and use of reporting sys- ents, 34.4% reported any incident of physical or verbal
tems, 2) effect on healthcare worker engagement, and abuse in the proceeding twelve months, of which 13.5%
3) report of posttraumatic symptoms. The project were identified to be physical assault (Table 2).
was deemed IRB exempt. Data was reviewed with a TagedPResults found that the emergency department had
statistician. the highest incidence and prevalence of violence, with
TagedP83.7% reporting any incident, and a mean of 28.22 TagedPevents, 50.9% of direct daily care providers reporting a
events per respondent. Medicine wards, psychiatry, and mean of 3.6 events, and 21.9% of physicians reporting
critical care areas had the next highest percentages of any incident of violence with a mean of 2.73 events.
affected employees, with the highest mean number of TagedPIncidence was not statistically different between
events. The lowest percentage of affected employees, male and female respondents but mean number of
and lowest number of events per worker were within events was, with 31.4% of men reporting a mean of 7.5
women’s health areas. incidents per worker, and 35.2% of women with mean
TagedPResults by healthcare worker role were significantly 2.25 events, and a p value of < .001. Thus, males
different, with 39.1% of nurses reporting a mean of 4.15 reported far more events per person than women.
TABLE 2. Physical and Verbal Assaults by Sex, Race, Clinical Role, Duration of Employment, and Department
Variable Total Any assault Physical assault Verbal assault
N % No. of CI % No. of CI % No. of CI
incidents incidents incidents
(mean) (mean) (mean)
Total sample 802 34.4 3.35 3.23 3.48 13.5 0.36 0.32 0.40 31.9 2.70 2.58 2.81
Sex
Male 169 31.4 7.50*** 7.10 7.92 9.4 0.46 0.36 0.57 22.2 6.31*** 5.95 6.70
Female 630 35.2 2.25*** 2.14 2.37 10.5 0.33 0.29 0.39 32.7 1.73*** 1.63 1.84
Race
White 585 33.8 3.68*** 3.53 3.84 14.4 0.41*** 0.36 0.46 31.5 2.94*** 2.80 3.08
Black 65 40.0 1.40*** 1.14 1.72 10.8 0.22 0.13 0.36 36.9 1.02*** 0.80 1.29
Latino 36 41.7 5.00*** 4.32 5.79 13.9 0.19 0.09 0.41 38.9 4.22*** 3.60 4.95
Asian 88 29.5 2.10*** 1.82 2.43 11.4 0.24 0.16 0.37 28.4 1.74*** 1.48 2.04
Clinical role
Physician 160 21.9 2.73*** 2.49 3.00 8.1 0.26* 0.19 0.36 21.3 2.43* 2.20 2.69
Clinical nurse 435 39.1 4.15*** 3.96 4.35 18.2 0.49*** 0.43 0.56 35.4 3.19*** 3.03 3.37
APP (APN/PA) 71 11.3 0.38*** 0.26 0.55 4.2 0.04*** 0.01 0.13 11.3 0.27*** 0.17 0.42
Direct daily care 57 50.9 3.60 3.14 4.12 19.3 0.46 0.31 0.67 47.4 2.96 2.55 3.45
Social worker 23 52.2 2.13* 1.61 2.82 4.3 0.04* 0.01 0.31 52.2 1.83* 1.35 2.47
Duration of employment
Employed 1 to 192 37.0 6.16*** 5.82 6.52 20.8 0.58*** 0.48 0.70 34.9 5.18*** 2.20 2.69
< 3 years
Employed 3 to 201 37.8 3.04* 2.81 3.29 14.9 0.43* 0.35 0.53 35.3 2.45* 3.03 3.37
< 8 years
Employed 8 to 213 34.7 2.59*** 2.38 2.82 11.3 0.23*** 0.17 0.30 30.5 1.72*** 0.17 0.42
< 16 years
Employed 16+ years 196 28.1 1.76*** 1.58 1.95 7.1 0.20*** 0.15 0.27 27.0 1.57*** 1.41 1.76
Hospital department
Emergency 49 83.7 28.22*** 26.78 29.75 38.8 1.92*** 1.57 2.35 81.6 25.04*** 23.68 26.48
Medicine 190 41.6 3.11* 2.86 3.37 18.9 0.44* 0.36 0.55 39.5 1.99*** 1.80 2.20
Critical care 88 43.2 2.16*** 1.87 2.49 23.9 0.52* 0.39 0.70 37.5 1.61*** 1.37 1.90
Psychiatry 40 47.5 1.83*** 1.45 2.30 7.5 0.08* 0.02 0.23 47.5 1.55*** 1.21 1.99
Neurology 36 33.3 1.47*** 1.12 1.93 11.1 0.36 0.21 0.62 33.3 1.08*** 0.79 1.48
Orthopedics 16 25.0 1.44*** 0.96 2.16 18.8 0.44 0.21 0.62 18.8 0.75*** 0.43 1.32
Postsurgical 61 29.5 1.41*** 1.14 1.74 9.8 0.15* 0.08 0.28 29.5 0.95*** 0.74 1.23
Cardiology 35 22.9 1.03*** 0.74 1.43 8.6 0.31 0.17 .057 22.9 0.80*** 0.55 1.16
Operating room 53 17.0 0.91*** 0.68 1.20 0 13.2 0.81*** 0.60 1.09
Procedural areas 50 26.0 0.66*** 0.47 0.93 4.0 0.04* 0.01 0.16 24.0 0.36*** 0.23 0.57
Oncology 55 27.3 0.58*** 0.41 0.82 12.7 0.18* 0.10 0.34 20.0 0.35*** 0.22 0.54
Research unit 5 20.0 0.20* 0.03 1.42 0 20.0 0.20* 0.03 1.42
Women’s health 109 9.2 0.17 0.11 0.27 1.8 0.03*** 0.01 0.09 7.3 0.15*** 0.09 0.24
TagedPrace, 16% (N = 404) on gender, and 13% (N = 333) on TagedPworkers endorsing verbal and physical violence. This
sexual identity. study did not assess whether these symptoms were further
associated with missed work or career and job dissatisfac-
TAGEDH1DISCUSSIONTAGEDN tion, which could be investigated. A published prospective
study of violence did not report percentages of affected
TagedPA large number and variety of healthcare workers indi- respondents but did find significant impact of posttrau-
cated feeling threatened or abused by patients and fami- matic symptoms on work engagement.12 There did not
lies in a tertiary care, urban, academic medical center. appear to be significant difference in impact between types
The highest impact appears to be on nurses and direct of healthcare workers, despite differences in incidence.
care providers, and overall numbers were far higher TagedPThis survey demonstrated barriers to reporting at
than formal system existing reports suggested. Hospi- our institution, including healthcare worker knowledge
tal areas most affected included the emergency of reportable events and appropriate reporting chan-
department, medicine floors, psychiatry unit, and crit- nels. Tested knowledge of reporting was far below per-
ical care areas. ceived knowledge. There was clear opportunity for
TagedPThe results of this survey may have a significant improvement in recording and addressing these inci-
response bias, as those employees affected by violence dents, as has been recommended after other investiga-
may have been more likely to respond to a survey invi- tions.13 Healthcare organizations could do a better job
tation on that topic. Thus, actual rates and impact of actively tracking events, which in turn must be
might be lower than suggested. In addition, the survey prompted by mandated reporting.
was at a single, urban, academic institution, and results TagedPOur community is responding, including the crea-
may have been influenced by factors at this site not pres- tion of a single pathway for reporting violence against
ent at others. However, the numbers reported are simi- employees, with creation of specific reports to assess
lar to, and in some cases lower than, prior reports, physical assault, attempted assault, and verbal violence.
including incidence and impact.9,10 Existing reporting systems described above have new
TagedPHigher rates of reported violence in newer employ- methods of coordination to appropriately funnel all
ees may suggest that they are more vulnerable to vio- types of violence against healthcare workers to the
lence, or may be more likely to interpret events as Human Resources system. For example, if a patient
violent or assaultive. Alternatively, long-term health- safety event report includes a description of violence
care workers may become desensitized, or be less likely against a healthcare worker, Human Resources is now
to interpret events as violent. alerted to the report. Education about the reporting
TagedPReported impact on work engagement was found process has been expanded, through intranet educa-
to be highly significant, suggesting that many affected tional campaigns, mandated employee education about
healthcare workers miss work and consider leaving violence and reporting, and dissemination of informa-
their jobs, or even their careers, because of both ver- tion by security and nursing leadership. The human
bal and physical violence. The reported impact was resources department contacts every employee who sub-
the same whether the worker endorsed verbal or phys- mits a report, and the health system is investigating
ical assault. The large reported impact of verbal options to expand the response to employees who may
harassment is particularly striking because verbal be less interested in counseling (for example, offering
assault reports are not mandated or otherwise encour- classes in self-defense).
aged. Verbal assault, along with all lessor injuries that TagedPComments received in this survey included many
do not require medical treatment beyond first aid, are variations of complaints about “psych patients,” though
not considered Occupational Safety and Health Asso- the majority of healthcare worker injuries from violence
ciation reportable events or included in most pub- are due to neurocognitive disorders and intoxication.1
lished numbers despite some prior evidence that the Education with colleagues and staff is critical to the liai-
prevalence is high.11 son function of consulting psychiatrists, to address
TagedPAlthough this survey was not designed to diagnose assumptions or biases about psychiatric disorders and
acute stress disorder or posttraumatic stress disorder, the etiologies of violence, and to ensure that clinical
posttraumatic symptoms appear prevalent among pathways are adequate. Consulting psychiatrists are
TagedPalso uniquely posed to address burnout and wellness in TagedPThe authors report no proprietary or commercial
our co-workers. interest in any product mentioned or concept discussed
in this article.
TAGEDH1CONCLUSIONTAGEDN
Acknowledgments: W TagedP e are thankful to our col-
TagedPPhysical and verbal abuse of healthcare workers is preva- leagues and team members on AQSI project “Calling Dr.
lent and has a significant impact on employee engagement Fairbanks”: Ross York-Erwin, Billy Wade, Robert Tur-
and posttraumatic spectrum symptoms. These results are elli, Arielle Sommer, Mohammad Salahuddin, Patricia
based on a cross-sectional survey at one institution and Roberts, Geetha Reddy, Sylvia Ranalli, Megan Oakford,
may have a significant response bias. There is opportunity Megan Malladi, Eric Gausche, Rachel Cyrus, Alan Tony
for improvement in reporting and addressing these inci- Amberg, the AQSI leadership, and the faculty and staff
dents, and assessment of both verbal and physical aggres- of Northwestern Memorial Hospital who participated in,
sion against healthcare workers should be standard. and supported this survey.
References
TagedP 1. Phillips JP: Workplace Violence against Health Care Workers in TagedP 8. CDC website accessed August 16, 2017. https://www.cdc.gov/
the United States. N Engl J Med 2016; 374(Apr(17)):1661–1669 niosh/docs/2002-101/default.html
TagedP 2. Gerberich SG, Church TR, McGovern PM, et al: An epide- TagedP 9. Pompeii LA, Schoenfisch AL, Lipscomb HJ, Dement JM,
miological study of the magnitude and consequences of work Smith CD, Upadhyaya M: Physical assault, physical threat,
related violence: the Minnesota Nurses' study. Occup Environ and verbal abuse perpetrated against hospital workers by
Med 2004; 61(Jun(6)):495–503 patients or visitors in six U.S. hospitals. Am J Ind Med 2015;
TagedP 3. Gates DM: The epidemic of violence against healthcare work- 58(Nov(11)):1194–1204
ers. Occup Environ Med 2004; 61(Aug(8)):649–650 TagedP10. Nachreiner NM, Gerberich SG, Ryan AD, McGovern PM:
TagedP 4. Kuehn BM: Violence in health care settings on rise. JAMA Minnesota Nurses’ study: perceptions of violence and the
2010; 304(5):511–512 work environment. Ind Health 2007; 45:672–678
TagedP 5. Occupational Health and Safety Administration. Workplace TagedP11. Behnam M, Tillotson RD, Davis SM, Hobbs GR: Violence in
Violence in Healthcare. US Department of Labor. Print. the emergency department: a national survey of emergency
Occupational Safety and Health Administration (OSHA). medicine residents and attending physicians. J Emerg Med
3826. 12/2015. Accessed on 4/5/18. https://www.osha.gov/dsg/ 2011; 40(May(5)):565–579
hospitals/workplace_violence.html TagedP12. Kowalenko T, Gates D, Gillespie GL, Succop P, Mentzel
TagedP 6. Guidelines for Preventing Workplace Violence for Health TK: Prospective study of violence against ED workers. Am J
Care and Social Services Workers (EPUB j MOBI). OSHA Emerg Med 2013; 31:197–205
Publication 3148, (2015). TagedP13. Findorff MJ, McGovern PM, Wall MM, Gerberich SG:
TagedP 7. Bureau of Labor Statistics, online data tool. accessed on 4/5/ Reporting violence to a health care employer: a cross-sec-
18: https://data.bls.gov/gqt/InitialPage tional study. AAOHN J 2005; 53:399–406