Schulte JM, Nolt BJ, Williams RL, Spinks CL, Hellsten JJ. Violence and Threats of Violence Experienced by Public Health Field-Workers. JAMA. 1998;280(5):439-442. doi:10.1001/jama.280.5.439
From the Bureau of HIV/STD Prevention, Texas Department of Health, Austin.
Context.— Public health workers may work with clients whose behaviors are risks
for both infectious disease and violence.
Objective.— To assess frequency of violent threats and incidents experienced by
public health workers and risk factors associated with incidents.
Design.— Anonymous, self-administered questionnaires.
Setting.— Texas sexually transmitted disease (STD), human immunodeficiency virus
and acquired immunodeficiency syndrome (HIV/AIDS), and tuberculosis (TB) programs.
Participants.— Questionnaires were completed by 364 (95.5%) of 381 public health workers
assigned to the programs. The STD program employed 131 workers (36%), the
HIV/AIDS program, 121 workers (33%), and the TB program, 112 workers (31%).
Main Outcome Measures.— The frequencies with which workers had ever experienced (while on the
job) verbal threats, weapon threats, physical attacks, and rape, and risk
factors associated with those outcomes.
Results.— A total of 139 (38%) of 364 workers reported 611 violent incidents.
Verbal threats were reported by 136 workers (37%), weapon threats by 45 (12%),
physical attacks by 14 (4%), and rape by 3 (1%). Five workers (1%) carried
guns and/or knives while working. In multiple logistic regression, receipt
of verbal threats was associated with worker's male sex (odds ratio [OR],
2.4; 95% confidence interval [CI], 1.5-4.0), white ethnicity (OR, 2.4; 95%
CI, 1.4-4.1), experience of 5 years or longer (OR, 2.2; 95% CI, 1.3-3.8),
weekend work (OR, 1.8; 95% CI, 1.1-3.1), and sexual remarks made to the worker
by clients (OR, 2.0; 95% CI, 1.2-3.5). Receipt of weapon threats was associated
with worker's male sex (OR, 5.7; 95% CI, 2.4-15.3), white ethnicity (OR, 4.0;
95% CI, 1.8-9.3), age of 40 years or older (OR, 2.5; 95% CI, 1.1-5.8), work
experience of 5 years or longer (OR, 2.7; 95% CI, 1.2-6.0), rural work (OR,
3.6; 95% CI, 1.3-10.1), being alone with the opposite sex (OR, 3.7; 95% CI,
1.6-9.7), and interaction with homeless clients (OR, 5.2; 95% CI, 1.7-18.8).
Physical attacks were associated with sexual remarks made to the worker by
clients (OR, 4.2; 95% CI, 1.4-13.9). No risk factors predicting rape were
Conclusions.— Violence directed toward public field-workers is a common occupational
hazard. An assessment of what situations, clients, and locations pose the
risk of violence to public health workers is needed.
SOME CLIENTS seeking traditional public health services for sexually
transmitted diseases (STDs), human immunodeficiency virus (HIV) infection,
the acquired immunodeficiency syndrome (AIDS), and tuberculosis (TB) may engage
in behaviors that are associated both with a risk of acquiring those diseases
and with violence. For example, crack cocaine use and other illicit drug use
are recognized risk factors for HIV infection, syphilis, and TB.1- 5
Crack cocaine use is also linked to violence. The drug's pharmacological
effects include paranoid thinking and aggressive behavior.6,7
Fatal injury after cocaine use, as measured by detection of cocaine metabolites
in blood and urine, would rank as 1 of the 5 leading causes of death in New
York City residents aged 15 to 44 years.8
Violence does spill into emergency departments and psychiatric care
and has been well documented as an occupational hazard for health care workers
and social workers.16- 18
However, risks for on-the-job violence have not been studied among public
health field-workers who routinely work outside clinics and in communities
with patients and other persons who have been exposed to STDs, HIV/AIDS, and
The public health field-workers are not physicians or nurses, but their
fieldwork is vital for patient follow-up and to stop disease transmission.
Workers in STD, HIV/AIDS and TB programs notify patients of their exposure
to disease, arrange disease-specific testing, schedule appointments, conduct
interviews with patients and exposed persons, administer tuberculin skin tests,
draw blood for HIV and STD testing, transport patients to the clinic, administer
directly observed therapy for TB, and monitor outcomes for TB and STD therapy.
Typically these workers have college degrees or some college and undergo training
to work with patients with STDs, HIV/AIDS, and TB using training materials
and classes developed by the Centers for Disease Control and Prevention.
In 1994, an HIV outreach worker in Austin, Tex, was physically attacked
during the late afternoon by a client who had received condoms from her. Two
of us (J.M.S., R.L.W.) had prior experience with verbal threats and weapon
threats (display of a gun) made by STD and TB patients to ourselves and to
other public health workers conducting fieldwork or epidemiologic studies
in other states. During follow-up of the Austin incident, 3 managers in Texas
STD and HIV programs indicated 1 or more of their public health workers had
experienced verbal threats or been kicked or hit while doing fieldwork. Because
the exact number and frequency of such incidents were unknown, we decided
to evaluate the frequency of the presence of violent incidents and risk factors
associated with verbal threats, weapon threats, physical attacks, and rapes
among public health workers in STD, HIV/AIDS, and TB programs.
The questionnaire was jointly developed by the STD and HIV/AIDS programs
and reviewed by STD or TB program managers in Dallas, Houston, and Austin.
Pilot testing of the questionnaire was done with 10 Florida public health
workers who had worked with one of us (J.M.S.) on prior outbreak investigations.
Program managers in 48 regional and local health departments in Texas identified
a total number of 381 public health field-workers employed in HIV/AIDS, STD,
and TB programs; they did not provide workers' names.
During a 6-week period beginning in late 1994, we did 1 mailing of 381
questionnaires for return by February 28, 1995. Each questionnaire included
a cover sheet outlining reasons for the study, describing the Austin attack,
and asking each worker to complete the anonymous, self-administered questionnaire.
Participation was not mandatory, and no attempt was made to recontact nonresponders.
The questionnaire was part of an evaluation of ongoing program operations
in STD, HIV/AIDS, and TB programs and was exempt from review by the health
department's institutional review board.
We asked about workers' demographics, the types of clients with whom
they worked, work hours and environment, job duties, 4 different types of
violent incidents ever experienced during their fieldwork,
and safety precautions routinely taken at the time of the survey. We asked
public health workers to provide yes/no responses as to whether they planned
their route ahead of time, did fieldwork early in the day, sent 2-person teams
to the field, carried a protective weapon (gun and/or knife), used a beeper
and/or cellular phone, and did not complete paperwork in the field. With the
exception of carrying a weapon, all those safety precautions are common in
public health programs.
We defined 4 violent outcomes: verbal threat, weapon threat, physical
attack, and rape. We asked if the worker had ever experienced those events
while doing public health fieldwork and when the last occurrence was. A verbal
threat was defined as a client's vocalization of intent to harm or injure
the worker. A weapon threat was one made to the worker while a client displayed
a weapon; we asked what weapon was shown. We defined a physical attack as
a punch, kick, hit, or slap to the worker's body. We specifically asked if
workers were raped or sexually assaulted while doing fieldwork. Descriptive
statistics were used to characterize the workers' demographics and potential
risk factors associated with violent outcomes. Univariate analyses were done
to measure the association between the individual risk factors and each of
the 4 outcomes. Stepwise multiple logistic regression using maximum likelihood
estimates was used to create 4 multivariate models using each of the outcome
variables. Epi Info version 6 was used for data entry, descriptive statistics,
and univariate analyses19; SAS version 6.11
was used for multiple logistic regression.20
In the logistic regression models, 22 risk factors were eligible to enter
into each model using a P value of .10 as the significance
level for entry. Results of univariate analyses and logistic regression are
presented as odds ratios (ORs) with 95% confidence intervals (CIs).
Preliminary analyses showed that black and Hispanic ethnicity did not
have significant associations with any of the 4 violent outcomes, and race/ethnicity
was used as a dichotomous white/nonwhite variable in the models. Safety precautions
taken by workers were not associated with any of the 4 violent outcomes and
were excluded from the final models presented in this article.
Overall, more than half the workers were men with a mean age of 38.7
years, and approximately equal proportions of the workers were white, black,
and Hispanic. The workers had a mean of 4.8 years of experience, and most
worked in urban locations (Table 1).
The public health workers worked in 43 Texas counties, and 165 (45%) were
assigned to the state's 3 most populous counties (Harris, Dallas, and Bexar)
where Houston, Dallas, and San Antonio are located.
Respondents to the questionnaire totaled 364 workers (95.5%); 139 (38%)
reported 611 violent incidents, an average of 4.4 incidents per worker exposed
to violence. Verbal threats were reported by 136 workers (37%); weapon threats
by 45 workers (12%), physical attacks by 14 workers (4%), and rapes by 3 workers
(1%). The 45 workers experiencing weapon threats reported the weapon last
used was a gun (28 threats [62%]) or knife (14 threats [31%]); unspecified
weapons were used 3 times (7%). The median number of verbal threats was 1
(range, 1-30); weapon threats, 1 (range, 1-10); and physical attacks, 1 (range,
1-3). No worker reported more than 1 rape. The year that workers last reported
the most recent occurrence of violent incidents is shown in Table 2.
These 611 incidents included 489 verbal threats, 94 weapon threats,
25 physical assaults, and 3 rapes. Of the 139 workers experiencing violent
events, 90 (65%) reported only 1 type of violent incident, 39 (28%) reported
2 different violent incidents, and 10 (7%) reported 3 different violent incidents.
No worker reported all 4 types of violent incidents. The types of violent
incidents that workers reported are in Table 3.
Reported regular safety precautions included mapping a route ahead of
time (277/364 [76%]), not doing paperwork in the field (108/364 [30%]), always
working as a 2-person team (67/364 [18%]), carrying a weapon (gun and/or knife)
(5/364 [1%]), carrying a beeper (137/364 [38%]), carrying a cellular phone
(68/364 [19%]), and doing fieldwork early in the day (197/364 [54%]). No safety
precautions were associated with any violent outcome. Of the 5 workers who
carried a weapon for protection, 3 reported only verbal threats; 1 reported
both a verbal threat and a weapon threat; and 1 reported a combination of
a verbal threat, a weapon threat, and rape.
In univariate analysis (Table 4),
verbal threats were associated with worker's male sex; white ethnicity; experience
of 5 years or longer; STD work; patient transport; being alone with the opposite
sex; contact with certain clients (pimps, homeless persons, alcoholics, gang
members, and intoxicated drug users); visits to specific establishments (flophouses,
bars, and crack houses); and client behavior (asking for needles, making sexual
remarks to the worker). Weapon threats (Table 4) were associated with worker's male sex; white ethnicity;
age of 40 years or older; experience of 5 years or longer; work with STD clients;
rural work; contact with certain clients (pimps, homeless persons, alcoholics,
gang members, and intoxicated drug users); visits to specific sites (flophouses,
bars, and crack houses); and contact with clients who asked for needles. Physical
attacks (Table 5) were associated
with sexual remarks that clients made to public health workers. No risk factors
were associated with rape in univariate analysis.
In multiple logistic regression (Table 5), verbal threats were associated with worker's male sex,
white ethnicity, experience of 5 years or longer, weekend work, and sexual
remarks that clients made to the worker. Receipt of weapon threats was associated
with worker's male sex, white ethnicity, age of 40 years or older, experience
of 5 years or longer, rural work, being alone with the opposite sex, and contact
with homeless clients. Physical attacks were associated with sexual remarks
clients made to the worker. No risk factors predicting rape were identified.
Violence was declared a public health emergency in 1992,21
and we have found that violent threats and incidents are commonly made by
clients to public health field-workers. Almost 40% of Texas public health
workers reported 1 incident, and more than one third of those workers had
experienced 2 or more types of violent threats or incidents. The years in
which violent incidents last occurred were reported for 142 violent incidents;
47% were in the year preceding the study. One percent of the workers carried
weapons as protection while working and did so before Texas passed a concealed
weapons law in 1996. The frequency of such incidents may be underestimated
because we surveyed only current employees and did not reach former workers,
who could have resigned because of actual or potential violence.
Violence and its consequences are already documented occupational hazards
for social workers and health care workers employed in emergency departments
and psychiatric care facilities. Between 42% and 100% of nurses, psychiatrists,
and other therapists in selected US psychiatric care facilities have been
assaulted at least once.11- 14
In a survey of 127 emergency departments in US teaching hospitals, 43% reported
at least 1 physical attack on a staff member each month.17
During 2 years, 1 Canadian teaching hospital reported 242 injuries related
to physical abuse and 646 incidents of verbal abuse or physically threatening
behavior.9 In published studies, social workers
have reported that verbal threats (23%-83%), weapon threats (18%), and physical
attacks (3%-40%) are common.16- 18
Whether public health workers have occupational rates of violence that equal
or exceed those of social workers or other health care workers is uncertain;
1 study of minority workers in Los Angeles found that psychiatric hospital
workers had an assault rate 38 times higher than that of public health workers.22
The violence that public health workers experienced in Texas differs
from that found in emergency departments and psychiatric care facilities,
which are fixed locations where guards, metal detectors, and other security
devices can provide some protection to health care workers. Guidelines to
protect health care workers and social service workers in fixed locations
have been published.23,24 Such
measures are not completely applicable to mobile public health workers who
must visit sites that include crack houses, housing projects, and bars where
activities can include prostitution, drug sales, and violent crime. In addition,
results from Texas may differ from the experience of workers in other states.
The challenge facing public health field-workers is working safely in communities
where they must visit repeatedly while imparting and seeking information that
clients may not want to hear or reveal. Clearly, public health workers have
duties that the National Institute for Occupational Safety and Health25 characterizes as increasing a worker's chance of
workplace assault. These duties include delivery of services to clients, working
alone or in small numbers, working in high-crime areas, and working in community-based
This study provides a preliminary look at the occupational violence
experienced by public health workers and identifies risk factors possibly
associated with violence. For example, we found that workers who were male
and of white ethnicity and who had 5 years' or more experience are more likely
to experience verbal and weapon threats. We also found that sexual remarks
made by a client to a public health worker are associated with verbal threats
and physical attacks. However, these findings do not provide the complete
context of the violence. We did not collect detailed information about the
triad of setting, client, and public health worker as related to specific
incidents. In addition, since we asked whether public health workers had ever experienced violent incidents during their fieldwork,
those with longer work experience would have greater exposure.
Given the work that a public health worker performs, verbal threats
may not be unexpected. In an STD or HIV interview, for example, a public health
worker attempts to learn the client's sexual orientation (heterosexual, homosexual,
or bisexual), the types of sexual behavior (vaginal, anal, and/or oral intercourse),
and numbers of partners (regular, casual, and/or anonymous) in specific time
periods. The TB workers administering directly observed therapy, a technique
important in preventing emergence of drug-resistant strains, may face hostility
from patients during their biweekly visits that continue for months. For confidentiality
purposes, public health workers are often alone when such interviews are done,
and fieldwork in Texas is not assigned on the basis of the worker's or client's
sex or ethnicity. In addition, public health workers looking for a client
or exposed person cannot tell anyone except that individual why he or she
is being sought and cannot tell the individual who named them as being potentially
exposed to STD, HIV/AIDS, or TB.
We did not find any safety measures routinely used by public health
workers to be protective or predictive of any of the 4 violent outcomes studied.
More information is needed about the dynamics involved in violent threats
and incidents. For example, it is possible that sexual remarks that a male
client makes to a female public health worker may differ from those a female
client would make or those made when both client and field-worker are of the
Whether these preliminary findings mean that public health workers should
more closely resemble the communities they serve is debatable. Our limited
analysis suggests that some types of workers (white, male, with 5 years' or
more experience) may be more likely to experience verbal and weapon threats.
We did not assess the types of clients associated with violent threats and
incidents. In addition, public health dollars are limited, and it may not
be realistic to match workers and communities when the public health workers
in Texas include broad representation of both sexes and the state's 3 major
ethnic groups (white, black, and Hispanic).
Public health workers should not have to accept occupational violence
as part of the job.21 Education of public health
workers is important so that they are capable of assessing or managing a given
situation, place, or client that may escalate into violence. Public health
researchers studying injury and violence22,23
should collaborate with their counterparts in STD, HIV/AIDS, and TB programs
to conduct research that will help ensure safer working conditions for public