Violence Prevention in the Primary Care Setting: Title and subTitle BreakA Program for Pediatric Residents FREE

INTERPERSONAL VIOLENCE is a leading killer of our nation's children and youth.¹ It must be addressed aggressively as a public health concern. Over the last 2 decades, there have been successful public health campaigns, often led by pediatricians, that have enhanced child safety through increased use of car seats, bicycle helmets, and childproof caps and the reduction of home hazards. It is time to add violence prevention to this list of important interventions.

Substantial literature exists that demonstrates that there are specific actions physicians can take if they recognize and treat violence prospectively.^{2 - 11} Physicians have the opportunity to address violence through various means, which include teaching parents about discipline, media exposure, and firearm safety; assessing youth for violence exposure, anger threshold, and weapon carrying; and working directly with youth in conflict resolution, anger management, and weapon avoidance. In fact, because pediatricians see most children and families repeatedly and confidentially, they can be pivotal participants in an interdisciplinary effort to reduce violence. However, a major barrier exists to pediatricians filling this role—they lack the knowledge and comfort to do so.¹²

Two overriding objectives guided this project. The first objective was to develop a resident program that would train future pediatricians to comfortably address violence during pediatric visits. The second objective was to explore the effect of such a program on pediatric residents' anticipatory guidance activity in a continuity clinic.

A 212-hour violence prevention program was provided for the pediatric residents in the hospital's primary care center, which predominantly serves nonadolescent children. The program included an introductory talk, role-playing sessions, a printed resident guide, and supplemental reading materials. The length of the intervention was chosen for 2 reasons: it could incorporated into a monthly seminar series within the residency program and it was sufficiently brief to replicate in other programs. One of us, a pediatric resident (C.D.J.), gave the introductory talk and summation and facilitated the role-playing sessions.

The introductory talk provided statistics regarding the problem of interpersonal violence, and outlined various skills and techniques that the resident would need to discuss the topic during routine well-child visits. The concept of educating patients to the point of eliciting behavioral change was introduced by outlining a simple process. The person must recognize the problem, have the desire and the skills to change, and weigh the costs and benefits of the change. Finally, if the benefits outweigh the costs, and the person trusts his or her skills, the change may be sustained. The program emphasized that, although facts may help parents and older children with recognizing the problem and having the desire to change, specific skills are needed to advance toward true behavioral change.

The residents were also given a framework to organize their own approaches to guiding families on sensitive issues such as violence. This framework included setting the stage, so the patient or parent understands why the topic of violence is being discussed, assessing the specific needs of the patient or parent within the realm of violence prevention, and targeting advice tailored to the patient's specific situation.

The residents practiced these skills on surrogate patients. The role playing was not scripted; however, adolescent peer educators assumed the roles of predetermined characters, including a 17-year-old mother with her 15-month-old child, an 11-year-old girl with her mother, and a 16-year-old boy alone in clinic. The residents rotated through all 3 role-playing scenarios, and time was allotted at the end of each scenario for feedback from the residents and the peer educators. Table 1 highlights the issues and skills addressed in each role-playing session.

A printed resident guide summarized the information presented in the program. It included statistics on violence, ideas for addressing psychosocial morbidity, and specific suggestions for discussing sensitive topics with patients and parents. Local violence prevention resources and references from the literature on the topics of discipline and violence prevention were provided. The resident guide and program details are available from the authors.

A survey was administered to approximately 100 parents or guardians at each of 3 points: during the weeks just before the program, immediately following the program, and 6 months after the program. The survey was administered to the parent or caretaker at the signout desk after the visit. Research assistants were trained to administer the questionnaire in a nonsuggestive manner. To include caretakers who might have difficulty reading the survey, the questions were read aloud, and answers were recorded on standardized forms. The parent or guardian was asked to select from a list of violence-related and non–violence-related topics to discuss during the well-child care visit: feeding and nutrition, infant walkers, sleep problems, immunizations, poisons, guns or violence, car and bicycle safety, discipline, and smoking. Discipline was listed separately because this topic may have been taught in other aspects of the pediatric residency program. Families that chose "guns or violence" from the list were considered to have received violence prevention guidance (VPG) during the well-child visit.

The survey was administered at 2 of the 3 resident continuity clinic sites. The residents were not aware of the survey administration at any time during the study, and were identified on the survey only by sex and level of training. Three of 28 primary care attending physicians (2 site directors, coauthor) were aware that the surveys were being used to assess resident anticipatory guidance. The research assistants were not blinded to the study objective.

Resident satisfaction with the program was assessed using a questionnaire distributed to all resident participants at the end of the 212-hour teaching session. The questionnaire requested that they rate the content, methods, and potential usefulness of the material presented, and provided space for comments or suggestions. A copy of this questionnaire is available upon request. Residents who graded questions as 4 or 5 on a scale of 1 to 5 were considered to have regarded the program positively.

The study was approved by the hospital's Committee for the Protection of Human Subjects (institutional review board).

Comparisons of resident performance regarding VPG were made using χ² or exact testing methods. Data were analyzed using the Systat version 5.2.1 software (SPSS Inc, Chicago, Ill). Odds ratios were calculated using confidence interval analysis.¹³

Three hundred eight patient visits (70.3% of eligible patients) were sampled; 93 prior to the program and 215 after the program. Of this latter group, 106 encounters were surveyed within 2 weeks of the program and 109 were surveyed 6 months later. The patients ranged in age from newborns to 15 years, with a mean of 4.3 years and a median of 3 years. One or both parents accompanied 90% of the patients, and the remainder were accompanied by grandparents or foster parents. Fifty-three (60%) of the 88 pediatric residents were assigned to 1 of the 2 primary care sites sampled. The residents whose patients were sampled prior to and after the program did not differ with respect to level of training (intern vs upper-level resident), sex, or primary care site (Table 2).

Overall, pediatric residents were more likely to offer VPG to their continuity clinic patients after the program (19.1% vs 9.7%,P=.04) (Table 3). Violence prevention guidance was offered in 17% of visits within 2 weeks of the program and 21.1% of visits 6 months after the program.

Certain patient or provider characteristics were assessed in relation to VPG (Table 4). Encounters that were not a "first visit" for the patient and encounters in which 1 or both of the patients in the room were girls were more likely to result in improvement in VPG after the program. Families with at least 1 child over the age of 7 years were more likely to receive VPG compared with families with younger children (odds ratio, 3.72; 95% confidence interval, 1.95-7.08). There was a trend for the program to enhance VPG in both age groups; however, this enhancement did not attain statistical significance. The resident's level of training did not influence the effect of the program on VPG.

To determine whether discussion of other anticipatory guidance topics increased over the study period, topics were analyzed separately (Table 5). Guns or violence was the only topic that was discussed more frequently after the 6-month period. This remained true when patients older and younger than the age of 7 years were analyzed separately.

Seventy-two (82%) of 88 residents attended the program; of these, 58 (81%) returned satisfaction surveys. Forty-eight residents (83%) thought the program increased their fund of knowledge regarding interpersonal violence, and 49 (84%) thought they learned skills for addressing this issue with their primary care patients and families. Fifty-four residents (93%) thought that they would use these skills during subsequent patient encounters.

Violence among children and youth is a well-recognized national problem. Homicide is the leading cause of death for African Americans aged 15 to 34 years and is the second-leading cause of death for all American youth aged 15 to 19 years.^{1 ,14} These racial differences are related to poverty rather than race—if the poverty level is factored out, the comparative increase among African Americans disappears.^{15 - 16} The US homicide rate for 0- to 14-year-olds is 5 times the combined rate of 25 other nations with comparable incomes and populations,¹⁷ and firearms rank second as a cause of injury death in children aged 10 to 19 years.¹⁸ The overall firearm-related death rate among US children younger than 15 years old is almost 12 times higher than the combined rate of the other 25 countries surveyed.¹⁷

Many children are at risk for observing a violent event during their childhood. One urban survey reported that 88% of urban adolescents and 57% of suburban adolescents had witnessed an assault, shooting, stabbing, robbery, or murder.¹⁹ Research has demonstrated that exposure to media violence^{6 ,20} and more violent forms of discipline^{21 - 22} are associated with subsequent violent and antisocial behavior in children. In addition, the availability of guns is associated with an increased risk of homicide and suicide in the home.^{23 - 24} Cummings et al²⁵ demonstrated a decline in unintentional shooting deaths in children less than 15 years old in states with gun safe storage laws. By teaching parents to limit media exposure, use nonviolent forms of discipline, and avoid or safely store firearms, one can hope to reduce the risk of violent injury among children and teenagers.

The purpose of this program is to teach pediatric residents how to discuss these issues with families, with the goal of decreasing future violent injuries in these children. An important barrier to discussion of sensitive topics is the physician's discomfort; we designed and implemented this program to begin to improve residents' comfort with the issue of violence. To do this, we focused on why this is an important issue for all families and provided practical, concrete suggestions for use in the primary care setting. The program sought to provide the residents with skills to implement realistic change and to teach them that small steps in a variety of areas can have an impact on violence prevention. The residents were able to practice these ideas using specially designed role-playing scenarios. Other teaching efforts in our residency program, such as mock codes in the emergency department and adolescent scenarios in the adolescent clinic, use role-playing scenarios as a teaching modality. Hopefully, the residents' familarity with this concept enhanced its usefulness during our training session.

Our results indicate that pediatric residents discussed violence issues more frequently after the educational program, and the improvement was sustained 6 months after the program. This last result surpassed the usual expectations surrounding behavioral change—that change is made immediately after an educational intervention, but that the learned behavior decreases in frequency without benefit of a booster intervention. Violence was formally addressed only once within the residency program; however, subsequent reinforcement of the general importance of anticipatory guidance could have increased VPG. In addition, the use of the written guide could have provided a small booster effect if the resident had postponed reading the guide until a few months after distribution. The natural maturation of the physician and the physician-patient relationship probably does not explain the increase in VPG, since there were no similar increases for other topics during the study period. The inclusion of these topics in the survey also served to disguise our interest in VPG from the parents and guardians responding to the survey.

Violence was discussed more frequently with older patients, possibly due to the perception that violent injuries occur more often in older children and that the effect of educating them is more immediate and pressing. Interestingly, prior to the educational program, violence was discussed more often with new patients than with established patients. New patients are assigned longer time slots, and residents may be able to discuss more anticipatory guidance topics during that time. The significant postintervention increase in violence guidance for established patients may reflect the program's educational goal of addressing this topic in a short period.

Overall, the residents felt the program was informative and useful. The positive response may have been due in part to the fact that one of their peers (C.D.J.) helped design and implement the program.

There are several limitations to our study. Our small sample size decreases the power to detect small differences within a stratified analysis, thereby limiting our ability to look for a sustained improvement in any of the subgroups analyzed. The data may have been influenced by recall bias on the part of the families. It is also possible that caretakers may have underestimated the amount of violence prevention offered because they may not have recognized when the pediatrician was indirectly addressing the issue of violence. For example, discussion of topics such as media exposure that does not include the word "violence" may still have direct bearing on the issue, particularly in younger patients. Finally, adjustments in our program may be necessary to stress the importance of addressing violence-related issues with younger children or to more appropriately target rural parts of the country.

Although the prevalence of violence in our neighborhoods is alarming, we must not accept it as inevitable. Currently there are very few active programs teaching health professionals how to address violence in the primary care setting. This is only one of several areas in which physicians can make a difference. We can educate ourselves, other health care workers, and families regarding violence-related issues and prevention strategies. We can advocate gun control and gun safety on the local and national levels. In addition, we can assist in the implementation of mentoring programs and community programs that provide children with activities to help keep them off the streets. We can advocate in the present to make our youth believe in their future.

Despite the demonstrated increase in addressing violence and gun safety, residents in our hospital still discussed it far less than the scope of the problem merits. Larger advocacy efforts should enhance trainees' recognition that to affect the health and safety of our patients, violence must be addressed routinely and repeatedly. This project illustrates that even a brief, directed violence education program can provide future pediatric physicians with the motivation and skills to discuss violence prevention with their patients and families. Sustained educational efforts are needed to increase skills and comfort with violence intervention—so that ultimately it will be routine in all pediatric offices.

Accepted for publication October 15, 1998.

Corresponding author: Joel A. Fein, MD, Division of Emergency Medicine, The Children's Hospital of Philadelphia, 34th Street and Civic Center Boulevard, Philadelphia, PA 19104 (e-mail: fein@email.chop.edu).

Editor's Note: Can someone explain why 80% of the residents thought this was a useful program and 93% said they'd use these skills in the future, but only 19% of the subsequent visits included these skills? Banking knowledge is not like banking money; you'll lose interest.—Catherine D. DeAngelis, MD