1. Introduction
The World Health Organization defines intimate partner violence (IPV) as “behavior by an intimate partner or ex-partner that causes physical, sexual or psychological harm, including physical aggression, sexual coercion, psychological abuse, and controlling behaviors” [
1]. IPV transcends all geographical and cultural boundaries and is found worldwide [
2]. In the United States, IPV is considered a major public health concern, with uniform reporting guidelines at the state and national levels [
3].
Victims of IPV include both sexes, although reports worldwide and in the US indicate females have a higher rate [
1,
4]. A Centers for Disease Control and Prevention report indicated that at some point in their life, “1 in 2 women and 1 in 4 men suffered from contact sexual violence, physical violence, and/or stalking victimization in the hands of an intimate partner” [
4]. Worldwide, 1 in 3 women has suffered from IPV [
1].
IPV may have a diverse impact on a victim’s health, including sexually transmitted diseases, depression, post-traumatic stress disorder, and many chronic illnesses [
5]. Additionally, it has been linked to structural changes in the DNA of victims’ offspring [
6].
Due to its common occurrence, there is a high likelihood that healthcare professionals, physicians especially, will encounter and treat those affected by IPV. Various studies have looked at screening rates across different healthcare settings. Screening rates by a healthcare professional in an emergency department or primary care clinic are estimated to be 10–12% [
7].
Resident physicians in community hospital-based residency training programs are in a unique setting to identify, assist, and manage IPV victims. However, physician knowledge, level of comfort, screening, and skills for managing IPV patients are deficient [
8]. A survey of medical students and primary care physicians revealed that half never talked about IPV to their patients, and there is a need to improve training [
9].
In general, physician education regarding domestic violence recognition and intervention is lacking in many hospital systems. A recent scoping review found 57% of articles were mostly for medical students, with lecture and standardized parents as the common forms of training [
10]. Most training sessions are delivered using lectures and standardized patients, and a standard curriculum is lacking [
10]. Many of the current studies are in academic health centers or specialized clinics [
10].
A few recent studies beyond the scoping review included a two-part, 2.5 h IPV training program for IM residents at an academic health center that improved overall knowledge, confidence, comfort, and preparedness when measured immediately after training [
11]. One limitation mentioned was that they did not measure the training’s effect longitudinally. A multi-site study in Canada and the US focused on educating healthcare workers in seven academic fracture clinics found that a year after a 2 h training, significant improvements were attained [
12,
13]. The specialized clinics utilized a train-the-trainer model and an educational intervention built upon prior curricula developed by IPV researchers, surgeons, and psychologists initially created for orthopedic surgery trainees. In another study, a half-day training for orthopedic trainees in an academic setting included lectures on foundational IPV knowledge and recent research outcomes [
14]. It found the training to be effective after 3 months. A study on Greek physicians and trainees indicated that results were not statistically significant one year after IPV training [
15].
Our hospital embarked on improving domestic violence screening and recognition starting in 2015. We have partnered with Turning Point, a domestic violence shelter that provides outreach and helpful resources to victims. To our knowledge, we have not found a study that looked at the longitudinal effect of an IPV training program provided by a community domestic violence shelter expert on community hospital-based residency training programs. The primary objective of this retrospective study was to analyze first-year or Program Graduate Year One (PGY1) residents’ perceptions, practice, and knowledge regarding intimate partner violence before and after they complete the community domestic violence expert’s educational program.
2. Materials and Methods
2.1. Study Design
This is a retrospective cohort study. The study was approved by McLaren Health Care’s ethics committee (SARC 4.23.2019). All residents at this community-based hospital were included in the training sessions. Completely de-identified, matched pre- and post-data from Program Graduate Year 1 (PGY1) resident physicians’ participants were retrospectively accessed. A well-studied and established, validated tool, Physician Readiness to Manage Intimate Partner Violence Survey (PREMIS), was used to assess changes [
16].
The PREMIS tool was administered pre-, post immediate, and one year post the IPV educational program implemented from July 2015 to July 2017. The study analyzed PGY1 residents’ knowledge and comfort level in discussing and addressing domestic violence with patients both before and after their educational intervention.
2.2. Population and Sample Size
A total of 20 matched cases were analyzed. Pre- and post-questionnaire responses of all participating residents at McLaren Macomb who completed training between 2015 and 2017 were included. This study employed convenience sampling of available complete pre- and post-data. Only de-identified, anonymized pre- and post-training data were accessed for evaluation. Due to the small sample size, we did not collect demographic other than gender or programmatic data to ensure the anonymity of the respondents.
2.3. Description of Educational Program and Assessment Tool
This curriculum was delivered and adapted for our community by one of the co-authors of this study Carmen E. Wargel, LMSW., a social worker at the Turning Point domestic violence shelter who has IPV expertise. Turning Point is McLaren Macomb Hospital’s county domestic and sexual violence provider. The training utilized the Project Connect curriculum called CUES, created by Futures Without Violence and the United States Office on Violence Against Women [
17,
18]. CUES stands for Confidentiality, Universal Education + Empowerment, and Support. It was unique in that it was solely delivered by our social worker IPV expert, and no physicians were part of the instruction. The training sessions delivered information about domestic and sexual violence, as well as specific best practices for confidentiality, universal education, referral, intervention, and harm reduction for patients. To our knowledge, there are no other studies utilizing this curriculum.
In the different sections of the training, participants worked on nine learning objectives across three main domains:
Recognizing and Responding to Domestic and Sexual Violence in a Healthcare Setting
How domestic and sexual violence impact the health of their patients
Four best practice interventions with patients on the topic of domestic and sexual violence
Why healthcare providers should discuss domestic and sexual violence with their patients
Confidentiality, Universal Education, Handling Disclosures, and Supported Referrals
Sharing the limits of their confidentiality with patients
Universal education on domestic and sexual violence
Best practices for handling disclosure of domestic and sexual violence from patients
A supported referral for patients experiencing domestic or sexual violence
Harm Reduction, Targeted Interventions, and a Survivor Speaker
Describe and demonstrate the use of harm reduction strategies with survivors of domestic and sexual violence
Describe and demonstrate the use of targeted interventions with survivors of domestic and sexual violence
Describe real-life interactions between survivors of domestic and sexual violence with healthcare providers
PGY1 residents completed an average of 5 h of training. The curriculum began with an introduction to the four evidence-based interventions (confidentiality, supported referrals, harm reduction, and targeted interventions). Educational materials were provided regarding the lethality of IPV, what to say to a survivor, harm reduction, and targeted interventions. One session addressed how to maintain confidentiality using best practices and how to make a survivor-centered mandated report. As an experienced educator on domestic and sexual violence, C.E.W. knew that information was vital but not sufficient to create change. Each skill was reinforced with role plays on evaluating an IPV survivor so that residents could develop their own language reflecting each concept and have an opportunity to practice it out loud. The training also featured videos and worksheet guides to reinforce the concepts of the previous lessons. A guest survivor speaker spoke about her personal experiences with many medical providers during and after being abused, providing valuable insight into a survivor’s perspective. During the final lessons, residents observed videos and learned the concept of targeted interventions and risks survivors face, including sexually transmitted infections, strangulation, and traumatic brain injury.
The evaluation utilized the PREMIS tool, a standardized, validated, and publicly available measure of physician knowledge and readiness and the educational program’s effectiveness in IPV training [
16]. It is a 15-min survey that allows a comprehensive evaluation in four main areas: background knowledge, preparedness, opinions, and practice issues.
During the 3 year period, we adjusted the training schedule and the number of sessions. These changes proved helpful in maximizing resident attendance and attention. However, the content and practice stayed largely the same, relying heavily on a local survivor speaker and experienced educator, as well as Futures Without Violence videos, framing, and evidence.
2.4. Outcome Measures and Statistical Analyses
Outcomes were measured at three points: prior to the educational intervention (Pre), immediately after the intervention (Post immediate), and one year after the intervention (Post one year) using the PREMIS tool. Primary outcomes are the different domains of the PREMIS tool administered pre- and post-IPV education programs. Composite scores for Background, Actual knowledge, and Opinions defined by the PREMIS tool were generated and evaluated. In addition, individual item scores also were generated to provide feedback for continuous improvement. For Practice Issues, the composite score was not calculated due to missing data for several items. However, individual scores for those with enough data were generated.
Perceived preparation was measured by asking the residents their level of preparedness on a 7-point scale (1 = Not prepared; 2 = Minimally prepared; 3 = Slightly prepared; 4 = Moderately prepared; 5 = Fairly well prepared; 6 = Well prepared; 7 = Quite well prepared). Perceived knowledge was measured by asking residents how much they feel they know using a 7-point scale (1 = Nothing; 2 = Very Little; 3 = A little; 4 = A moderate amount; 5 = A fair amount; 6 = Quite a bit; 7 = Very Much). Opinions were measured by asking residents their responses to statements using a 7-point scale (1 = Strongly disagree; 3 = Disagree; 5 = Agree; 7 = Strongly agree).
Descriptive data such as percentages, frequencies, total and mean scores, and standard deviation were generated. A repeated measures ANOVA was performed on continuous variables to determine the presence of statistical differences. Pairwise comparisons with Bonferroni correction were performed on variables that were found to be significant. In the results tables, mean separation was represented by a letter, whereby means sharing the same letter are not statistically significant. Statistical significance was set at p < 0.05. IBM SPSS Statistics for Windows, Version 25.0 (Armonk, NY, USA) was used to analyze these data.
3. Results
Seven programs were represented during the training: Emergency Medicine, Family Medicine, Internal Medicine, Radiology, Orthopedic Surgery, Obstetrics and Gynecology (OB-GYN), and General Surgery. Of the participants, 61% were male.
Prior to their IPV training at McLaren Macomb Hospital, residents recorded an average of 1.39 h of IPV training. Immediately post-training, the average increased to 5.21 h, and one year later, had increased to 5.79 h (p = 0.0002).
Table 1 shows the composite scores for each of the main PREMIS tool categories. Under Background, Perceived preparation (
p = 0.0001) and Perceived knowledge of IPV management (
p = 0.0001) for post-mean scores were statistically significantly higher than pre-scores. While the Post one year mean scores were lower, these were not statistically significantly different than the post immediate scores. Actual knowledge of the IPV total score increased by a statistically significant amount post immediate; however, it decreased to a level Post one year statistically significantly lower than pre-intervention. Individual items for these subcategories will be discussed in detail.
Under the Opinions category, Preparation (p = 0.058), Workplace issues (p = 0.11), Alcohol and drugs (0.57), and Victim understanding (p = 0.21), mean scores were not statistically significantly different before and after the intervention. Means scores ranged between neither disagree nor agree (4) and agree (5). Since the composite scores for these subcategories were not statistically significant, we did not perform a per-item test for significance. However, it is important to note that Post immediate and Post one year mean scores were higher than Pre for both Preparation and Workplace issues but not for Alcohol/drugs. Legal requirements and Self-efficacy were statistically significantly higher post-intervention, although there is no statistically significant difference between Post immediate and Post one year.
3.1. Background
Table 2 shows the scores for the individual items for the Perceived Preparation subcategory. Scores for all the individual items reflect the trends of the composite score in
Table 1. Overall, an increase in score was observed Post immediate, which then decreased slightly Post one year. Participants progressed from feeling ‘slightly’ or ‘moderately prepared’ to ‘fairly well’ or ‘well prepared’ immediately post-IPV training. The only exception to this trend was in conducting a safety assessment of the victim’s children, where the mean returned to the ‘moderately prepared’ region, a change that was statistically significant. Participants also believed they were less able to help an IPV victim create a safety plan than they were one year prior.
Table 3 shows the scores for the Perceived knowledge subcategory. Scores for all the individual items reflect the trends of the composite score in
Table 1. Residents initially reported perceived knowledge prior to IPV training in the ‘very little’ to ‘moderate amount’ range. After formal IPV training, their perceived knowledge improved into the statistically significant range of ‘a fair amount’ to ‘quite a bit.’ One year later, their overall perceived knowledge persisted in a statistically significant trend in the ‘moderate amount’ to ‘quite a bit’ range. An exception, such as their perceived preparation, was a statistically significant decrease in developing a safety plan with an IPV victim from Post immediate to Post one year.
The percentage of correct responses for Actual knowledge of individual items is shown in
Table 4. Of the 34 questions, about 70% (24) started with higher than 70% correct responses and stayed high. This indicated that although residents did not report a high number of IPV training hours pre-intervention, they were knowledgeable about the basics of IPV. The question about the strongest single risk factor for becoming a victim of intimate partner violence, which is being female, was a challenging question for participants. It started with a very low score, improved Post immediate but decreased again Post one year. The rest of the questions saw improvements Post immediate, but scores declined Post one year.
3.2. Opinions
Table 5 shows results for Legal requirements and Self-efficacy. Under Legal requirements, both Post immediate and Post one year training, residents significantly improved their awareness of legal requirements for reporting suspected IPV, child abuse, and elder abuse. Under Self-efficacy, immediately post-training, residents were better at asking new patients about abuse in their relationships. One year later, there was a slight decline in this response, though not statistically significant. Immediately post and one-year post training, residents agreed they increased their comfort level in discussing IPV with patients, were more able to detect IPV without asking a patient about it, and were better able to garner information to determine IPV as the reason for a patient’s illnesses. When it comes to matching therapeutic interventions to an IPV patient’s willingness to change, the gains post-training were lost after one year. No statistical significance due to the training was found with the residents’ ability to recognize victims of IPV by the way they behave.
3.3. Practice Issues
One practice issue question was centered on asking patients about the possibility of IPV when presenting with symptoms such as injuries, chronic pelvic pain, or eating disorders. The residents improved from a mean score of 1.52 (0.58) to 2.51 (0.79) Post immediate and 2.53 (0.52) Post one year (p = 0.0001).
Compared to a pre-intervention rate of 90% who did not identify IPV in the past 6 months, Post immediate and Post one year, this rate decreased to 60%. For those who identified IPV in the past 6 months, pre-intervention, 10% of respondents provided information to and counseled the patients. In addition to these actions, in both the post and Post one year intervention time periods, residents who identified IPV in the past 6 months also indicated counseling patients about options, conducting a safety assessment for the victim and the victim’s children, and helping the patient develop a safety plan.
Pre-intervention, 79% were unsure, and 21% indicated there was a protocol for dealing with adult IPV at the clinic or practice, compared to post-intervention responses of 35% and 65%, respectively. Post one year of the intervention, the response was 41.2% and 41.1%, respectively, in addition to 17.7% who responded that there was no protocol in place.
On familiarity with institutional policies regarding screening and management of IPV victims, pre-intervention, 15% answered affirmatively. This number increased to 75% in both the post and Post one year periods.
Regarding whether they are in practice in a state where it is legally mandated to report IPV cases involving competent adults, the percentage of respondents who indicated Yes was 25% (Pre), 65% (Post immediate), and 40% (Post one year).
Figure 1 shows the percentage of new IPV diagnoses. Pre-intervention, only 6% of respondents indicated 1–5 new diagnoses, and more than 90% indicated no new IPV diagnosis. Post immediate and Post one year of the intervention, there were 60% and 65% new diagnoses, respectively, for both 1–5 and 6–10 new diagnoses, and the number of those reporting no new diagnoses decreased to 40% and 45%, respectively.
The number of residents who screened increased post-intervention (
Figure 2). Post-intervention, training participants demonstrated screening of all new female patients and patients periodically and during an annual exam and those with abuse indicators.
Table 6 indicates an increase in mean responses post and one-year post. The exception is for hypertension which, while it did not significantly increase Post immediate, it increased significantly one year post, compared to pre-training.
4. Discussion
The IPV training’s biggest impact is on Perceived preparation, Perceived knowledge, Legal requirements, and Self-efficacy. In addition, increased new IPV diagnoses and the number of residents screened were also improved due to the training. Opinions on alcohol/drugs were not impacted, while the rest of the opinions increased but were not statistically significant. The residents’ performed well on the actual knowledge questions even before the training on about 70% of the questions. The challenge was in retaining performance on several questions Post one year. The heavy work burden of residents’ patients and other educational responsibilities could be a reason and points to the need for ongoing training in this area to reinforce the concepts.
A systems model approach is a more coordinated and comprehensive structure and process that link hospitals, clinics, physicians, and other healthcare providers with community resources such as hotlines, domestic violence shelters, and support group [
19,
20]. This approach has been recommended as one solution to the IPV problem [
19,
20]. Our partnership with our local domestic violence shelter, Turning Point, is one example of this model to increase knowledge, awareness, and action to alleviate IPV in our community. This partnership has also led to our hospital using the Futures Without Violence curriculum. Futures Without Violence is a nonprofit organization, and Project Connect was a federally funded initiative that involved a comprehensive approach from victim identification, linking to resources, and prevention [
18]. Nationwide, Project Connect has impacted greater than 400,000 patients through training more than 7000 healthcare providers in over 80 clinical settings [
18].
While most curricula found in a recent scoping review were taught to Family Medicine, Internal Medicine, and Emergency Medicine [
10], we included Radiology, Orthopedic Surgery, OB-GYN, and General Surgery in our training. As a result of the broad range of practice settings residents in a community hospital-based residency program encounter, the training was applied to multiple real-life practice situations, which provided another layer of learning. For example, in the office setting, residents were trained to separate the patient suspected to be a victim of IPV from any other person in the room in a non-threatening manner, such as by asking them to provide a urine sample in the bathroom. Concomitantly, the literature would have previously been placed in the bathroom to allow the patient to signal to the office staff that they were in need of assistance. In an alternative setting example, a trauma victim can be provided the literature in a safe manner by providing them with a list of community resources in a manner that can be easily concealed from their abuser as they exit the hospital. In any setting, multiple factors may prohibit a patient from immediately leaving their abuser. Therefore, residents were trained to develop safety plans with the patient with the expectation of eventually delivering the patient to safety.
Several studies are of interest when comparing approaches and outcomes to our study. Similar to our study, the EDUCATE study evaluated the residents one year after the training [
12,
14]. It used a train-the-trainer model where one lead from each of the specialized fracture clinics was trained, who then, in turn, trained everyone at the site. The participants included not only orthopedic surgeons and surgical trainees but also nonphysician health care professionals and research and administrative staff. Our approach was focused on first-year residents in a multidisciplinary model. Similar to ours, they showed significant gains immediately post-training and one year after [
12,
13]. Their studies also showed the same trends, where immediately after training, there was a statistically significant increase in scores, but one year after, the scores decreased, albeit not statistically significantly, from the immediate post-training period. Both models point to the longevity of the gains from training but also indicate that refresher courses to bolster these gains against erosion could be beneficial.
In another study, training was delivered by two physicians who have expertise in IPV research with patients and curriculum development [
11]. Their training was also delivered only to first-year Internal Medicine residents. They also used the experience of a local women’s shelter during training, but not to the extent conducted here. The direct insight and expertise of the Turning Point trainer provided our residents with up-to-date information and recommendations on how to interact with these victims and offer effective care.
The length of training delivery is another aspect worth noting. One study involved the delivery of a 45 min IPV training by a community IPV trainer from a domestic abuse referral source [
21]. Participants involved family medicine physicians, residents, midwives, and nurse practitioners at multiple training sites of a community-based urban family medicine residency in Chicago [
21]. However, the results were mixed. On the other end of the spectrum, in a study involving Greek General Practitioners (GP), a 9 h IPV training was delivered by a team comprised of a GP, nurse, psychologist, social worker, lawyer, and administrators from an abused shelter [
15]. Results were compared to a control group with no training. The training did not result in a statistically significant improvement for residents.
As documented in the results, residents improved their overall number of hours in IPV training. One year after their training at McLaren Macomb Hospital, the total number of hours reported increased, suggesting they had received additional training. This additional training, while not documented as part of this study, likely occurred in the form of additional reading and/or formal training on the resident’s behalf.
While the training resulted in very positive results, certain areas of improvement were also uncovered. In both Perceived preparation and Perceived knowledge, residents showed a decreased overall comfort level in helping an IPV victim create and develop a safety plan. This suggests that further training specific to this aspect would be beneficial for the long-term retention of knowledge. Victim understanding showed insignificant change among the residents evaluated and is another area where training can be targeted in the future. Having the necessary skills to discuss the abuse with an IPV victim who is female, male, or a different cultural ethnicity continued to be an area of improvement after training.
Continuous improvement and education are also needed to increase and sustain a higher screening rate. One example is the Veterans Health Administration (VHA) experience. In 2014 IPV screening was rolled out in Veterans Affairs medical centers (VAMCs) [
22]. Through this policy, over half of VAMCs had adopted IPV screening by Fiscal Year 2020, an exponential increase from one VAMC in Fiscal Year 2014 [
22].
The training provided many practical aspects. Through a series of educational seminars, role-playing scenarios, survivor testimony, and education from crisis center social workers and more experienced physicians, residents learned to screen for, identify, and provide an appropriate response to a given situation. Many practical tips were discussed. The relatively small size and specialty diversity of the first-year residency class at McLaren was a great benefit. Having clinical faculty in the room and directly involved in conversations about the lessons also strengthened its importance. The other benefit is that while some specialties are more likely than others to embrace this topic, the multi-specialty participation meant that the needs of survivors across our hospital system would be met. In a larger hospital setting, a single specialty residency could replicate this model by emphasizing the importance of domestic and sexual violence by centering it in lectures, engaging teaching faculty, and modeling involvement from the highest levels.
The uniqueness of our project is partnering with our local domestic and sexual violence program to leverage their expertise and experiences with survivors in our community compared to a more usual training model where physicians may not have the day-to-day immersion in a survivor-serving setting. An area for future research and program improvement would be to pair a domestic and sexual violence advocate with a physician who is well-versed in this area. In this way, the physician could offer concrete examples from their own practice, could be a part of training other teaching faculty, and could model behaviors during clinical instruction on shifts.
Given the difficulty in managing IPV, pre-written scripts were provided to residents as a starting point for a conversation with the patient as well as to ensure no critical details were omitted. Focus is also provided on personal relationships to be developed as the patient exits a relationship involving IPV. When a referral to a community center is made, all local resources are provided, highlighting the name and office number of a social worker that the provider frequently works with so the patient can start a personal relationship with a social worker in the outpatient setting.
Other benefits of this training include improving the identification of IPV victims that may have otherwise been missed by insufficient provider training and increasing the comfort level of providers when dealing with emotionally fragile patients. This study provides information and a model for training future residents at community hospital-based residency programs.
There are limitations to this study. The small sample size, while beneficial for a more engaged discussion, is a limitation of the study from the analysis perspective. This precluded us from performing subgroup analysis. Future research should include a bigger sample size, potentially through a multi-center study. As this was a voluntary survey, missing values were unavoidable. In our setting, the attending physicians were not previously trained and therefore were unable to engage in the discussions fully. In a see one, do one, teach one model of learning in residency programs, this is a significant limitation to ongoing behavior integration. One recommendation is to provide training to attending physicians as part of the need for a more comprehensive approach to IPV education in residency programs. We also recommend further education and practicing communication skills with patients of different genders and cultural backgrounds. An example will be incorporating more small group sessions or workshops to have hands-on practice and role-playing to address these gaps.