National implementation of a trauma-informed intervention for intimate partner violence in the Department of Veterans Affairs: first year outcomes

Over the past decade numerous research studies have indicated that high rates of intimate partner violence (IPV) among U.S. military veterans may convey risk for physical and mental health problems, as well as social, occupational, and legal difficulties [1, 2]. Women veterans are at high risk for experiencing IPV compared to their civilian counterparts [3], and male veterans with mental health disorders, particularly posttraumatic stress disorder (PTSD), evidence high rates of IPV use compared to both civilians and other veterans who do not have mental health disorders [4]. In response to this issue, in 2012 the Department of Veterans Affairs (VA) convened a Domestic Violence (DV)/IPV task force to develop recommendations for a national program. One year later, the task force finalized 14 recommendations to expand screening, prevention, and intervention for women and men veterans, as well as to introduce a VA employee assistance program for employees experiencing IPV [5]. The recommendations also included adopting non-stigmatizing language, specifically “IPV use” instead of IPV perpetration and “IPV experience” instead of IPV victimization. This program is relatively unique among healthcare systems because it places special emphasis on providing treatment to those who may be using violence or aggression in their relationships. Specific to veterans who have used IPV in their relationships, one task force recommendation in particular was that the VA implement a pilot of the Strength at Home program (SAH).

Prior to the implementation of the VA DV/IPV taskforce recommendations described above, few VA hospitals offered any programming for veterans using IPV beyond general anger management. This manuscript describes program outcomes from the first year of a pilot training program to introduce SAH at 10 VA hospitals. SAH is a 12-week cognitive-behavioral and trauma-informed group treatment to end IPV in military and veteran populations. Based on the social-information processing model of IPV in military populations [6], SAH has led to reductions in the use of physical and psychological aggression in pilot, effectiveness, and efficacy trials [7–9]. Importantly, until the trial of SAH [9], no prior intervention had demonstrated efficacy in reducing IPV in a military or veteran sample in a randomized clinical trial [10]. In fact, randomized clinical trials of IPV interventions have been rare, and results from civilian samples have shown these programs to have very small benefits in reducing IPV relative to legal system monitoring alone [11, 12]. Treatment programs that are most commonly available, such as those viewing IPV exclusively as a gender-based expression of power and control [13], depart substantially from the SAH model, in part because they do not account for the influence of trauma exposure, do not programmatically address core themes linking trauma to relationship problems and abuse, and often discount the importance of mental health symptoms commonly linked with IPV [6, 14]. Trauma and trauma-related consequences such as PTSD are consistently among the strongest risk factors for use of IPV in both civilian and military populations, and are therefore important considerations in the treatment and prevention of IPV [4]. As noted in a recent review paper authored by 17 leading IPV treatment experts, many existing interventions are not grounded in empirical research evidence or best practices [15].

In light of the above considerations, SAH was developed with the goal of establishing a model IPV treatment program for military and veteran populations. SAH consists of twelve 120-min group sessions organized into four treatment phases designed to address deficits in social information processing that increase risk for IPV: 1) psychoeducation on IPV and common reactions to trauma; 2) conflict management skills; 3) coping strategies and negative thought patterns; and 4) communication skills. All treatment components use military and veteran relevant language and examples. Practice assignments are completed at home. For example, veterans complete anger analysis worksheets designed to facilitate early recognition of anger signs and cognitive distortions, as well as the use of new skills and strategies learned in group. Group sessions start with a review of the practice assignment completed by each group member with feedback and discussion provided by the group members. This is followed by the introduction of new skills.

Prior to enrollment in SAH, all veterans complete an intake assessment, during which the assessing clinician prioritizes rapport building and uses motivational interviewing skills to enhance readiness for treatment [6]. During that intake, veterans complete measures of recent IPV, PTSD symptoms, and alcohol misuse. Due to high co-morbidity with both PTSD and IPV, alcohol misuse is included to help clinicians assess whether this factor may contribute to the veteran’s use of IPV or may complicate treatment efforts [16, 17]. Although SAH was not designed as a treatment for PTSD, the program addresses the impacts of trauma and core themes underlying trauma that may relate to IPV in relationships. Therefore measures of PTSD symptoms are also included.

A key component of SAH is partner outreach to ensure that any affected partners receive hotline numbers and information about other resources, safety planning, and the opportunity to complete a telephone assessment regarding their IPV experiences. Partner outreach is an important component of IPV treatment programs because partners may not have had the opportunity to reach out to available sources of aid and support, and they may not be aware of the resources available to them in the community. Research has suggested this direct outreach is an often overlooked opportunity to support those who have experienced IPV [18].

Using anonymous program evaluation data from veterans who completed the program at 10 VA hospitals, this manuscript describes patient outcomes from the first year of the VA pilot roll-out of SAH. We also present information on our initial approach to implementing this trauma-informed treatment for IPV. Specifically, we report on clinician success in the training program (number of clinicians facilitating groups within the first year of the program). We report on the number of veterans initially enrolled in the program and their demographic information. We also examine changes in the number of different types of IPV used, changes in the prevalence of physical and psychological IPV, and changes in both PTSD symptoms and alcohol misuse among those veterans who attended the majority of the program sessions (8 or more). We also examined veteran satisfaction with the treatment.

This manuscript describes the format, structure, and initial therapist and patient outcomes of the first year of a pilot rollout of SAH at 10 VA hospitals. Readers should evaluate the potential causal effect of SAH on patient outcomes in the context of a naturalistic program evaluation of veterans who attended at least some SAH sessions and completed a post-treatment assessment. Significant reductions in the number of types of IPV used and changes in the proportion of the sample who used physical and psychological IPV toward their partners suggest the training model and initial implementation efforts were effective. Moreover, veterans were satisfied with the treatment they received. Reductions in PTSD symptoms from pre to post-treatment also suggest that the program may be helpful in reducing PTSD-associated distress and symptomatology. With regard to therapist outcomes, 70% of sites and 34% of clinicians were successful in launching the program. Combined with the reductions in IPV and PTSD symptoms observed in veterans, this suggests that the training program was successful overall.

The average length of time between receipt of the training and facilitating the first group was longer than desired and there were three sites that did not successfully implement the program within the first year. This suggests a need for improved attention to the implementation strategy, and in particular how to best ensure institutional support. Prior research has shown that an important facilitator of successful VA implementation is the alignment of resources to support delivery of the new treatment [31]. For example, starting a new program requires substantial infrastructure support such as identifying space to hold a group, identification of billing codes, and application for clinic status in the official medical record. In addition, the availability of the new program must be advertised to referring providers, who in turn, require information on the program and how to refer veterans to the program. If this work begins only after receiving the training, as was the case at some sites, the time to start the first group can be considerably delayed, underscoring the need to initiate implementation support and outreach efforts prior to the in-person training.

The number of clinicians successfully completing the training program was also lower than desired and there was a high rate of clinician dropout, primarily due to change in position and workload. Collectively this suggests the need for enhanced efforts to improve institutional and leadership commitment to offering the program by protecting clinician time to participate and improving leadership understanding of the amount of time required. This is a similar finding to those from the early stages of other VA rollouts of evidence-based treatments [32]. Specifically, in the early stages of the VA rollouts of Cognitive Processing Therapy and Prolonged Exposure, there were lower rates of training completion when clinician trainee requirements were less defined. An application process with leadership concurrence was introduced and this had a positive and significant impact on adoption of the therapy [32].

In response to these challenges and drawing from the literature describing successful implementation of other treatments within VA, the implementation strategy was improved for a second cohort of seven sites currently under way. First, we are attempting to enhance commitment from the local sites by requiring sites and clinicians to apply to receive the training. This provides an opt-in process that may exclude sites that do not have institutional support, and increases the likelihood that site leadership members understand the staff and resource requirements to successfully implement the program. Second, we now offer pre-training calls to sites and set requirements regarding activities that need to be completed prior to training. This requirement is intended to help them conduct the pre-implementation work needed to shorten the time between training and the first group session. Third, we specifically encourage sites to begin outreach efforts prior to training. Fourth, we now emphasize team-based implementation in which all clinicians assist with outreach and referrals.

The observational research design of program evaluation imposes important limitations on the interpretation of findings. The lack of a control group limits our ability to account for regression to the mean in IPV and PTSD symptoms. The naturalistic process of assigning patients to groups may increase effects by excluding patients who are less likely to benefit from the treatment, and non-randomized assignment precludes an intent-to-treat analysis. However, a previous randomized controlled trial [9] demonstrated that SAH was more effective than enhanced VA treatment-as-usual (ETAU) and the current results are consistent with that earlier research. Second, the outcome measures were limited in scope, most notably because we obtained only prevalence and not frequency counts for specific acts of IPV. However, the summary scores used in the current study are closely modeled after a widely used scoring method for measures of IPV, labeled “variety scoring,” which involves dichotomizing reports for each specific act of IPV and then summing the number (or variety) of different aggressive behaviors used [33]. Third, the amount of data available from partners, who are generally considered to provide the most accurate reports of IPV, was quite limited. Finally, all participant data for the program evaluation came from men. Likewise, only male participants were included in the pilot and efficacy studies of SAH. Further research is needed on program efficacy with women.

The observed effect size for pre- to post-treatment reduction in IPV in the current evaluation (g_rm = .56) is reasonable compared to a prior randomized clinical trial of SAH. The current effect size is larger than the effect size found in the prior research control condition (Hedge’s g = .36 and .26 for physical and psychological IPV, respectively), but smaller than the prior research treatment condition (Hedge’s g = 1.02 and .84 for physical and psychological IPV, respectively). Some gap in efficacy is to be expected in initial broad-scale implementation as a result of clinician expertise, experience with the population and intervention, and treatment integrity. However, methodological factors likely contributed here as well. Most notably, the selection criteria for the original controlled trial required all participants to have engaged in recent physical IPV, whereas the implementation sample included participants with less recent IPV or only psychological IPV, thus reducing the potential magnitude of observed pre- to post-treatment reduction in physical IPV. In addition, the controlled trial also used more precise measures of IPV frequency. Importantly, in the implementation sample, the prevalence of any physical IPV at post-treatment for those who reported physical IPV at baseline (22%) is essentially identical to rate reported for SAH in the controlled trial (23%), and considerably lower than the rate observed in the treatment-as-usual control (43%).

Preliminary results from the implementation of SAH suggest that the intervention can be effectively delivered in real-world clinical practice at the VA. The observed reductions in IPV use in this implementation evaluation were consistent with the reductions observed in a highly controlled clinical efficacy trial [9]. Given data indicating other IPV treatments have minimal effects on repeat IPV above and beyond arrest and legal system monitoring, this is a very promising outcome [10–12]. However, delays between the completion of training and the start of groups, as well as clinician and site drop-out were limitations. Results identified four opportunities to strengthen the implementation strategy in order to increase the impact of the intervention. These strategies are being evaluated in a second cohort of seven sites.