Promoting Cross-Sector Collaborations to Address Intimate Partner Violence in Health Care Delivery Systems Using a Quality Assessment Tool

Introduction

Violence against women has significant consequences for women's health and is a major public health threat. ^1,2 The Institute of Medicine in the United States (and the World Health Organization globally) recognizes the health care system, including the public health infrastructure, as a key site for identification of and support for survivors subjected to intimate partner violence (IPV; emotional, physical, and sexual abuse by an intimate partner), as well as for violence prevention. ^3,4 Examples of multisectoral collaboration in addressing IPV are limited, ^{5 –7} despite over two decades of public health literature, which has emphasized integration and collaboration across different sectors as critical for tackling complex public health problems. ^{8 –16}

Health care delivery systems are particularly critical sites for addressing IPV. Those who experience IPV are frequent users of the health care system because of the profound impact on health, although they may not recognize that their health concerns are associated with their experiences of abuse. ^17,18 Exposure to violence can lead to costly, chronic health conditions, increased utilization of emergency and inpatient services, and poor mental health. ¹⁹ Health care providers and staff are on the front lines for supporting patients who have experienced or are experiencing IPV and connecting patients to resources. Studies have indicated that when asked, women are willing to discuss abuse with trusted health care providers, although disclosure rates may be low. ²⁰ Through universal education (in which all patients are provided education and resources on IPV), routine inquiry, and brief counseling interventions, the health care setting can have a significant impact on increasing safety for survivors and preventing future abuse across a patient's life span. ^18,21

Since 2013, women's preventive health care services in the United States now include IPV screening and counseling. ²² This policy creates opportunities and incentives for both IPV victim service advocates and community-based primary health care providers to think creatively about strategic partnerships. Evaluations of innovative prevention and intervention programs are needed to transform the service delivery model. This will ensure that evidence-based resources are offered where they are needed, in both social service organizations and health care settings.

Addressing IPV in a health care setting is multifaceted and requires provider-level as well as system-level changes. System-level changes include trauma-sensitive policies and protocols at the clinic level, environments and training that facilitate care for survivors, and funding for organizations and providers serving IPV survivors. ^7,18,23 Without these clinic and health care delivery system changes and supports, it is difficult for providers to make and sustain changes in their interactions with patients. ²⁴ Providers report that the primary barriers to addressing IPV in clinical settings include concerns about time, lack of knowledge on how to start the conversation, and uncertainty about resources for IPV survivors. ²⁵ Some system-level changes that can help assuage these concerns are streamlined clear protocols for assessing for IPV, scripts, safety cards, and other resources for providers to utilize, as well as the knowledge of how to safely refer patients to the local victim service agency. ²⁶

In 2013, Futures Without Violence, a national nonprofit organization dedicated to ending violence against women and children, launched “Domestic Violence and Health Care Partnerships” (DVHCP), funded by the Blue Shield of California Foundation. ¹ The primary focus of this initiative was to promote collaboration between IPV advocacy agencies and health care delivery systems serving vulnerable and marginalized populations (called safety net health care providers) to improve systems of IPV care in California and to increase access to high-quality integrated health and IPV services for vulnerable populations. Through a competitive application process, 19 partnerships between a safety net health care provider and IPV agency committed to working collaboratively on policy and practice improvements within both the health care setting and IPV agency were selected. Health care providers and IPV advocates received educational materials about the impact of IPV on health and training, on how to communicate with their patients and clients about IPV, and available services. Using a comprehensive training program featuring the evidence-based CUES (Confidentiality, Universal Education and Empowerment, and Support) approach, health care providers and staff were encouraged to receive training from Futures Without Violence on the importance of providing universal education by offering information about IPV and relevant resources using an educational safety card to patients (with a particular emphasis on all adult women, all female-identifying patients including LGBTQIA+ patients, and all adolescents given the prevalence of IPV in these populations; see ipvhealth.org for examples of these materials). ¹⁸ These safety cards contained information on healthy relationships as well as national and local domestic violence resources. Leadership teams also participated in an ongoing learning community with bimonthly webinars and biannual in-person meetings to encourage colearning.

As part of a developmental evaluation of this initiative, sites also completed a systematic assessment of clinic procedures, policies, and protocols regarding IPV. The practice manager or clinic lead completed a Quality Assessment/Quality Improvement (QA/QI) tool every 6 months throughout the project to guide improvements in protocols related to IPV (such as confidentiality and how to respond to positive disclosures), staff trainings, support for staff related to secondary trauma, referral processes to the local domestic violence agency, and educational materials and resources related to IPV (including posters, handouts, and educational safety cards). This article describes the feasibility of implementing this QA/QI tool and the changes in clinic protocols and practices that occurred over time with the goal of demonstrating one potential strategy for guiding system-level changes and promoting cross-sector collaborations with victim service agencies to support a more robust response to IPV in clinical settings.

Materials and Methods

This study was approved by The University of Pittsburgh institutional review board. The DVHCP initiative selected 19 leadership teams that each consisted of administrative representatives from a safety net health care delivery system (mostly community health centers) and a victim service agency. Each health site completed the QA/QI tool at baseline and every 6 months following enrollment. Data were collected between 2014 and 2017; only those health sites (15 total) that completed at least two QA/QI tools were included in this description of the feasibility of this tool. Nine of the partnerships started in 2014, and six were added in 2015–2016. Ten sites completed three or more QA/QI tools. The time interval between baseline and end line varied with four sites completing the project in 1 year, two sites in 2 years, and two sites continuing for 3 years.

The QA/QI tool was designed with input from health care providers, IPV advocates, and public health practitioners, and modified from a clinical assessment tool developed by Futures Without Violence using a Delphi process. A practice manager and/or medical director at each health site completed the tool. A copy of the tool is available online at https://ipvhealthpartners.org.

The first “protocols” section of the tool included questions about clinic protocols specific to IPV as well as whether staff were provided with sample scripts to help guide patient interactions during IPV assessment. In addition, items about managing confidentiality asked about policies to see all patients alone at least for some portion of a health care visit and confidentiality in the electronic health record.

The “assessment methods” section inquired about how patients were assessed for IPV, including whether assessment occurred in a private space and whether they were offered information on healthy relationships and a palm-sized educational safety card with information on how violence can impact health.

To explore the referral process between health care providers and IPV agencies, the “documentation of assessment and response” section included questions about documentation of referrals to the clinic's partnering domestic violence agency or advocate and of provision of harm reduction strategies.

In the “intervention strategies” section, sites were asked how staff were supported in handling situations in which IPV was disclosed, including whether scripts were available for providers, whether someone was on-site (such as an advocate) who could provide support, whether providers had up-to-date contact information for the partnering domestic violence agency, and whether providers had lists of relevant resources to share with patients.

To learn more about the health care centers' networks within their communities, they were asked whether, within the last 2 years, representatives from IPV programs, child protective services, or law enforcement were contacted or had contacted them to either schedule or conduct a training. Sites were asked if internal resources were available to support staff if a patient disclosed IPV (such as whether staff who were survivors of IPV could “opt out” of treating a patient). Staff training was evaluated through two questions about training for new and existing staff.

Staff support was evaluated through three questions that asked about ways that staff could discuss concerns (including through individual clinical supervision) and whether the clinic had an employee assistance program. Three items assessed how clinics document IPV assessment and disclosures, as well as whether clinic protocols are reviewed annually. The final section of the tool included questions about resources available for patients to take and whether those resources were inclusive of various populations served by the health care delivery system.

Once a practice manager or medical director completed the tool, it was immediately sent to the University of Pittsburgh evaluation team. Data were analyzed using Stata statistical software (version 14.1). Reports were created for each site to inform them of their progress with making changes on each of these indicators.

Results

Fifteen of the 19 partnerships completed the QA/QI tool at least twice. While some sites completed multiple tools throughout the project, only the first completion of the tool (“baseline”) and final completion of the tool (“end line”) are reported (as above, project periods ranged from 1 to 3 years). Improvements were seen in the majority of categories of the QA/QI tool. Eighty-four percent of indicators improved by at least 5%. The greatest improvements emerged in the number of health care delivery systems that: (1) reported having written protocols for assessment and response to IPV during visits addressing depression, suicidality, and alcohol and other drug use, (2) provided opportunities for staff to discuss issues and concerns with difficult cases, (3) had IPV agencies or law enforcement schedule and/or come to their clinic to conduct a training, and (4) annually reviewed all protocols related to IPV (Table 1). Notably, in terms of building a stronger partnership with a victim service agency, at follow-up, 100% of sites reported that representatives from an IPV program scheduled and/or came to clinic to conduct a training (within the past 2 years).

In the protocols section, improvements were seen in the number of clinics with written protocols for assessment and response to IPV during multiple appointment types. In addition, at follow-up, more health care delivery systems indicated that they had privacy screens on their computers and policies to ensure that patients are asked about IPV when they are alone. All sites indicated at the beginning of the project that they had a designated place to speak with clients privately and this remained true at the end of the project.

In addition to inquiring about clinic policies and protocols about assessing for IPV, the tool also asked about whether all patients were being provided universal education in the form of two palm-sized educational safety cards. At baseline, 60% of clinics provided safety cards to all patients compared with 87% at follow-up. Only a small improvement was seen in the number to clinics providing information on healthy relationships with patients (80% at baseline and 87% at follow-up).

Little improvement was seen in the documentation of assessment and response category. Seventy-three percent of clinics reported that they documented the provision of referrals to a domestic violence agency or advocate at baseline, and 80% at follow-up.

In the networking and training section, all but one of the domains showed improvement from baseline to follow-up. The most improvement was seen in the number of clinics reporting that an IPV agency came to their clinic within the past 2 years to conduct a training (73% at baseline and 100% at follow-up). Collaborations with law enforcement and child protective services remained low.

The self-care and support section of the tool saw improvements on all three items. The greatest improvement was reported in having options other than individual clinical supervision for staff to discuss concerns relating to difficult cases (40% at baseline and 67% at follow-up).

The data and evaluation section saw improvements on all three items. The most improvement was seen in the number of clinics annually reviewing all their protocols, which improved from 33% at baseline to 67% at follow-up. The least improvement in this section was with tracking the number of patients assessed for IPV (53% at baseline and 60% at follow-up).

Lastly, in the environment and resources section, two out of the four items saw improvement. Most clinics (93%) reported at baseline that they had materials on IPV available for patients to take and that the number remained the same at follow-up. Eighty percent of clinics had materials that focused on adolescent relationship abuse and that number did not change at follow-up. Only a small improvement was seen in the number of clinics offering materials on how violence exposure affects children (60% at baseline and 67% at follow-up).

Discussion

This descriptive qualitative study showed that implementation of the QA/QI tool was feasible, with most participating health center sites completing the tool at least twice. Multiple changes emerged in health sites' policies, protocols, and practices. Improvements were especially notable for changes in protocols for provider/patient interaction as well as with staff training and support, clinical operations, and data collection. At the provider/patient interaction level, more health sites reported that patients were being assessed for IPV in private settings and offered universal education. At the staff level, improvements were seen in the supports offered to staff when caring for a survivor as well as supports available for staff who were currently or had previously experienced IPV. The number of training opportunities offered to staff increased as well, indicating the site's commitment to staff members' professional development. And lastly, with clinical operations and data collection, improvements were seen in the number of sites with protocols concerning assessment for IPV (including how often those protocols should be reviewed) and increase in their record keeping and documentation of IPV assessments, disclosures, and referrals.

This feasibility study involved implementation of a quality assessment tool within health care delivery systems to guide improvements in the health care response to IPV. System-level changes have been identified as key to implementation and support of provider-level practice change in IPV assessment and response. ⁷ IPV is similar to other health issues, which have seen improvement in screening in the clinical environment thanks to the driving force of health care systems change. ²⁴ Not only does this type of change provide comprehensive support and guidance for providers in assessing for IPV, it also ensures long-term sustainability of the practice. ⁷ Not only were these health care sites able to implement the tool, participation in this initiative and use of the tool appeared to support improvements in clinic protocols, policies, and practices as well. The largest gains were seen in the number of sites providing universal education to all patients, the number of sites whose protocols instructed providers to assess for IPV during various types of visits, and the number of sites that received training from a victim service agency. The areas that did not change were either those areas where sites were already at a high level of implementing a specific practice (“ceiling effect”) or sites that perceived the area to not be a primary focus of DVHCP (such as partnership with child protective services).

The fact that many more providers were giving patients educational safety cards with information on how violence can impact health could be a result of both the training that the providers received and changes that were seen in many areas of the QA/QI tool. First, the increase in the number of sites whose protocols dictated that providers assess for IPV at various types of visits could have impacted the number of providers distributing the safety cards, as the safety card provision is meant to be a part of comprehensive IPV assessment and universal education using the CUES intervention. This increase in clinical protocols may reflect the fact that the DVHCP initiative recognized that achieving changes in provider practices is difficult without organizational- and system-level protocols in place that support such changes, especially intentional collaboration with a local victim service agency. Therefore, such system-level changes were encouraged throughout the project, including during the learning collaborative webinars led by Futures Without Violence.

In addition to changes in written protocols, many sites began providing scripts to help providers navigate situations in which a patient disclosed IPV and to help with safety planning. The educational safety cards are meant to be used in situations such as these, as they provide resources for survivors and can help guide discussions about IPV. Having specific scripts and prompts might have reminded providers to give out the card and engage in discussion about IPV with their patients.

System-level changes in how IPV is assessed in health care settings must be sustainable. Sites reflected recognition of the importance of sustained attention to collaboration, as all sites reported at the end of the project that a victim service agency had either come and conducted a training or was scheduled to do so. The number of sites receiving trainings from IPV agencies increased more than the number of sites receiving trainings from other outside organizations. To nurture multisector collaborations, continued effort to invite other agencies that address other social influences on health to train staff may be an important strategy to nurture these partnerships and ensure that providers and staff are aware of the resources available to help patients. While not a requirement for DVHCP partnerships, formal memoranda of understanding between victim service agencies and health care delivery systems may also assist with making this collaboration routine practice.

Fewer changes in some areas could be due to several factors. First, some domains (such as the number of clinics with brochures about IPV that patients could take) were already so high at baseline that improvement could not be seen (i.e., “ceiling effect”). While we cannot say that DVHCP had an impact on such areas, it was promising to see that those numbers remained high throughout the duration of the project. A second factor influencing sites' ability to implement certain changes could be related to time and technology costs. Few changes were seen in some of the questions about tracking and documentation, areas that require extra time on the part of clinic staff and new systems for tracking and data extraction (including updating the electronic health record). Further research is needed to determine an efficient way to implement these changes.

There were several limitations to this exploratory study. First, some sites completed multiple QA/QI tools throughout the duration of the project (either because they were involved in the project longer or because they were more responsive to requests to complete the tool every 6 months). As sites were provided with continuous feedback after each completion of the tool in the hopes that this information would guide and motivate future changes, it is possible that those sites who completed more tools may have had more favorable outcomes on the “end line” tool than those who completed fewer tools. This was certainly one of the desired effects of the QA/QI tool in terms of using the evaluation tool as a mechanism for promoting continuous quality improvement. Second, while sites were advised to have a clinic administrator or someone else familiar with the site's policies and protocols complete the tool, it is possible that different people completed the baseline and follow-up tools for sites and might therefore have had different knowledge of the site's policies and protocols. In addition, the QA/QI tool is completed by the site, without objective confirmation of reported changes, introducing social desirability and recall biases. Site managers were asked to be as honest as possible, and the low scores across several domains suggest that respondents completing the tool were attempting to capture actual site-specific practices. Finally, as a U.S.-based project, this tool was developed to promote changes within clinical settings and encourage cross-sector collaborations with victim service agencies and other community organizations in the United States, and thus would need to be modified for use in global settings.

Conclusions

IPV is a major public health threat, and health care systems are critical points for providing assessment of and support for survivors of IPV. Utilization of the QA/QI tool supported effective and substantive changes from clinical operations to interpersonal levels in delivering care to IPV survivors. This study demonstrates the feasibility and effectiveness of implementing such a tool to encourage sustainable cross-sector collaboration between victim service agencies and health care delivery systems.

Addressing IPV in health care settings can have an enormous impact in promoting health and well-being of survivors of IPV. The QA/QI tool evaluated in this study is a low-cost, promising means of guiding system-level change, which in turn may have a larger impact on providers' practices and patients' health outcomes. Two of the common barriers to addressing IPV in health care settings that are often identified are health care provider's lack of knowledge and concerns about time. These barriers can be addressed through system-level changes, as shown by the implementation of the QA/QI tool. Promoting system-level changes that facilitate staff education and support may help to address IPV in clinical settings and, in turn, to provide improved care for populations at particularly elevated risk for IPV.

In summary, the implementation of a QA/QI tool in clinical settings was feasible and appears to have helped document substantive changes in multiple areas of policy and practice in community health care settings. In addition to training health care providers about how to address IPV in the clinical setting, implementation of the QA/QI tool can be feasibly integrated into health care settings to promote cross-sector collaboration with victim service agencies and to increase survivor-centered policies, protocols, and practices within health care delivery systems.