Evaluating a health care provider delivered intervention to reduce intimate partner violence and mitigate associated health risks: study protocol for a randomized controlled trial in Mexico City

Our study utilizes a cluster-randomized controlled trial design and will take place between 2012 - 2015. Briefly, 42 public health clinics (first level of care facilities) were randomized to either the treatment or control conditions and women are to be assessed for information at baseline, time 2 (3 month follow-up), and time 3 (15 month follow-up). Further details are described below.

Mexico City, the capital and most populous urban center of Mexico, provides the study setting in which our research will occur. In this megacity, almost half of the population is uninsured (approximately 3,859,963 residents) [35]. The uninsured population receives health care, through the federal health care program, Seguro Popular, at MoH clinics in Mexico City, this public insurance scheme aims at providing universal health care; a right recognized in the Mexican constitution of 1983 [36].

The MoH health system in Mexico City comprises 206 health clinics and 31 hospitals. The 206 MoH clinics are classified into Type I, Type II and Type III based on level of care. Clinics classified as Type III are large clinics that offer a broader array of services than Type I or II clinics. For instance Type III clinics offer laboratory tests. Clinics classified as Type I and II, are community clinics with 1 or 2 doctors and about 2 nurses; their focus is on community outreach and immunizations with Type I being the smallest level of community health center. Type III clinics were selected to participate in the study after conversations with the MoH because of their lack of IPV services and larger patient volume compared to Type I and Type II. These Type III clinics, chiefly located in low and middle income neighborhoods, serve primarily a low income population, and are thus eligible for the Seguro Popular. The number of women served by these health clinics varies greatly from borough to borough. Over 2,336,000 residents use the services of the MoH health clinics and hospitals annually [37].

At these Type III primary care clinics, the first points of contact for patients are nurses. The flow of patients at these clinics is as follows: patients line up early in the morning in order to get an appointment on a first come first served basis; although the MoH is currently implementing an appointment system that will reduce waiting times for patients. Nurses take somatometry measurements on all patients before they are seen by the doctors. Social workers do not have contact with every patient unless they are specifically needed in instances if patients need a referral to a hospital or to a different health clinic.

Out of the 60 Type III health clinics, one was excluded from potential recruitment because it was located in the borough of Milpa Alta; a remote and rural location with limited access to community resources and a small populace (<2% of Mexico City’s population). Two additional clinics were excluded to reduce contamination. These hospitals were located less than three blocks from hospitals and some hospitals also had violence against women programs.

Approximately 900 women from 42 health clinics were calculated to be sufficient sample size in order to achieve 80% power to detect a 15% difference in IPV frequency between treatment versus control arms at the α < 0.05 statistical significance level. A conservative intraclass correlation of 0.07 was assumed for clustering at the clinic level. This total sample size takes into account an attrition rate of 45%. The study proposal has budgeted for tracking participants that are relatively more difficult to track in order to reduce attrition.

Health clinics were first stratified by city zone (e.g. Center, North, East, West, South). Of the 57 health clinics that met inclusion criteria, 42 were randomly selected and randomized to treatment or control conditions, based on sample size calculations. To select the 42 health centers, all centers were assigned random numbers in Excel and sorted from smallest to largest; health centers were selected based on city zone and in order of their random number. Randomization was completed in STATA.

Women were eligible to participate in the study if they were between the ages of 18-44 years, currently in a heterosexual relationship with a male partner, responded in affirmative to past year sexual or physical IPV, and was not pregnant or pregnant in their first trimester. Study exclusion criteria included cognitive impairment (e.g. slurred speech), seeking treatment for life threatening emergency care, and intending to relocate within two years. Research assistants approached women in the waiting rooms of participating clinics, verified eligibility for the study, and asked women to take an assessment screening that contained items from an abuse assessment screen that is widely used by International Planned Parenthood Federation/Western Hemisphere Region (IPPF/WHR) in Latin American and Caribbean countries and has been previously used in studies occurring in Mexico City [38‐40]. Based on feedback from piloting the assessment, the final tool consisted of eleven questions in order to build rapport between the research assistant and participant. The first nine questions were in regards to the woman’s health and relationship with her partner, including emotional abuse, as directly asking about physical or sexual IPV at the very beginning of the assessment was rendered as too sensitive during our piloting phase. Based on feedback from focus groups carried out with IPV survivors at a community domestic violence agency, concrete examples of physical and sexual IPV were included in the questions (see Figure 1). If a woman answered at least one affirmative response to the validated screening items on physical and sexual IPV contained in the assessment, they were invited to participate in the study; those interested completed written and/or verbal informed consent. Both the assessment and the informed consent process took place within a private area of the clinic.

At baseline data collection, women were asked for their contact information, including the names of three other people in case the woman could not be reached for follow-up appointments. Research assistants called every number provided by the participant to verify that the phone numbers were functional. Participants received a call on average 3 days after they completed the baseline to verify contact information. All participants agreed that it was okay to call them and research assistants were trained to ask if it was a safe time to talk and did not mention violence or the study if another person answered the phone.

Data were collected through a computer-assisted survey which has been shown to improve response rates to sensitive behaviors [41]. Participants completed the survey in a private space in the clinic and were able to listen to the questions in Spanish through headphones and answer on the keyboard. Research assistants were available during this period in case the participant had any questions. Given that clinics were randomized and not individuals, research assistants were not blinded to whether they were in a treatment or control clinic.

Women who participated in the study at treatment clinics received the following intervention: (1) integrated IPV and health screening; (2) supportive care; (3) safety planning and harm reduction counseling; (4) supportive referrals; and (5) booster counseling sessions at three months post baseline. The majority of materials utilized for the intervention were adapted from existing IPPF/WHR materials used in other Latin American and Caribbean contexts and were adapted to the characteristics of Mexico City [38]. A referral directory was also created with the contact information of local community agencies that provide services to IPV victims. During the development phase of the referral directory, IPV staff at the MoH suggested that only community agencies providing free services be included. All community agencies listed in the directory were visited by research staff and addresses and phone numbers were verified. Furthermore, for each community agency, research staff obtained the name of a specific contact person to include in the referral directory. Contact names were included in the referral directory based on pilot focus group research with IPV survivors enrolled at a community agency. Pilot work revealed that women felt more comfortable asking for services if there was a specific person to contact at an agency; and that only having the name of an agency was a deterrent.

Of the agencies, two programs were highlighted: (1) SEPAVIGE, the MoH IPV program, (Gender based Violence Special Services) and UAVIF (Family Violence Prevention and Care Unit). There are approximately 14 SEPAVIGE units located primarily within hospitals. Key services include free psychological treatment and group therapy to IPV victims. UAVIF is associated with the Ministry of Social Development and it consists of 16 units in each district of Mexico City. Each unit offers legal and psychological services. They are also in charge of identifying women eligible for an IPV conditional cash transfer program. The referral directory was handed out to nurses during the training Details are presented in Table 1.

Nurses from all 42 health clinics were invited to participate in the training based on the following criteria: morning shift nurses (due to the walk-in basis of appointments and afternoon shifts not occurring at all clinics) and not a field nurse (due to their limited time at the health clinics). A total of 197 nurses (approximately 8% male nurses and 45% in treatment clinics) received the training. 147 nurses are actively participating in the study (49% in treatment clinics) (26% staff dropped out primarily due to staff turnover at clinics). On average, 4 nurses were trained at each health clinic. Nurses in the intervention group received a 3 day training that covered all the topics of the intervention and then nurses received up to 3 visits by research staff to practice delivering the intervention through role – playing exercises individually and in the clinics. In-depth training was needed as nurses did not feel very comfortable talking about IPV and several rounds of role playing exercises were needed for nurses to gain confidence. Role-playing exercises with research staff were chosen as it preserved rapport between nurses and the research team as well as increased confidentiality of nurses requiring additional training. All training activities were conducted by the research team, IPPF training consultants, and invited guests from local organizations and government offices who specialize in this area. Specific topics in the training included an introduction to IPV, health consequences of IPV, legal considerations in Mexico City, methods to screen for violence and assess for health implications such as reproductive coercion, safety and ethical considerations, referral methods and linkages to other organizations in the community. An educational video modeling the counseling sessions was also developed by the study team for the training of treatment nurses. This video was shown and discussed during research assistant visits to the health clinics as part of the training.

Women in the control clinics were given a referral card containing general information on IPV and a list of resources, which was consistent with the current goal for standard of care in the Mexico City MoH. The referral card is the size of a business-card so that it may easily be hidden from male partners, and is widely used in other IPV screening programs [42]. Staff in control clinics received a one day training that focused mainly on sensitizing nurses and training them on using the abuse assessment. Control clinic staff will receive the intervention training upon completion of the study.

Conducting research with women who have recently experienced violence represents a number of safety concerns [43]. Ethical procedures were guided by World Health Organization standards for conducting research on domestic violence [44] and were taken into account during study preparations. First, all women will provide informed consent which stresses the voluntary nature of the study, the ability to skip questions as necessary, confidentiality, and outlines potential risks and benefits of participation. At the end of the survey, health care providers explain available resources for emotional support and provide a list of resources. Second, the study was designed so as not to reveal to the partner the purpose of the study or the woman’s participation in the study. At all times a woman’s disclosure of IPV remained confidential, including the three friends or family members she listed in case she could not be contacted at follow-up. During follow-up calls, the specific purpose of the study was not given; rather, it was referred to as a study on “women’s health and families”. All research staff were trained by senior study personnel on study ethics. Regular de-briefings, including recommendations for self-care, occurred with research assistants to reduce vicarious trauma among staff. Research staff and nurses participating in the study were encouraged to report any emotional distress, including if they were struggling with experiences of victimization in their own lives, to principal investigators. Study procedures have been approved by the Yale School of Public Health (Protocol # 1202009793), Innovations for Poverty Action (Protocol # 555.23May-001), and National Institute of Public Health (Mexico) (Project #1089) institutional review boards.

Our primary outcome is past-year IPV [32] and our secondary outcomes include past-year reproductive coercion [7, 23], use of community-based resources [33], safety planning measures [33], and quality of life [34]. Descriptive statistics will be used to assess achievement of Aim 1. For the remaining aims, the follow analyses plans will be undertaken: (1) assess if randomization of health clinics was successful by examining if there are statistically significant differences between women’s demographics, by treatment arm, through chi-square or t-tests; (2) assess for differential loss to follow up by demographics and baseline IPV severity; (3) assess if IPV and secondary outcomes (reproductive coercion, safety planning, use of community resources, quality of life) differed at baseline, by treatment arm; and (4) conduct multilevel analyses of the effect of the intervention on outcomes utilizing appropriate models that will account for potential clustering for repeat measures, time, and at the clinic level. We will subsequently estimate adjusted models if any demographics significantly varied at baseline. Analyses will first be undertaken using the intent-to-treat approach. Following this, a per protocol analysis will be undertaken to determine whether effectiveness varied by adherence (i.e. attending the three month booster counseling session) and potentially nurse adherence.

During endline data collection, research assistants will conduct 45 qualitative interviews with female participants to provide insight into observed quantitative results. In-depth interviews will consist of questions regarding perceptions of IPV, safety planning behaviors, reproductive autonomy, and quality of life. Underscoring all questions will be the woman’s perception of how, if at all, participating in the intervention created or affected these changes. Additionally, fifteen interviews were conducted with nurses at treatment clinics to understand their experiences with the HCP delivered approach between the three month follow-up period and the final endline at fifteen months. Interviews will last no longer than one hour and be conducted in Spanish. Upon transcription of audio recordings, interviews will be translated into English. All interviews will be coded using inductive thematic analysis [45] through which emerging themes will be discussed with principal investigators.

Throughout the study period, a process evaluation is conducted, consisting of three key activities: (1) pre-post tests of the nurse training; (2) fidelity assessments with mock clients; and (3) exit interviews with women participating in the study. First, during the training of nurses, we conducted a pre-post test before and after the training to assess the change in knowledge, attitudes, and skills among nurses in regards to screening and referral for IPV. In addition to assessing nurses’ skills for addressing IPV, these pre-posttests were undertaken due to previously noted challenges that may hinder effective interventions, such as health care providers’ victim-blaming attitudes and misconceptions of IPV [28]. Throughout the study, nurses are asked to fill out the short survey at three month intervals to assess any changes, although test-retest bias is a limitation of this process evaluation method.

Second, anonymous mock clients are used in order to assess nurses’ adherence to protocols and to serve as an opportunity to provide feedback during the training phase. Existing literature has also underscored the lack of feedback given to health care providers as a reason for infidelity to screening protocols [18]. Therefore, after completing the training, all intervention nurses are visited by trained research staff. The staff member screens positive for IPV, and goes through the clinic visit, including screening, referral, and counseling following a scripted scenario. Following the health visit, the staff member completes a fidelity checklist to determine areas in which the nurse is exceeding expectations and areas in which the nurse needs improvement. All nurses are informed that they will be visited by a mock client and that all mock clients will adhere to the script used in the training. No health nurse receives more than two mock client visits throughout the study period. Feedback was given during on-site training prior to the launch of the baseline; subsequent visits were used for fidelity monitoring only. Nurses in the control group also received one visit from the mock client, but feedback will only center on standard of care, rather than comprehensive screening and counseling women. Fidelity checklists from these mock clients do not contain identifying information and were entered into a secure database upon completion. Tape recording counseling sessions were also considered as a method of assessing fidelity. However, due to safety concerns for the women, mock clients were chosen as the more appropriate method to assess intervention fidelity.

The final activity to monitor fidelity was exit interviews with women in both treatment and control clinics. Women who were participating in the study in treatment clinics were randomly selected by research staff in the clinic after they had received the intervention, based on selection of questionnaire identification numbers. Exit interviews are designed to last between 5-10 minutes where all women were asked to provide an additional written informed consent prior to completing the survey. All women in the intervention and control groups were eligible to participate in the exit interviews.

A costing analysis is also included in the study protocol. Trained research assistants have been identifying human, financial, physical, and time resources required for the HCP delivered program throughout the study. Data has been collected and categorized by research-related costs or operational costs in order to calculate the monetary value of resources needed to conduct the intervention. The costs and effectiveness of this approach will be compared between treatment versus control groups and will mirror methodology from a previous cost-effectiveness analysis of an IPV intervention [46]. This costing exercise will provide insight into the scalability of the intervention and economic impacts of broader policy recommendations.

Study recruitment is currently underway at the time of initial manuscript submission. Upon resubmission, the baseline and three-month follow-up surveys have been completed.