MOSAIC (MOthers' AdvocateS In the Community) for Pregnant Women and Mothers of Children Under 5 with Experience of Intimate Partner Violence: A pilot randomized trial study protocol

Background: Pregnancy and motherhood increase the risk for long-term exposure to physical, psychological and sexual intimate partner violence (IPV; sexual or physical violence by current or former partners). Pregnant women and mothers with children under 5 who have experienced IPV exhibit poor physical and mental health and obstetric outcomes. Depression and posttraumatic stress disorder (PTSD) are the two most common mental health consequences of IPV. There is good evidence that irrespective of the severity of IPV, women with good social support have better mental health outcomes. Methods: This study will develop MOthers' AdvocateS In the Community (MOSAIC) Plus intervention for pregnant women and mothers with children under the age of 5. MOSAIC uses trained mentor mothers and has been found to reduce subsequent IPV. This study will blend the original MOSAIC intervention with principles of interpersonal psychotherapy (IPT) to address symptoms of depression, PTSD, and prevent subsequent risk of IPV. We will conduct a pilot randomized trial of the MOSAIC Plus intervention compared to the traditional MOSAIC intervention to determine its feasibility and acceptability. Study samples include focus groups (n=36), open trial (n=15), and a randomized pilot trial including 40 pregnant women and mothers with children under 5 who report current/recent of IPV and elevated symptoms of maternal depression and/or PTSD. The studys primary outcome will be changes in maternal depressive and PTSD symptoms as measured by the Patient Health Questionnaire and Davidson Trauma Scale respectively. Secondary outcomes will include reduction in subsequent IPV as evaluated by the Composite Abuse Scale, improvement in functioning, measured by the WHODAS-II, changes in social support and effectiveness in obtaining resources as assessed by the RAND social support scale. Discussion: This is a formative study evaluating the feasibility and acceptability of a mentor mother intervention for pregnant women and mothers with children under 5. Promising results of this study will be used for a larger, fully-powered randomized trial evaluating the effectiveness of a mentor mother intervention in preventing subsequent IPV and reducing depressive and PTSD symptoms in this population.

Abstract (300/300) 3 4 Background: Pregnancy and motherhood increase the risk for long-term exposure to 5 physical, psychological and sexual intimate partner violence (IPV; sexual or physical violence 6 by current or former partners). Pregnant women and mothers with children under 5 who have 7 experienced IPV exhibit poor physical and mental health and obstetric outcomes. Depression 8 and posttraumatic stress disorder (PTSD) are the two most common mental health 9 consequences of IPV. There is good evidence that women with good social support have better 10 mental health and IPV outcomes.  Discussion: This is a formative study evaluating the feasibility and acceptability of a 25 mentor mother intervention for pregnant women and mothers with children under 5. Promising 26 results of this study will be used for a larger, fully-powered randomized trial evaluating the 27 effectiveness of a mentor mother intervention in preventing subsequent IPV and reducing 28 depressive and PTSD symptoms in this population. Violence against women is a global human rights concern and a significant public health 3 challenge; one in three women experience lifetime physical and/or sexual violence.(1) As many 4 as 5.3 million U.S. women experience intimate partner violence (IPV; referring to sexual or 5 physical violence by current or former partners) annually(2). Depression and posttraumatic 6 stress disorder (PTSD) are the two most common mental health consequences of IPV. In turn, 7 developing depressive and PTSD symptoms after experiencing IPV increases future risk of 8 IPV(3). IPV is also associated with chronic physical health problems(4, 5) and self-harm (6).

9
Pregnancy and motherhood increase the risk for prolonged exposure to physical, 10 psychological and sexual IPV(7). IPV can start or worsen during the perinatal period. As many 11 as 4% to 8% of women report IPV during pregnancy (8). Compared to those who do not 12 experience IPV, mothers who experience IPV during pregnancy or in the year following 13 childbirth are at elevated risk for experiencing physical injury, other medical challenges, 14 homicide(9) and mental health problems including suicide (10,11). Twenty percent of 15 postpartum deaths are attributable to suicide(12). 16 Pregnant women and mothers who experience IPV may not seek care when they need 17 it(13). For example, perinatal women with experience of IPV may not initiate prenatal care 18 until third trimester (14). They are more likely than women without IPV to miss three or more 19 prenatal visits(15). Perinatal women with experience of IPV are twice as likely to have babies 20 with low birthweight(16), higher rates of severe acute maternal morbidity(17) and have infants 21 requiring intensive care(18). IPV increases risk of depressive and PTSD symptoms, which in 22 turn increases risk of subsequent IPV(19-21) (8,21). 23 Interventions are needed that address both, especially for pregnant women and mothers of 24 young children. (22)(23)(24). There are very few interventions that effectively reduce IPV among 25 pregnant women and mothers with children under 5 who report IPV(22-24). One of them, 26 Mother AdvocateS In the Community (MOSAIC) forms the basis of this study(25). Very few 27 existing interventions address maternal depressive and PTSD symptoms that often follow IPV

28
(3, 26,27). An integrated intervention that addresses elevated symptoms of maternal depression 29 and PTSD while reducing subsequent IPV is needed.

30
Mother AdvocateS In the Community (MOSAIC)(25) is a non-professional intervention 31 delivered by mentor mothers from the community to reduce IPV in pregnant women and 32 mothers with children under 5(28). The intervention combines elements of mentoring and IPV-33 specific support provided by mentor mothers. MOSAIC was tested in a fully-powered RCT for 34 pregnant women and mothers with children under 5 in Australia and was found to reduce . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted April 16, 2022. ; https://doi.org/10.1101/2022.04.14.22273885 doi: medRxiv preprint 1 subsequent IPV(24). However, the effects of the intervention depression were not significant 2 and PTSD symptoms were not evaluated. The proposed study will augment MOSAIC with 3 principles of an evidence-based intervention to improve maternal mental health. Interpersonal 4 psychotherapy (IPT) is the front-line treatment for perinatal depression(29) and has been found 5 to reduce PTSD symptoms in the perinatal period(30). IPT addresses maternal mental health 6 by helping women increase their general social support systems and build communication skills 7 and confidence to access needed resources and help. IPT can be effectively delivered by lay 8 providers(31). IPT is theoretically consistent with the MOSAIC approach and addresses the 9 consequences of IPV. IPT focuses on building or better utilizing one's social support network 10 within the context of a life stressor (32, 33). The proposed study aims to integrate IPT principles 11 into MOSAIC to address both IPV and its mental health sequelae.

12
Specifically, the study aims to:    2.1.1. Project advisory board. We will establish a project advisory board that will 24 include 6-8 pregnant women and mothers with children under 5 and potential mentor mothers 25 from the community. They will provide feedback on the content of the intervention in the 26 development stage and on study procedures and results generally.

27
Manual development. The goal of the development phase will be to develop the MOSAIC 28 Plus intervention manual. We will begin with the MOSAIC intervention manual which was 29 developed to reduce subsequent IPV experiences in perinatal women and mothers with children 30 under 5 (28)but was not specifically designed to address concomitant mental health issues. The

31
MOSAIC intervention uses non-professional befriending, domestic violence-specific advocacy 32 and mentoring by mentor mothers from the community(34). The mentor mothers may or may 33 not have lived-experience with IPV, maternal depression or PTSD. MOSAIC Plus will expand 34 the social support offered by mentor mothers in MOSAIC with IPT approaches for broadening . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted April 16, 2022. ; https://doi.org/10.1101/2022.04.14.22273885 doi: medRxiv preprint 1 women's social support network generally. In MOSAIC Plus, we will expand the existing 2 MOSAIC manual by integrating IPT approaches to address depressive and PTSD symptoms 3 among IPV-exposed mothers. IPT conceptualizes interpersonal functioning and disturbance in 4 social support as intimately linked to psychological symptoms(35). IPT identifies a difficult 5 life event or current interpersonal problem and addresses it by helping the individual to build 6 or better utilize the individual's extended social support network with specific IPT strategies 7 that include psychoeducation, communication skills practice, and exploration/encouragement. 8 Focus group discussions. We will conduct focus group discussions with potential study 9 participants and potential mentor mothers to enhance and tailor the MOSAIC intervention to 10 address mental health outcomes (reduction in depressive symptoms and PTSD symptoms). We 11 will use information generated from focus group discussions to ensure that our manual 12 expansion optimally addresses the needs of our target population, the mentor mothers, and the 13 eventual implementing systems. We will conduct six focus group discussions of 6-8 members 14 each (for 30-44 in total). Three groups with pregnant women and mothers of young children 15 with current, recent or prior experience of IPV, 2 groups with potential mentor mothers, and 16 one group with staff from community agencies (the settings where we would envision 17 MOSAIC Plus being implemented if found promising). Focus group discussions will explore 18 perspectives of potential clients and mentor mothers regarding the potential acceptability, 19 benefits, and risks of the proposed MOSAIC Plus intervention and of the IPT components we 20 plan to integrate. The goal of the 3 focus groups with pregnant women and mothers with 21 children under 5 with IPV experiences will be to explore: (1) women's perspectives on the 22 greatest needs of IPV-exposed perinatal mothers; (2) what they would like most from mentor 23 mothers to help with IPV, depression, and PTSD symptoms; and (3) feedback on the proposed 24 MOSAIC Plus intervention and on the IPT-based components to be integrated. The goal of the 25 two focus groups with nurse-family partnerships will be to explore ways to maximize 26 acceptability of MOSAIC Plus, to integrate the intervention into their ongoing programs, and 27 how to design the MOSAIC Plus program to maximize future implementability.

28
Participants for potential clients focus group discussions will be recruited from Hurley 29 Medical Center's perinatal clinic and YWCA using our standard study recruiting procedures 30 and criteria. For the other two focus group discussions, we will recruit participants from local 31 agencies serving mothers.

32
Data analysis and manual development. A moderator accompanied by a note taker will 33 facilitate the focus group discussions, which will be recorded by using a digital audio recorder.

34
The recordings will be transcribed verbatim by an experienced transcriber. We will use the . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted April 16, 2022. ; https://doi.org/10.1101/2022.04.14.22273885 doi: medRxiv preprint 1 framework approach by using a framework drawn from the data. We will modify our coding 2 framework based on the newly emerging themes. For rigorous and high-quality data 3 management, we will use NVivo qualitative analysis software, version 12(36). We will develop 4 an outline of the MOSAIC Plus intervention manual based on the original MOSAIC 5 intervention manual. We will embed contents relevant to reducing depressive and PTSD 6 symptoms to the new manual based on results of the focus group discussions. We will enhance, 7 refine and adapt the MOSAIC Plus intervention for the outcomes of interest based on the results 8 of the focus group discussions. Dr. Hailemariam will conduct member checking with selected 9 study participants to validate the fit, credibility and transferability of the results (i.e., main 10 themes and translation of them into the manual itself) from the focus group discussions.

11
Open trial. The second part of the development phase will include conducting a small open 12 trial to further refine the intervention manual and assess feasibility of research procedures. We 13 will recruit 15 pregnant women and mothers with children under 5 with experience of IPV, 14 who also report elevated symptoms of depression and/or PTSD, meeting the same inclusion 15 criteria as participants in the randomized trial (see sampling and recruitment section below), 16 and who will receive the MOSAIC Plus intervention. We will review our strategies for 17 recruitment and retention of participants, strategies for recruitment, supervision and retention 18 of the mentor mothers and mentor mothers' compliance with the study protocol. Participants 19 will be requested to complete an intervention specific End-of-Treatment Questionnaire 20 assessing perceived helpfulness of various components of the intervention and their comfort 21 with the research processes and assessments. We will review the responses of participants at 22 an exit interview to elicit information for further refining the intervention protocol.

23
Program and procedure modifications. Based on data from the development phase, exit 24 interviews, and ongoing feedback from the study team, the intervention manual, fidelity scales 25 and the training program will be modified. Study recruitment and retention procedures will 26 also be revised based on experiences of the research team and participants. These procedures 27 will enhance study feasibility by ensuring that our assessments do not place undue burden on 28 participants that measures are feasible and acceptable; and that dropout rates are minimized.

30
In the pilot study phase, the MOSAIC Plus intervention will be compared to the traditional 31 MOSAIC in 40 women. The goal of this phase will be to assess the feasibility of the research 32 design, the acceptability of the MOSIAC Plus intervention and to explore 95% confidence 33 intervals of reduction around depressive and PTSD symptoms, and decrease in the experience . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted April 16, 2022. ; https://doi.org/10.1101/2022.04.14.22273885 doi: medRxiv preprint 1 of IPV within a 9-month period from intake. The duration of both interventions will be 9 2 months. Study assessments will take place at baseline, 3, 6 and 9 months.
3 Sampling and recruitment. The randomized pilot study will involve 40 women who are 4 1) pregnant and/or are mothers of children under 5 and report IPV experiences in the past 6 5 months (as assessed by Composite Abuse Scale (37) ) 2) aged 18 or above, and 3) have elevated 6 depressive and/or PTSD symptoms (including full disorder). This will be determined using the 7 Patient Health Questionnaire (PHQ-9) with a cutoff point of ≥9(38), and the Davidson Trauma 8 Scale (DTS) with a cut-off point of ≥40(39). We anticipate that the majority of study 9 participants will have both elevated depressive and PTSD symptoms, given that 50% of 10 individuals with PTSD also have comorbid major depressive disorders, (40, 41) especially 11 among women with IPV(21). Participants will be excluded if they (1) cannot provide the name 12 and contact information of at least two locator persons (~6%), and/or (2) do not have access to peer support for pregnant women and mothers?". The assessing provider will also explain that 27 their decision whether to obtain additional information about the project has no potential impact 28 on their access to other services at the agency. Women who are willing to meet with study staff 29 will be given a referral slip by their providers which they will present to the study research 30 assistant (RA). The study RA will meet potential participants in private locations at each 31 agency. The RA will explain all aspects of the study, including confidentiality and its limits, 32 and address questions. If the participant agrees, they will provide a written informed consent 33 complete the baseline assessment. We will ask participants if they would like us to read consent 34 forms aloud. The RA will emphasize that enrollment in the study is completely voluntary.
. CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted April 16, 2022. ; https://doi.org/10.1101/2022.04.14.22273885 doi: medRxiv preprint 1 Those who consent to participate will be provided with a copy of the informed consent 2 document.

3
The MOSAIC Plus intervention: The MOSAIC Plus intervention will be delivered to 4 those in the intervention arm as specified in the manual, for the duration of 9 months. After 5 completing the screening interviews, the study research assistant will introduce consenting, 6 eligible participants to a MOSAIC Plus trained mentor mother (See training procedures below).

7
The mentor mothers will work as a mentors and advocates in the areas of IPV safety planning, 8 building social support, managing depressive symptoms and enhancing effectiveness obtaining 9 community resources. During the initial phase, mentor mothers will explain their roles as a 10 MOSAIC Plus intervention provider and exchange contact information (i.e., the participant's 11 contact information and the mentor mother's study cellphone number). In the following four 12 weeks, the MOSAIC Plus mentors will meet their clients in a private, confidential location to 13 assess if there is ongoing IPV, conduct safety planning, strength and needs assessment. They 14 will collaboratively set a realistic treatment plan for the intervention period. MOSAIC Plus 15 mentor mothers will maintain regular contact with their clients at least once a week (may be 16 more as needed). Information obtained during initial assessment (safety, strengths, needs) will 17 be used to plan the rest of the mentoring intervention in the subsequent months.

18
After the initial phase is complete, as in the original MOSAIC intervention, mentor mothers 19 will continue meeting their clients (in person or by phone) for a session no less than 60 minutes.

20
This level of contact is high but was feasible and acceptable in the original MOSAIC trial(24).

21
In person meetings will take place in the community at convenient, private, and safe locations 22 (libraries, treatment facilities, public buildings) as suggested by the clients. Telephone 23 meetings will only be made from a study telephone line given to each mentor mother. The 24 mentor mothers will review their clients' activities since the initial planning meeting and 25 problem-solve any challenges they may encounter. Mentor mothers also facilitate their clients' 26 access to other services by informing, educating and advocating for them when needed. Their 27 services will involve 1) building social support networks by helping their clients solve any 28 interpersonal difficulties they may have within their networks. Mentor mothers will also work 29 with their clients by linking them with new social support groups and networks available in the 30 community. They will focus on social support because strong social support protects against 31 the negative effects of IPV on mental health (42). 2) Mentor mothers will inform women about 32 services and advocate for their access. Women with experience of IPV may have limited access 33 to information, and hence poor access to services. Mentor mothers will help improve their 34 access to services by actively identifying the needed services, creating proper linkage with . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. but utilizes a broader conceptualization (i.e., help the mother build her own social support 10 network) and evidence-based IPT strategies for doing so.

11
The control condition. Those in the control condition will receive the MOSAIC 12 intervention without the added IPT component. Participants in the control condition will also 13 receive ongoing monitoring of depressive and PTSD symptoms and suicide ideation with 14 appropriate referrals from mentor mothers. They will also be provided a county-specific IPV 15 resource guide after consenting to enroll in the study.

16
Randomization. Randomization to MOSAIC Plus or MOSAIC interventions will occur after 17 the baseline assessment, in a 1:1 ratio. All baseline assessments are blind. Immediately after 18 randomization, the research assistant will review the study outcome assessment schedule, 19 means of contacting the research staff, and participants' contacts with all participants. The 20 study statistician will prepare the randomization schedule using opaque sealed envelopes 21 before the enrollment of the first participant. developing social support is useful. Social support also helps with depressive and PTSD 28 symptoms(42). Lack of social support is associated with antepartum depressive symptoms(44).

29
Studies suggest that social support, received in the form of informational, instrumental and 30 emotional support, reduces depressive and PTSD symptoms in women with IPV experience(5, 31 42, 45). Studies in population-based samples of women have found that women with 32 experience of IPV with stronger social support are protected against the negative effects of IPV 33 on mental health including depression and PTSD (42,46). In contrast, lack of perceived social 34 support is a primary risk factor for developing PTSD following a traumatic experience(47-52).
. CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) 1 Therefore, social support is a hypothesized target mechanism of the effects of MOSAIC Plus 2 on IPV and mental health outcomes. Mentor mother training, supervision and supervision: We will recruit 8 mentor mothers 8 and train them to deliver the MOSAIC and the MOSAIC Plus intervention with fidelity.

9
Turnover in community interventionists can be high. Therefore, by training 8 mentor mothers, 10 the study is likely to have at least 6 for the trial. If needed, additional mentor mothers can be 11 recruited and trained. Mentor mothers will be selected with the help of our Community-Based 12 Organization Partners (CBOP), our community partner for this study. With the support from 13 CBOP and its network of partners, we will identify and interview women from Flint Genesee 14 County. Mentor mothers will be: 1) 21 years of age and above, 2) have good listening skills 15 and embody a non-judgmental attitude, 3) a permanent resident in Genesee County, 4) report 16 that they are able to use a computer to upload intervention recordings and have access to a 17 computer with internet, 5) are mothers, 6) have access to reliable transportation and 7) are 18 available up to 2 hours per week by phone or in person. Mentor mothers will have weekly 19 group supervision and case discussion with Dr. Hailemariam. Dr. Hailemariam will also listen 20 to audio recordings of the mentor mother intervention sessions and will provide feedback.

21
During weekly supervision or in the audio files, if a mentor appears to be triggered by a case 22 discussion or talks about being overwhelmed in a group supervision meeting, Dr. Hailemariam 23 will promptly schedule an individual meeting and debrief session with the mentor mother.

24
Mentor mothers will be constantly reminded that if they are feeling overwhelmed by a case 25 (eg., thinking about the case more than they want to, poor sleep, changes in appetite, dreaming 26 about the case, feeling anxious) they will be encouraged to contact Dr. Hailemariam for a 27 debrief and supportive supervision.

29
Safety. We will give strong emphasis to the safety of the mentor mothers. Briefly, we will 30 develop a comprehensive protocol, based on the protocols used successfully in previous 31 MOSAIC studies (24,53), that covers safety planning with the client and procedures to ensure 32 safety of the mentor mothers. Contents of the protocol will involve leaving the scene if there is 33 an apparent violence and procedures for immediate reporting of any safety concerns and 34 questions. Mentor mothers will be given a study cellphone to communicate with their clients.
. CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted April 16, 2022. ; https://doi.org/10.1101/2022.04.14.22273885 doi: medRxiv preprint 1 They will meet their clients in public places with private space such as health facilities, libraries 2 and parks to reduce potential risk of victimization. To protect mentor mothers' privacy, mentor 3 mothers will not share their residence addresses and last names with their clients. We will train 4 mentor mothers on how to respond to clients' substance abuse. Moreover, safety plans will be 5 adapted and tailored to each participating woman based on her needs. Mentor mothers will also 6 be trained on burnout management, self-care and will be offered an opportunity to debrief with 7 study investigators quarterly, and to call on an as needed basis.

9
Fidelity ratings. Intervention adherence and competence ratings will be developed based 10 on the manual and previous IPT adherence and competence scales(30, 54). Adherence rating 11 scales for both conditions will consist of a checklist of tasks to be completed at each meeting.

12
Competence rating scales will be completed from audio recordings of mentor mother visits and 13 will reflect mentor mothers' general skills, such as reflective listening, safety planning, linking 14 women to services, advocacy and effective goal setting. The study team will use these scales 15 to rate a randomly selected 33% of the open trial tapes. Scale development will continue until 16 the item content is satisfactory and interrater reliability is acceptable (> .80). Assessments: The main goal of this pilot study is to explore the feasibility and 21 acceptability of the MOSAIC Plus intervention. We will also evaluate the feasibility of the 22 research and training procedures. In addition to these main outcomes, we will conduct several 23 assessments to evaluate changes in depressive and PTSD symptoms, reduction in future IPV 24 and changes in effectiveness in obtaining community to resources. The study research assistant 25 (RA) will be trained to fidelity to administer clinical and other assessments. The clinical 26 training process will also be supplemented by standard case vignettes showing elevated 27 depressive and PTSD symptoms to acquaint trainees with common manifestations of IPV, 28 symptoms of depression and PTSD. We will record all RA interviews and do regular quality 29 and fidelity monitoring. Study mentor mothers will not play a role in the data collection 30 process. Data will be collected at baseline, 3 months, 6 months and 9 months (post-intervention 31 completion) (see Table 1: the SPIRIT schedule). We selected the following feasibility, 32 acceptability, and safety metrics based on face validity, clinical experience, and relevant 33 literature.

34
Demographic/screening measures will include age, educational level, marital status, . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted April 16, 2022. ; https://doi.org/10.1101/2022.04.14.22273885 doi: medRxiv preprint 1 occupation, employment (status, # hours per week), income, race, gestational age/age of 2 children. At follow ups, occupation, employment (status, # hours per week), and income will 3 be repeated. We will also assess relationship information, such as current relationship to the 4 abuser/s and length of relationship. We will use the Composite Abuse Scale (CAS)(37) to 5 include women who report any physical and sexual violence in the past six months. We will 6 also use the Patient Health Questionnaire (PHQ-9) with a cut-off point of ≥ 9(55) and/or the  Primary: Study Feasibility and Treatment Feasibility/Acceptability and safety. We will 10 assess feasibility of the research procedures by examining study recruitment and refusal rates, 11 participants' willingness to be randomized, follow-up rates, reliability and range of responses 12 to study questionnaires. We will assess the feasibility and acceptability of MOSAIC Plus by 13 examining treatment completion (based on the jointly established treatment plan) and drop-out. 14 We will examine reasons for termination from study and/or intervention for consistent patterns. 15 We will examine the acceptability of the MOSAIC Plus intervention by using data from CTQ-8 16 treatment satisfaction questionnaire, the End of Treatment Questionnaire, and detailed exit 17 interviews. Clients' experiences with the mentor mothers, the quality of their working 18 relationship and their level of satisfaction with the service will be evaluated using the Working

19
Alliance Inventory-Short Revised (WAI-SR). Moreover, as in the original MOSAIC study(34), 20 we will also explore perspectives of the mentor mothers regarding acceptability of the 21 MOSAIC Plus intervention. We will use Competency Assessment Inventory (CAI), which tests 22 the attitudes, knowledge, and skills needed to provide high-quality mental health care (i.e., 23 learns and respects clients' preferences about intervention, creates opportunities for clients to 24 practice skills).

25
At the end of the open trial and at 2 time-points during the RCT (i.e., after enrolling 20 26 and 40 participants), we will compute the descriptive data on all target outcomes described in 27 table 2. The scientific team will convene to discuss how our actual outcomes compare to target 28 outcomes. If the team identifies discrepancies between target outcomes and actual outcomes, 29 we will 1) investigate the reason for failure to meet the stated outcomes, and 2) discuss among 30 the research team. Depending on the nature of the discrepancy, we may modify recruitment or 31 assessment procedures, instructions to participants, training procedures, or other trial aspects.

32
Table 2 below presents assessment methods. is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted April 16, 2022. ; https://doi.org/10.1101/2022.04.14.22273885 doi: medRxiv preprint 1 2 Clinical and services outcomes. 3 Mental health outcomes. Reduction in depressive symptoms will be evaluated using the 4 nine-item Patient Health Questionnaire (PHQ-9), a reliable self-report measure assessing the 5 presence and severity of depressive symptoms(38). We will measure changes in PTSD 6 symptoms by using the 17 item Davidson Trauma Scale (DTS)(39), a reliable and well-7 validated self-report measure(56) of trauma assessing DSM-5 symptoms of PTSD rated on a 8 five point frequency. We will use the DTS as a cross-sectional measure of PTSD symptoms in 9 three domains including intrusion, avoidance/numbing and hyperarousal. Respondents will be 10 asked to identify the most troubling traumatic experience in the past week and measure its 11 severity and frequency in 5-point scales.

MOSAIC Plus Feasibility and Acceptability ( MOSAIC Plus participants only) Acceptability (End of Treatment Questionnaire)
Endorsing an average score of 3 or more (out of 5) for each rated intervention component). Quality of working relationship (WAI-SR) Average > 36 (or a mean score of 3 or more [out of 5] on each of the 12 items). We will also descriptively examine goal, tasks, and bond subscales to see where mentor mothers are doing well and less well overall. Satisfaction (Client Satisfaction Questionnaire-8) Average > 24 (or a mean score of 3 on each of the 8 items) Intervention attendance 70% of all participants complete at the initial session and at least 2 of 3 booster sessions Intervention Fidelity MOSAIC Plus intervention achieves at least 80% adherence on a random subset of sessions Safety (all participants) Adverse events No serious adverse events or injuries that are possibly, probably, or definitely related to study participation.

Recruitment rate
Average of 4-8 enrolled per month MOSAIC Plus intervention procedures 80% of participants complete MOSAIC Plus intervention sessions at 3 months and 70% at 6-month follow-up Timeliness of assessments 80% of follow-up assessments occur within 3-weeks of due date Completeness of self-report instruments and interviews Self-report instruments have 80% of items completed and interviews 90% of items completed in 90% of cases Retention rate 80% complete 3-mo and 70% complete 6-mo follow up assessments Participant burden Qualitative responses from exit interviews do not suggest undue burden of intervention or research procedures . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted April 16, 2022. ; https://doi.org/10.1101/2022.04.14.22273885 doi: medRxiv preprint 1 DAS-II)(57) will be used to measure functioning in the domains of cognition, mobility, self-2 care, getting along/interaction with other people, life activities and participation in community 3 activities. A nine-item assessment tool measuring Self-care and Self-worth will be used assess 4 changes in self-care. The scale was developed based on our team's previous qualitative work 5 with high-risk women(58). Items included "I take good care of myself," "I need to take care of 6 people around me before I take care of myself," "I am worth taking care of," "I am worth 7 protecting," "I deserve to be able to take care of myself," "I deserve to protect myself," "I 8 neglect myself," "I put myself in dangerous situations," and "I take care of my needs," and 9 were answered on a 5-point Likert Scale from "Strongly agree" to "Strongly disagree." The 10 scale has an internal consistency reliability (Cronbach's alpha) of .77. To measure general 11 health and wellbeing, we will use SF-36 (59). SF-36 has a good internal validity in research, 12 community and clinical settings (59, 60).

13
Target mechanisms. We will measure social support using the RAND social support It is reliable, valid, and shows significant changes over time in clinical samples (63)(64)(65).

24
Provider perspectives. We will use Competency Assessment Inventory (CAI) to evaluate 25 the skills, attitude and knowledge needed to deliver a high quality mentoring intervention(66). 26 We will also use the provider version of Working Alliance Inventory(67) to assess mentor 27 mothers' perceptions of agreement on mentorship goals, agreement on mentorship tasks, and 28 the quality of the mentoring relationship with each participant mother. 29 30 Data Analysis. As this is an intervention development study, our primary goal is 31 assessment of feasibility and acceptability of the MOSAIC Plus intervention and research 32 procedures. However, pilot data can be used to demonstrate whether the effects of treatment 33 look promising across a set of outcome variables, to begin to examine distribution of outcome 34 variables to inform future analytic strategies, and to suggest, in concert with results from larger . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted April 16, 2022. ; https://doi.org/10.1101/2022.04.14.22273885 doi: medRxiv preprint 1 scale clinical trials in related fields, the range of effect sizes that would be reasonable to expect 2 in a future trial. We will obtain the between treatment condition effect size estimates (with 3 95% CI) at each assessment (e.g., Cohen's d or h) and the correlation between the same 4 dependent variable at adjacent assessments. We will have complete data on approximately 40 5 participants (the minimum expected after participant loss due to attrition). Sample size 6 guidelines for treatment development from Rounsaville et al. (68) recommend 15 to 30 7 participants per cell. Given that group means begin to stabilize by around 15, we believe a 8 sample of 20 in each condition should provide some information relevant to demonstrating 9 potential promise for MOSAIC Plus.
10 Mental Health Outcomes. We will: (1) Calculate the effect size and 95% CI for 11 reduction in depressive symptoms (PHQ-9 scores). Exploratory tests for differences between 12 conditions will use HLM, with baseline scores as a covariate. (2) Calculate the effect size and 13 95% CI separately for reduction in PTSD symptoms (DTS scores) using HLM with baseline 14 DTS score as a covariate.

IPV outcomes:
We will calculate the effect size and 95% CI separately for reduction 16 in CAS scores using HLM with baseline CAS score as a covariate.

17
Functioning outcomes: For secondary outcomes, separate calculations of the effect 18 sizes and 95% CI will be performed. These include improvement in functioning (WHODAS-19 12) self-care (Self-care and Self-worth scale) and general health and wellbeing (SF-36).

20
Separate exploratory tests for differences between conditions will use HLM, with baseline 21 scores as covariates.

22
Target mechanisms. We will separately calculate the effect size and 95% CI for the 23 effect of the intervention on proposed target mechanisms including improved social support 24 (measured using MOS-SS) and effectiveness obtaining resources (measured using EOR scale).

25
Separate exploratory tests for differences between conditions will use HLM, with baseline 26 scores as covariates. We will then explore the association of each of these target mechanisms 27 with changes in our primary outcome (reduction in depressive and PTSD symptoms) from 28 baseline through 9 months. These exploratory analyses will inform full tests of mediation (i.e., 29 tests of the hypothesis that the effects of MOSAIC Plus on reduced depressive symptoms is 30 mediated through improvement in social support and effectiveness obtaining resources) in a 31 subsequent fully powered trial. Although we expect the MOSAIC Plus intervention show 32 evidence of preliminary effectiveness, should the intervention show no or limited evidence, 33 exploratory tests of target mechanisms will also provide some initial information about whether 34 MOSAIC Plus' limited effectiveness was due to failure to engage target mechanisms or to the . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted April 16, 2022. ; https://doi.org/10.1101/2022.04.14.22273885 doi: medRxiv preprint 1 target mechanisms not being associated with the final outcome. Target mechanisms are 2 presented in table 2 below. Questionnaire) (83) as predictors and moderators of treatment outcome. These variables were 8 also used in the original MOSAIC study(24) and other studies with women with IPV 9 experiences(25, 46, 84). We expect that MOSAIC Plus will be appropriate for a full range of 10 perinatal women with IPV experience. Preliminary analyses of dose-response effects will also 11 be conducted.

12
Treatment Integrity. We will calculate scale reliabilities of adherence and competence 13 ratings using both individual item correlations and total intraclass correlations. We will 14 compute scale validity by correlating adherence and competence ratings to intervention 15 outcomes, to each other, and to expert (MH and JEJ) ratings. Adherence and competence 16 ratings of trained raters will be compared to expert global ratings to determine cut-off scores 17 with sufficient sensitivity and specificity.

18
Ethics and dissemination. The MOSAIC (MOthers' AdvocateS In the Community) for . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted April 16, 2022. ;https://doi.org/10.1101https://doi.org/10. /2022  training certificate and good clinical practice (GCP) trainings. The study RA will meet potential 5 participants in private locations at each agency. The study research assistant will explain all 6 aspects of the study to the participants and a written informed consent will be obtained. We 7 will ask participants if they would like us to read consent forms aloud. The research assistant 8 will emphasize that enrollment in the study is completely voluntary. Those who consent to 9 participate will be provided with a copy of the informed consent document. The study was also 10 registered in www.clinicaltrials.gov under identifier # NCT05106361, date of registration 03 11 November 2021, https://clinicaltrials.gov/ct2/show/NCT05106361.

12
Dissemination and data sharing. The research outcomes generated from this R34 will be 13 disseminated in a timely fashion. The study investigators are committed to reporting the study 14 results both in national and international platforms per the proposed timeline of the study. After 15 data have been collected and study results published, de-identified electronic data will be made 16 available to other qualified researchers upon request. The request will be evaluated by the 17 investigators to ensure that it meets reasonable standards of scientific integrity. For rapid, 18 responsible and broad data sharing, we will follow the standard NIH-funded clinical trial data 19 sharing procedures. We will also ensure compliance of our data sharing and reporting 20 procedures with the Michigan State University's internal policy regarding clinical trials 21 registration and reporting results. We will place the de-identified dataset, along with the data 22 dictionary and documentation of data collected, into the NIMH Limited Access Dataset 23 Repository. We will submit primary results for publication by the end of the project period, 24 and will have final de-identified datasets and data dictionaries available by CD and on the 25 NIMH Limited Access Dataset Repository within required timeframes. We are committed to 26 doing research that will change clinical practice. We will also work with our other community 27 partners to share results of the study with their state and national networks, through presentation 28 at their conferences and meetings, newsletters, flyers, and other strategies that they deem to be 29 appropriate. The current study is designed to provide evidence for the feasibility and acceptability of a 33 mentor mother-delivered intervention for IPV and depressive and/or PTSD symptoms among 34 pregnant women and mothers with children under the age of 5. Moreover, the study also aims . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted April 16, 2022. ; https://doi.org/10.1101/2022.04.14.22273885 doi: medRxiv preprint 1 to pilot research procedures for a subsequent fully powered trial comparing MOSAIC Plus to 2 the original MOSAIC.

3
Addressing IPV and its associated mental health symptoms (i.e. depression and PTSD) is 4 important because unaddressed mental health challenges and IPV increase the woman's risk of 5 mortality, morbidity, future child abuse, suicidal ideation, and femicide(84). Women with 6 elevated depressive and PTSD symptoms are at greater risk of experiencing future IPV(3, 85).

7
Moreover, the stigma of mental illness also reinforces abusers' ability to manipulate, control 8 and discredit survivors, and weaken vital social support(85). Therefore, depressive and PTSD 9 symptoms not only follow IPV and are associated with suffering and morbidity, they potentiate 10 future IPV risk. This bidirectional relationship between depression/PTSD and IPV suggests 11 that addressing these issues simultaneously may help end the vicious cycle of victimization and 12 mental health vulnerability. Therefore, an integrated intervention for reducing subsequent IPV,  To date, no study has found that an IPV focused intervention for any group of women 30 significantly reduced both future IPV risk and depressive and PTSD symptoms. Addressing 31 depressive and PTSD symptoms while effectively reducing future risk of IPV in mothers who 32 report IPV is crucial.

33
This study expands the literature because it will develop and pilot-test a paraprofessional-34 delivered intervention to address symptoms of maternal depression and PTSD in a highly . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) 1 vulnerable sample; pregnant women and mothers with young children who report experience 2 of IPV.   . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted April 16, 2022. . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted April 16, 2022. ; https://doi.org/10.1101/2022.04.14.22273885 doi: medRxiv preprint . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted April 16, 2022. ;https://doi.org/10.1101https://doi.org/10. /2022