Post-traumatic stress disorders in women victims-survivors of intimate partner violence: a mixed-methods pilot study in a French coordinated structure

Data source and study population

We carried out surveys from July 2020 to June 2021 in three FPCs: one in the Mdf and two in MHC from the same department (MHC-1 in Saint Denis and MHC-2 in Aubervilliers).

All women aged 18 years and over consulting in one of the three FPCs, having suffered or suffering from IPV and able to understand the objectives of the study, were eligible (interpreters could be contacted by phone if necessary). Trained research assistants (RAs) were available in each of the study centres to screen women for eligibility, explain the study and ask for a written informed consent before recruitment. Women under 18 years old or under tutorship were excluded. RA also assisted participants in completing the questionnaire.

We contacted each participant by telephone 6 months later.

Patient and public involvement

For security and confidentiality reasons, it was not appropriate or possible to involve patients or the public in the design, or conduct, or reporting, or dissemination plans of our research.

Outcome measures

Data were collected using a self-administered questionnaire that included questions about participants’ sociodemographic characteristics, as well as a range of health and substance use data.

The main outcome was a PTSD diagnosis, measured using the PTSD self-report checklist of 20 PTSD symptoms defined in the DSM-5 (PCL-5).11 PCL-5 is a widely used self-administrated questionnaire to detect and evaluate a PTSD, with a validated French version.12 Each item on this scale is rated on a 5-point Likert scale reflecting severity of a particular symptom from 0 (not at all) to 4 (extremely) during the past month, with a threshold score of 33.

Other outcomes included: the possibility of reaching women by phone 6 months after the inclusion visit, the level of self-rated global health (Likert scale: ‘very good’, ‘good’, ‘quite good’, ‘bad’ and ‘very bad’), the self-reported number of emergency visits in the past 6 months, the substances use: smoking status, alcohol (evaluated by the Alcohol Use Disorders Identification Test-Consumption/AUDIT-C13), drugs (‘Did you use hypnotics, sleep pills, antidepressants or anxiolytics in the past 6 months?’), the readiness to change, the safety behaviours (evaluated by questions inspired by the safety behaviour checklist14) and the help seeking behaviours in the past 6 months (evaluated by questions inspired by Van Parys et al15).

Qualitative interviews

We conducted semistructured interviews with a subsample of the participants in the MdF and in the MHC-1, according to the grounded theory (the interview guide is available in online supplemental figure). The interviews were all conducted by the same researcher, an MD qualified in qualitative research who used an interview guide. The interview guide was developed by the coauthors and reviewed and tested by two psychologists and one social researcher to verify the comprehensibility of the questions. The guide included questions about history of violence, women’s perception of the effect of the care provided at MdF and in the MHCs, women’s perception of their needs and their mental and physical health. Interviews were anonymised, transcribed, analysed and interpreted following practical guidance for conducting qualitative research.16

As recommended by the WHO, our study paid particular attention to minimising the risk affecting the safety of the respondents: confidentiality, safe climate at all time, informed consent, and basic care and support available locally for victims-survivors.17

Analysis

This mixed-method study used a convergent parallel design.18 Quantitative data analysis was conducted using SAS for Windows (V.9.4). To describe the sociodemographic characteristics, perceived social support, health and substance use indicators descriptive statistics were used consisting of frequency, percentage, and mean and SD.

As concerned the qualitative interviews, we conducted an inductive content analysis using a grounded theory approach.19 The qualitative data were analysed with NVivo V.12 software. The transcribed text was coded, then the codes were sorted into categories and main themes and were illustrated using verbatim quotations. We used a checklist of quality criteria (ie, credibility, dependability, conformability, transferability and authenticity) to improve the trustworthiness of our results.19

Sample description

A total of 67 women responded to our questionnaire: 40 in the MdF, 12 in the MHC-1 and 15 in the MHC-2.

The characteristics of study participants are described in table 1 (more detailed characteristics are described in online supplemental table 1). The majority of the participants (57%) were aged below 35 years, with a mean age of 34 (SD=9.7), and had at least one child (73.1%). Slightly more than half of participants (53.0%) were not born in France. Around 25% of participants (n=17) declared having no one to turn to for help or assistance if they needed it, while more than one-third (n=25) had at least two people to turn to for help.

Six months after inclusion, half of participants (52.2%) had been reached by phone (65.0% in MdF, 25.0% in MHC-1 and 40.0% in MHC-2).

Prevalence of PTSD

Participants reported an average PCL-5 score of 37.1 (SD = 16.6) (table 2). Forty women (59.7%) had a PCL-5 score of at least 33, which is the accepted cut-off value for defining the presence of PTSD diagnosis (table 2). The prevalence of PTSD was quite similar between the three groups.

Health and substance use outcomes

Around 40% of participants (n=26) self-rated their global health as a good or very good. The same percentage of participants reported consulting at an emergency room in the past 6 months.

Less than 30% (n=18) of women were regular smokers, and only 7.5% of participants had a problematic alcohol use with an AUDIT-C score greater than or equal to 4, one out of five women used psychotropic drugs.

Qualitative data

For this pilot study, nine women were interviewed (six in the MdF, three in the MHC-1) (online supplemental table 2).

They were aged 27–55 years old (mean age: 38.8), six were employed. Seven women had at least one child. Only one was in a couple, and they suffered from domestic violence between 1.5 and 13 years (mean: 5.4 years).

With regard to the perception of difficulties encountered by the women victims of violence, four main themes emerged from the thematic analysis: a feeling of loneliness, the need to be listened to, the specificity of the symptoms of the victims-survivors and the difficulties in accessing healthcare (figure 1).

Figure 1

Perception of the difficulties encountered by the victims-survivors of violence.

I spend all day long alone like this, with my thoughts, I don’t know where to go, and I’m still turning in circles… (MHC1)

In fact I think we should be in a bubble with psychologists all the time [laughs] to be listened and to feel that we’re not alone. (MHC2)

We need real professionals, who understand what we’re going through (MdF5)

I wanted to go to another support group but I’ve been told that I have to wait because there are too many people… I cried not because there was no room for me but because we are so many, and there is no room for anyone… (MdF3)

With regard to the perception of the specific care of the MdF, six main themes emerged. Four of them correspond to the four themes developed in figure 1: social networking, multidisciplinary approach, specialised listening, healthcare facilities and the other two themes highlight additional advantages provided by the MdF: evasion and ‘feeling at home’ (figure 2).

Figure 2

Perceptions of the specific care of the Maison des Femmes.

We live in a society where we are forced to believe that women are our competitors… But here we are sisters. (MdF5)

That’s what interested me here, being able to get out of my head and to concentrate on a physical activity […] it makes it possible to find an appeasement, what I call a “air bubble” so I can restart (MdF4)

I’m in a house here, it’s a house that is made for us […]. There’s a roof like a house, I mean it’s friendly, at first I said to myself “what am I doing here?” […] and finally every time I come I can say everything, I feel like it won’t come out the walls, I feel like whatever I say I won’t be judged. (MdF6).

Prevalence of PTSD

These results are consistent with those reported by other authors who have described an association between the exposure to IPV and the presence of PTSD.20 21 The prevalence of PTSD among victims-survivors of IPV varies depending on the studies and on the tool used to quantify PTSD, ranging from 33% to 84%, with a mean of 61%.22

To assess the benefit of a multidisciplinary and cooperated approach on the mental health of the victims-survivors, as the MdF provides, a comparison between the three centres a few months after the inclusion with a repeated measure of PTSD would be advisable. The fact that the prevalence of PTSD at inclusion is almost the same between the three centres in our study seems to eliminate centre bias, and encourages us to consider a larger comparative study aimed at comparing the long-term impact of the MdF on the mental health of victim-survivors with that of standard-of-care structures. This study, the IROND-L study (‘Evaluation of the impact of the care of women victims of sexist and sexual violence, according to a coordinated multidisciplinary approach in women’s homes or traditional health centres or family planning, on mental and physical health: a prospective, quasi-experimental, multicentre, national study’), has just been funded by the French Ministry for Health, submitted to an ethic committee and is likely to start early in 2024. We were unable to contact a majority of women from the MHCs 6 months after inclusion, whereas two out of three women from the MdF had been reached. This could reinforce the fact that women may value the care provided by the MdF more than that provided by non-dedicated structures, but it will present an additional difficulty for a subsequent comparative study.

A multicomponent model

The MdF is a structure that provides multicomponent trauma-informed and holistic care. Getting out of IPV is a process with multiple stages.23 Our qualitative results reinforce the fact that MdF seems to fit to the needs of the victims-survivors throughout their trajectory. MdF supplies essential interventions recommended by the WHO to prevent VaW.24 These interventions correspond to models presented as highly efficient to improve the mental health of IPV survivors.25 On the top of that, respondents also described the MdF as a warm place where they could escape from reality, a new concept that needs to be explored in the future.

Perspectives

This study will serve as a basis for a larger comparative trial of the long-term impact of specialised care of the MdF on PTSD compared with the care in non-specialised structures: the ‘IROND-L’ study. We are planning a quantitative and a qualitative component, as well as a medico economic study. The main objective aims to assess the evolution of PTSD between the initial visit and 6 months later in women victims-survivors of violence, according to whether they are treated in structures offering a coordinated multidisciplinary approach (MdF), or in health centres or family planning. We also aim to assess the presence of sleep disorders, the quality of life, the presence of depressive and anxiety symptoms substances use, the reason for seeking care, and the women’s perception of their safety and well-being and that of their children.

In the pilot study, the mean PCL-5 score was=37.1 (SD=16.6). Therefore, a minimum of 150 women per group is required to achieve a power of 80% and 5% significance level (two sided), to detect a mean difference of 5.5 between the two groups, assuming that the SD of the differences is 17 (figure 2). Thus, 180 women per group will need to be recruited to account for a potential 35% rate of loss to follow-up. The IROND-L study will include 360 women victims of violence and will be conducted in 5 metropolitan department, and we hope it will increase generalisability of the results. However, generalisability may not be transposable as the MdF approach is new and quite unique worldwide.

Lastly, the future qualitative component will also require much greater recruitment to interview more profiles that are different and approach data saturation, which this pilot study could not achieve.

Strengths and limitations

This is the first French study assessing the prevalence of PTSD in victims-survivors of IPV in MHCs and the MdF, is the first French structure dedicated to the care of women victims of violence. The high prevalence of PTSD outlined in our study justifies the need for launching larger quantitative and qualitative researches on the mental health of the victims.

However, this study has several limitations. First, it was conducted in health facilities, and therefore, does not include women in IPV situations who do not have access to healthcare or those who face significant barriers to seeking care. Nevertheless, we do not believe that our population is biased towards more advantaged women. Indeed, even though limited, our sample embraces a wide range of situations, with 53% of women born outside of France, 37% having social security coverage reserved for those with no or low income from work or illegal immigrant status, suggesting that the facilities where the study was conducted have a broad recruitment base.

Even if we have no formal explanation for the low follow-up rate, financial difficulties have been described as an important factor in the loss to follow-up.26 As our study took place in the poorest area in France, we had anticipated, but without fair estimate, that the follow-up rate would be low. It was indeed one of our objectives to provide data to design the IROND-L study.

It is known, and we have shown10 that violent episodes occur more frequently during a break-up phase (separation, job search) and that women who are victims of violence are, therefore, logically more likely to move house or change their telephone number in order to escape their violent partner. It is, therefore, logical that these women are more difficult to monitor.

Moreover, we decided to focus on domestic violence in this pilot study and did not collect data on other stressful or traumatic events. Only the qualitative part of this study explored the duration and/or repetition of the violence and/or the duration since the possible end of the violence. This information will have to be considered and collected in the future quantitative and qualitative questionnaires of the larger IROND-L comparative study.

Finally, the health outcomes measured in this study are based solely on women’s self-reported perceptions, rather than on potentially more valid clinical observations. These limitations will be addressed in the future comparative study using a quasi-experimental design, where care pathways and the consumption of medical goods and services, will be assessed based on medical records and health insurance database.