Participants

Ethical approval

This study received ethical approval from the Joint South London and the Maudsley and the Institute of Psychiatry NHS Research Ethics Committee (ref 07/H0807/66).

Analysis

Interviews were analysed thematically. ^{Reference Braun and Clarke13–Reference Attride-Sterling15} It was not assumed that themes would ‘emerge’ from the data but that interpretive work would be needed to identify them. The analysis began with identification of patterns in the data that were coded by two raters (A.W. and K.T.). An initial coding frame was developed by the two raters based on simultaneous coding of five service user and five professional transcripts and supplemented by codes taken from the interview schedule. The appropriateness of the coding frame was checked through progressive iterations and reapplied to earlier transcripts as it developed. In total, 20% of the coding was cross-checked to ensure reliability and 81% agreement attained by the two raters. Nvivo 8 (2008, QSR International Pty Ltd, www.qsrinternational.com) on MS Windows was used for indexing material and for retrieval of text chunks pertaining to the same or similar codes. Data that did not seem to fit into the coding frame were actively sought and multiple coding conducted by two raters. ^{Reference Barbour16} Codes with similar information were merged and pruning of irrelevant codes and an initial identification of themes was conducted. Subsequently, these themes were themselves interrogated and a final level of abstraction was reached in the form of interpretive themes, with a complementary deviant case analysis where individual transcripts were interrogated for information that seemed to be discrepant with our overall analysis. At each step, it was ensured that saturation of themes had been reached. Concept maps were developed to represent the data in a visual format. The diagrammatic representations in the maps reflect the superordinate themes and subthemes that emerged from the data after repeated coding and refinement of the framework.

Dominant themes

The dominant themes from the interviews concerned disclosure of domestic violence. The two conceptual maps representing the two interpretive themes and their associated subthemes are given in Figs1 and2. For service users, the dominant theme was barriers to disclosure of domestic violence. For professionals the dominant interpretive theme was barriers to enquiry about domestic violence. The presentation of results will be given in the following order: themes from the data supplied by service users; themes arising from the interviews with professionals; and themes which overlap between the two groups.

Service users

As can be seen from the conceptual map (Fig. 1), fear is an important and multifaceted subtheme for service users when considering disclosure of domestic violence to mental health professionals. It encompasses fear of Social Services involvement; fear that disclosure will not be believed; fear that disclosure will result in further violence; fear of disruption to family life; and fear of the consequences for immigration status. Some illustrative quotes follow.

A further cluster of subthemes concerns blaming attitudes of others, self-blame, and shame and embarrassment, which again makes victims reluctant to disclose abuse.

Fig. 1

Service users’ barriers to disclosure of domestic violence.

Unsurprisingly, this was linked to psychological distress.

A further cluster of subthemes concerned not recognising the behaviours as abusive.

Service users talked about perpetrators’ actions to prevent disclosure and about how this could isolate them from friends and family as the perpetrator did not want them to know about the abuse.

Service users were critical of services’ failure to respond to signs of abuse and on the other hand they stressed the importance of engagement between professionals and service users if disclosure was going to be possible. This theme emerged in the professional data as well and is dealt with below.

Mental health professionals

As with service users, the interpretive theme that emerged was also disclosure, with an emphasis on barriers to asking clients about any abuse that they have experienced (Fig. 2). The dominant subthemes in this data are clear and are to do with whether enquiry about domestic violence is part of their role and within their competence.

The issue of professional roles is often linked to a medical diagnostic and treatment model in the sense that the professionals sought to focus on mental health symptoms.

In the following quote, a psychologist states that enquiry about domestic violence is not part of their role.

A contrasting view is described in the quote below from a consultant psychiatrist.

Competing demands, time constraints and the presence of the partner at consultations were all also mentioned as hindering enquiry into domestic abuse in a client's life. Many professionals talked about the importance of competency, and training, which many felt they had not received, as well as the boundaries of their role.

However one professional, a team manager, felt that colleagues in the team did have the expertise to identify domestic violence.

Fig. 2

Professionals’ barriers to enquiry of domestic violence.

Linked to competency, many staff expressed unease and lack of confidence in dealing with a client's experience of domestic abuse.

However, a small number of experienced staff, who had been qualified between 11 and 29 years, described feeling confident in approaching the subject of domestic violence.

Further analysis showed that six professionals stated they did enquire about domestic violence on occasion; this was not associated with gender, age, professional background or years qualified. However, these six professionals appeared to differ in their ‘interpretation of their roles’, compared with other professionals.

However, for some professionals, routine enquiry in relation to perpetration of violence was felt to be an easier subject to address with clients than enquiry about the experience of violence.

Although less evident than in the service user interviews, there was also a cluster of subthemes on the issue of fear. Professionals were worried that clients would be offended if they asked about domestic abuse, that there might be consequences for the victim or that discussing the experience might be traumatising for the client.

Professionals, like the service users, also identified factors that could facilitate discussion of domestic violence, including the importance of therapeutic engagement and a supportive relationship.

Overlapping themes between service users and professionals

It has already been seen that the issue of fear is common to service users and professionals although in different ways. A further common theme concerns therapeutic engagement. Both service users and professionals believed that a good and supportive relationship between them and their clients facilitated disclosure and the exploration of domestic abuse.

Therapeutic engagement is generally conceptualised as something developed over time but there was an example of a service user immediately ‘hitting it off’ with a worker.

We have already seen that some professionals queried whether asking about domestic abuse was part of their role. They saw themselves as dealing primarily with the client's mental health problems and ensuring stability. Both service users and professionals talk about the influence of the ‘medical diagnostic and treatment model’ on willingness to disclose or ask about domestic violence.

Issues of gender were evident in the interviews of both service users and professionals. The following male member of staff made his ‘excuses’ on the grounds of lack of gender matching.

In the following response, issues of both gender and culture appear.

Culture appeared as a strong theme in both service user and professional interviews.

Main findings

This is the first study to look at the views of both service users and mental health professionals in relation to disclosure of domestic violence. Our main findings were that service users were reluctant to disclose domestic violence because of their concerns about the potential consequences, and mental health professionals reported finding enquiry about domestic violence difficult because of their lack of knowledge and expertise in this area or because they did not think it was part of their role.

For service users an important barrier to disclosure was fear evoked by a number of factors. Men and women were afraid of the potential consequences of disclosure of domestic violence to mental health professionals and other professionals such as general practitioners, other doctors, the police or social workers. Women feared that if they disclosed to health professionals the perpetrator might come to know this and they could be at increased risk of violence or their children would be removed from them. Non-disclosure can therefore be a very rational decision based on the service user's knowledge of the risks and benefits of disclosure. These findings are similar to those reported by service users in other healthcare and non-healthcare settings. ^{Reference Feder, Hutson, Ramsay and Taket17} Service users also feared that they would not be believed if they disclosed domestic abuse because professionals would focus solely on their symptoms. One professional instead expressed concern that her seniors assumed that the client was not ‘telling the truth’. There is, however, evidence that psychiatric patients tend to underreport rather than overreport experiences of violence. ^{Reference Goodman, Thompson, Weinfurt, Corl, Acker and Mueser18} It is common in situations of domestic abuse ^{Reference O'Neill and Kerig19} for women to believe that the abuse must somehow be their fault and we found this in some of our interviews. Self-blame is exacerbated when the victim has a mental health problem as she may see this as part of her own involvement in provoking the abuse.

The main themes for professionals concerned role boundaries, competency and confidence. Professionals queried whether enquiry about domestic abuse was part of their role at all, as an assessment should focus on mental health. One psychiatrist expressed exasperation that so many recommendations were made by different stakeholders that they could not possibly all be accommodated. It is noteworthy that the team manager's view of staff competence in dealing with domestic abuse was in direct contrast to the staff themselves, who expressed a lack of expertise in addressing the topic of domestic violence and called for more training. Quantitative research involving a survey of US clinicians to identify barriers to the diagnosis and treatment of post-traumatic stress disorder (PTSD) in people with severe mental illness, found that these issues of competence and confidence were associated with the percentage of clients with whom trauma and PTSD had been discussed, documented and addressed directly in treatment; ^{Reference Salyers, Evans, Bond and Meyer20} lack of confidence and competence inevitably has implications for the ability of clinicians to address domestic violence. Finally, professionals talked about this being a ‘delicate’ issue with which they felt uncomfortable. Generally, the barriers to enquiry highlighted by the community mental health professionals in this study were similar to those reported in a systematic review of studies of barriers to screening for domestic violence in other healthcare settings: lack of training regarding domestic violence, lack of time, lack of effective interventions and patient-related factors such as non-disclosure and fear of offending the individual. ^{Reference Waalen, Goodwin, Spitz, Petersen and Saltzman21}

Service users and professionals expressed similar ideas about facilitators of disclosure. Both believed that a supportive and therapeutic relationship was necessary for enquiry and disclosure so that the issue could be explored further. This has also been stressed by women interviewed in primary care qualitative studies on domestic violence ^{Reference Feder, Ramsay, Dunne, Rose, Arsene and Norman22} and is clearly an important component of effective enquiry in clinical practice. However, research in mental health settings employing standardised screening tools to measure domestic violence suggest the experience of being asked about domestic violence may be sufficient to facilitate disclosure, despite not having an established relationship with the interviewer. ^{Reference Howard, Trevillion, Khalifeh, Woodall, Agnew-Davies and Feder4} Some service users and some professionals believed that the ‘medical diagnostic and treatment model’ was an obstacle to discussion of domestic abuse as the focus was always on symptoms. Gender and culture were subthemes in both sets of interviews, with particular emphasis on the perception that some cultures believed domestic abuse to be acceptable.

One professional made the point that professionals were used to asking people whether they had ever been violent or had a propensity to violence as this is a routine part of a mental health risk assessment but felt that professionals were less comfortable enquiring whether a client had been a victim of violence. The emphasis on the risk of violence by people with severe mental illness rather than on their (higher) risk of experiencing violence appears to be related to the stigma of mental illness, which, it has been suggested, has ‘turned the world on its head’. ^{Reference Eisenberg23} Mental health service users are 11 times more likely to have experienced recent violence than the general population ^{Reference Teplin, McClelland, Abram and Weiner24} and we therefore recommend that risk management should focus more on the risk of experiencing violence. It appears that at present professionals may not routinely enquire about such experiences. In this paper, we have reported some of the barriers and facilitators to enquiry and we are using this data to inform the development of interventions in mental health services to address domestic violence.

Limitations

This study was carried out in a very socioeconomically deprived setting where there is a high proportion of people from Black and minority ethnic groups; our results may not therefore be generalisable to other settings. As with other research, selection bias means the views of those who did not wish to participate in the study are not known and may be different.

Clinical implications

Disclosure of domestic violence is facilitated by a good service user–professional relationship and is likely to be facilitated further by domestic violence training of professionals. Training interventions in healthcare settings can improve identification and specialist referral of people experiencing domestic violence. ^{Reference Ramsay, Feder and Rivas25} A recent UK Department of Health initiative (the Victims of Violence and Abuse Prevention Programme, VVAPP) plans to increase training on abuse throughout mental health services and recommends that all service users are asked about violence and abuse as part of routine assessments. ^{Reference Itzin, Bailey and Bentovim26} However, the focus of the VVAPP is on training to address childhood sexual abuse. Current domestic violence may be more strongly associated with mental health symptoms than childhood sexual abuse ^{Reference Coid, Petruckevitch, Wai-Shan, Richardson, Moorey and Feder27} and will need different interventions if the violence is still occurring.

Training needs to address the deficits in knowledge identified in this study and include practical examples of how to ask about domestic violence, information on the needs of service users and details of local domestic violence support services. The current evidence base for interventions to support mental health service users experiencing domestic violence remains sparse. ^{Reference Howard, Trevillion, Khalifeh, Woodall, Agnew-Davies and Feder4} It is necessary therefore to develop and evaluate clear care pathways, involving the domestic violence sector when appropriate, to help mental health services address this underdetected but potentially life-threatening issue.