Introduction
Data from the Crime Survey of England and Wales suggests that ~4% of men report experience of one or more incidents of partner abuse in the last year (Britton, Reference Britton2012) with 11% reporting having experienced ‘non-sexual partner abuse’ since the age of 16 [Office for National Statistics (ONS), 2013]. It is important to note, however, that the measurement of the prevalence of domestic violence and abuse (DVA) in general populations surveys, with focus only on incidents of abuse rather than impact, may not be useful in the clinical setting (Williamson, Reference Williamson2012; Hester, Reference Hester2013).
Focusing on evidence from a clinical context, male patients, as victims or perpetrators of DVA, seek help from health services but usually do not spontaneously disclose abuse. The British Crime Survey (Hoare and Jansson, Reference Hoare and Jansson2008) reported that 18% of self-identified male victims of DVA reported seeking medical assistance. A UK survey of 175 male patients in four practices reported that 15% had experienced potentially abusive behaviours and 16% self-identified as perpetrators. In that study, the majority of male patients thought it was helpful for men to be asked about perpetration and experience of DVA and 73% said they would be comfortable discussing the issue with their general practitioner (GP) (Westmarland et al., Reference Westmarland, Hester and Reid2004). A US study of service use by perpetrators of DVA (Coben and Friedman, Reference Coben and Friedman2002) reported that 42% of those arrested for DVA perpetration had seen their health provider in the six months before arrest. This is consistent with findings from the United Kingdom (Hester et al., Reference Hester, Westmarland, Gangoli, Wilkinson, O’Kelly, Kent and Diamond2006), where 32 (from 45) perpetrators on a DVA perpetrator programme reported having consulted their GP before the programme attendance. While the evidence above might suggest that general practice could be a useful setting in which to address male patients’ experience and perpetration of DVA, the issue of men and DVA has received little attention over the years that may go some way to explaining why the issue is not higher on the clinical radar of practitioners (Hines and Douglas, Reference Hines and Douglas2009). Recent NICE guidelines [National Institute of Clinical Excellence (NICE) 2014] focusing on DVA and its impacts on all patient groups, has identified the need for clinicians to address the needs of male patients as both victims and perpetrators of abuse.
Background
A number of studies have identified and explored different barriers to effective clinician responses to DVA, although much of this research has focused on responses to predominately female patients. The barriers identified within this research include: inadequate preparation and training (Gutmanis et al., Reference Gutmanis, Beynon, Tutty, Wathen and MacMillan2007); perceived complexity of the issue in relation to perpetrators (Miller and Jaye, Reference Miller and Jaye2007); concerns about treating members of the same family where DVA might be an issue (Hegarty et al., Reference Hegarty, Taft and Feder2008); a reported lack of information about appropriate referral agencies (Miller and Jaye, Reference Miller and Jaye2007); concerns about differentiating between victims and perpetrators (Kimberg, Reference Kimberg2008); inconsistencies in the documentation of DVA within the medical record (Williamson, Reference Williamson2000; Miller et al., Reference Miller, Thow, Hall and Martin2005); time constraints in terms of availability for training and within patient consultations (Gadomski et al., Reference Gadomski, Wolff, Tripp, Lewis and Short2001; Gutmanis et al., Reference Gutmanis, Beynon, Tutty, Wathen and MacMillan2007); a reported lack of confidence in knowing how to respond to DVA (Miller and Jaye, Reference Miller and Jaye2007); attitudinal and preconceived ideas that might be detrimental to effective practice (Lo Fo Wong et al., Reference Lo Fo Wong, Wester, Mol and Lagro-Janssen2006); and uncertainty about the role of health clinicians in relation to DVA (Yeung et al., Reference Yeung, Chowdhury, Malpass and Feder2012).
When addressing the needs of male patients it is often the perceived need to differentiate between ‘genuine’ victims and perpetrators that raises concern for clinicians and creates a barrier to effective practice. Evidence collated from the National DVA helpline for male victims, run by the domestic violence charity RESPECT (RESPECT, 2010), indicates that of 1441 men who contacted the helpline between 2007–2010, helpline workers determined by the end of each call that only 56.5% of callers would be classified as victims of abuse. Of the remaining callers 13% were describing an unhappy relationship where abuse was not present; 1.6% talked about mutual violence; 16.3% described incidents where they had perpetrated abuse; and in 12.5% of cases the helpline worker, despite extensive training on the issue, was still not sure. This highlights how distinguishing between the role of victim and perpetrator is complex and understandably raises concerns for clinicians, and that it should be carried out by DVA experts rather than clinicians.
These barriers also highlight the need for training to give practitioners the tools to address DVA within their clinical practice. Different training models have been trialled that includes the development of core curriculum principles (Brandt et al., Reference Brandt2007), interventions aimed at improving the identification and management of DVA (Thompson et al., Reference Thompson, Rivara, Thompson, Barlow, Sugg, Maiuro and Rubanowice2000), and focus group training approaches (Lo Fo Wong et al., Reference Lo Fo Wong, Wester, Mol and Lagro-Janssen2006). A training and support intervention for female patients in UK general practice was tested in a cluster randomised trial of 48 practices: Identification and Referral to Improve Safety (IRIS) (Feder et al., Reference Feder, Davies, Baird, Dunne, Eldridge, Griffiths, Gregory, Howell, Johnson, Ramsay, Rutterford and Sharp2011). The intervention, delivered by an advocate-educator based in a third-sector organisation, included two 2-h training sessions for doctors and nurses, a 1-h training session for administrative staff, a referral pathway for women who disclosed DVA, and feedback to practices. The IRIS trial found a threefold difference in the identification of women experiencing abuse and a sixfold difference in referrals to the specialist local DVA agencies between the intervention and control sites.
The aim of the HERMES study was to evaluate the feasibility of a training intervention for family practice extending the IRIS model to improve the response to male patients who had experienced or perpetrated DVA. The objectives were: (i) development of the intervention, (ii) delivery of the intervention to four general practices, and (iii) mixed method evaluation of its feasibility.
Methods
Development of the intervention
The intervention was developed through meetings between the research team, primary care, and specialist DVA professionals. The content was based on RESPECT training principles (RESPECT, 2013) and streamlined within the process defined by the IRIS model (Feder et al., Reference Feder, Davies, Baird, Dunne, Eldridge, Griffiths, Gregory, Howell, Johnson, Ramsay, Rutterford and Sharp2011). The training included information about prevalence and associated presenting complaints of male patients experiencing or perpetrating DVA; appropriate responses and nature of the referral pathway; the acting out of enquiry about abuse; and a discussion of issues potentially impacting on men who have sex with men. The intervention used a flowchart with all the pertinent information from the training including prompts on how to ask about abuse, symptoms suggestive of abuse issues, and a clear referral pathway (Supplementary data 1). Unlike the IRIS intervention (Feder et al., Reference Feder, Davies, Baird, Dunne, Eldridge, Griffiths, Gregory, Howell, Johnson, Ramsay, Rutterford and Sharp2011), HERMES did not include a direct referral pathway to a local specialist DVA advocate. Instead, HERMES clinicians were advised to refer male patients to the two national RESPECT helplines for victims and perpetrators respectively by giving patients a card with both helpline numbers, one on each side.
Delivery of the intervention
A total of four general practices in Bristol that had already received the IRIS training were given the 2-h HERMES session by a GP and DVA specialist, using the IRIS interactive adult learning method. Clinicians in each practice were provided with the HERMES flowchart and the two-sided referral information cards. A prompt to ask about DVA was linked in the electronic medical record to conditions associated with DVA, such as anxiety and depression. Reception and administrative staff were also offered a 1-h information session on the intervention. Clinicians were also offered follow-up training.
Evaluation of feasibility
A range of methods were used to evaluate the feasibility of the intervention. This included measuring change in clinicians’ practice and behaviour through the PROVIDE Intervention Measure (PIM); data extracted pre- and post-intervention from the medical records of the participating practices; referral data from the third sector specialist agency identified within the referral pathway; and semi-structured interviews with participating clinicians. Each of these are outlined below.
-
(i) The PIM (Supplementary data 2 and 3) was developed from the validated PREMIS questionnaire (Ramsey et al., Reference Ramsay, Rutterford, Gregory, Dunne, Eldridge, Sharp and Feder2012), and edited down to include only questions about clinical practice and behaviours. As such, the version used within this study was not a validated measure. Practitioners undertaking the training were asked to complete it before the training and again after six months. Data from both time frames were entered into an Access database and analysed in STATA 12. We computed a Wilcoxon signed-rank sum test of the differences in responses between pre- and post-PIM training, imposing no distributional assumptions on the differences (Armitage et al., Reference Armitage, Berry and Matthews2002). We also report baseline and follow-up median scores, as well as a bootstrapped median of the differences (with 95% confidence interval), computed by re-sampling observations 50 times.
-
(ii) DVA identification data were extracted from electronic medical records of HERMES practices using codes designating DVA, six months before and after the intervention. For the analysis we compared numbers of patients identified as experiencing or perpetrating DVA. Cases that were identified through the searches were individually checked to ensure that they related to DVA cases. This procedure was carried out by a member of the clinical team in the presence of a researcher in order to adhere to ethical concerns about accessing personal information without the consent of individual patients.
-
(iii) Referral data were collected from RESPECT, who run the two national helplines specified within the referral pathway. Helpline staff asked all callers from the study city if they had heard of the service from their general practice and, if so, which one. To ensure the counting of patients who chose to self-refer to local DVA services, two local services, one for male victims the other for male perpetrators were contacted and data were collected.
-
(iv) A total of seven semi-structured telephone interviews were conducted with HERMES clinicians who had completed a post-intervention PIM measure and who agreed to take part. Analysis involved a thematic analysis of the transcribed interview transcripts. Limited summary data from these interviews is presented in this paper.
The study was approved by the South West research ethics committee (reference 10/H0106/22).
Results
PIM survey
A total of 25 out of 34 practitioners who attended the training completed the baseline PIM survey. Respondents were aged between 31 to 60 years (median 47, SD 7.7), 16 were female and nine were male. In all, 19 respondents were GPs, five practice nurses, and one respondent was a health advisor. Respondents had between 4 and 30 years’ experience working in family practice (median 15, SD 8.1), and reported seeing between 21 and 300 patients a week (median 80, SD 56.2).
Respondents reported 0–10 h previous training on DVA (median 4, SD 2.9) that included watching a video (n=2), completing a web-based programme (n=2), attending a lecture or talk (n=14), skills-based training (n=6), medical classroom-based (n=3), or clinic-based (n=2) training. The majority of respondents (n=19) had received training as part of the IRIS project (Feder et al., Reference Feder, Davies, Baird, Dunne, Eldridge, Griffiths, Gregory, Howell, Johnson, Ramsay, Rutterford and Sharp2011). Previous training included information about women as victims of DVA (n=24), heterosexual men as victims (n=11), heterosexual men as perpetrators (n=14), and men having sex with men as either victims or perpetrators of DVA (n=6).
Comparison of pre- and post-intervention responses (14 respondents) suggests an improvement in perceived ability to manage situations where patients either disclosed, or the clinicians suspected, exposure to DVA. There were statistically significant improvements in perceived competence in responding to heterosexual, gay, or bisexual male patients, but no significant change in dealing with female patients who may have been victims, the exception being ability to respond to disclosure (Table 1). This is consistent with all participating practices being part of the IRIS programme, and having previous training in relation to women experiencing DVA.
Table 1 Change in practitioners’ self-reported ability to identify and respond to DVA
This table reports number of observations; the non-parametric signed-rank sum test of differences in the distributions between baseline and follow-up measurements; baseline and follow-up medians; and the bootstrapped median of the differences with its 95% confidence interval (CI) (the median of the differences and its interval are bootstrapped, ie, calculated by repeatedly sampling observations a number of times, in this case 50).
Recording of DVA identification
DVA identified and documented in medical records across all four practices in the six months before and after the training is shown in Table 2. A total of six male patients were identified before the intervention as experiencing or perpetrating intimate domestic violence compared with 17 male patients during the post-intervention period [intimate partner violence (IPV) is identified here to differentiate from domestic abuse experienced from wider family members]. According to information from the medical records, four of the six male patients identified before the intervention were classified as perpetrators of IPV, one as a possible perpetrator of IPV, and one as a victim of wider family abuse (by his brother). Post-intervention, 2 of the 17 patients were identified as victims of IPV, four as victims of wider DVA including other family members. In all, seven men were clearly identified as perpetrators of IPV and four male patients were identified as possible perpetrators of IPV.
Table 2 Number of male patients documented within the medical record as experiencing or perpetrating DVA
DVA=domestic violence and abuse.
a Four of these six victims were victims of DVA within the wider family context and not DVA within an intimate partnership.
Contact with RESPECT helplines and local services
RESPECT, who run helplines for male victims and perpetrators of DVA identified three referrals to their helplines via HERMES practices during the data monitoring period. Other patients could have accessed other help offered by the service through their website. Direct contact had been made to the local domestic violence perpetrator programme from two patients from the study practices. This resulted in a total of five patients, from HERMES active practices, contacting support services following the intervention.
Interviews with practitioners
Views on domestic violence and male patients
When discussing male patients as perpetrators or victims of DVA, informants claimed this was not a common problem for their patients. As one female doctor stated:
I think the training was fine […] we just don’t seem to see it as much, the sort of male issues. Whether that’s because it’s not as common or whether they’re just not coming, […] People just don’t pitch up, even though my radar you know is more aware than it was before, it just doesn’t seem in practice to be as common (EW03).
Another caveat to HERMES raised by some clinicians was that they tended to see mostly female patients (EW05), or predominantly older patients that one informant perceived as not being at risk of experiencing or perpetrating DVA. In both these cases there was an assumption that, while it was useful to be aware of the issues raised in the HERMES training this was unlikely to be a regular aspect of their work. However, there is no research evidence that age is a protective factor for older patients experiencing DVA.
Also evident was an assumption, when discussing male patients who might be victims of abuse, that this would more likely affect men who have sex with men. When asked about men in general, three of the seven clinicians interviewed responded by referring to gay men. These clinicians were equating male victims of abuse with gay men despite this not being the focus of the questions or the HERMES training. While it is positive that practitioners recognise that abuse can occur in men who have sex with men relationships (Hester et al., Reference Hester, Williamson, Regan, Coulter, Chantler, Gangoli, Davenport and Green2012; Donovan and Hester, Reference Donovan and Hester2014), this becomes problematic if there is an underlying assumption that the impact of DVA on men is limited to gay men.
Safety concerns
One of the issues that arose in interviews and is reported in the wider literature discussed above, is a concern about the safety of clinicians themselves.
I suppose it would only be you know I wouldn’t feel particularly safe raising it with someone that was coming across aggressive. You know so I would choose not to, you know, and we do ha- we do have one patient at the moment that’s very anti female doctors (laugh). So, there’s nothing that I can think of inherent in me that would be a barrier. It would be more whether I would choose not to on safety grounds if say a potential perpetrator was perpetrating (laugh) in the room (Female doctor, EW03).
Addressing the safety concerns of clinicians is crucial in enabling them to raise issues about anger and violence with patients who might then direct that aggression towards them. Removing a requirement for clinicians to differentiate between victims and perpetrators helped, but did not eliminate these concerns completely.
Differentiations in the use of codes
During the collection of DVA disclosure data from individual patient records it became apparent that different clinicians used abuse codes differently. We asked clinicians about how they used DVA codes in medical records and their responses support the case for more research into practitioners’ current coding practices. As one doctor stated: ‘I probably don’t code it if I’m perfectly honest. I probably free text it in to the problem heading’ (EW01). This raises concerns about how inconsistencies in documentation might be impacting on clinical practice and research (Miller et al., Reference Miller, Thow, Hall and Martin2005).
Time for training
Most clinicians felt that the HERMES training would preferably be incorporated in a comprehensive package that also includes training on women experiencing DVA, specifically within the IRIS programme. While this preference was dictated often by concerns about time constraints, it also raises questions about how best to integrate training on tackling DVA with all patient groups.
Clinicians also highlighted the need for top-up training. ‘I think that in order for the message to stick, we need to be reminded regularly’ (SJ01). Both HERMES and the IRIS intervention had offered top-up training; however, this was not always practical to arrange, given the competing training demands of busy practitioners.
Conclusions
The findings from this pilot suggest that an intervention focused on the issue of DVA and male patients resulted in a number of positive outcomes. This pilot supports previous evidence from studies of female patients (McCaw et al., Reference McCaw, Berman, Syme and Hunkeler2001; Lo Fo Wong et al., Reference Lo Fo Wong, Wester, Mol and Lagro-Janssen2006; Gutmanis et al., Reference Gutmanis, Beynon, Tutty, Wathen and MacMillan2007; Feder et al., Reference Feder, Davies, Baird, Dunne, Eldridge, Griffiths, Gregory, Howell, Johnson, Ramsay, Rutterford and Sharp2011) that when provided with the relevant tools, through training, clinicians felt more prepared to address the needs of male patients experiencing or perpetrating DVA. Although the number of men identified from the medical records was low (n=20), there was an increase in documentation at the post-intervention point, which is a positive outcome. An examination of the specific cases within the medical records data revealed some interesting examples of good practice where clinicians followed up initial discussions at subsequent appointments asking whether things were any better, or if the patient had accessed the specialist support. We also identified cases of male patients disclosing their concerns about the impact of their behaviour on the wider family. The low number of male patients who contacted the referral and local organisations (n=5) raises additional questions about whether they might benefit from a local advocate referral (as in IRIS Feder et al., Reference Feder, Davies, Baird, Dunne, Eldridge, Griffiths, Gregory, Howell, Johnson, Ramsay, Rutterford and Sharp2011) or whether more research is needed to establish how male victims and perpetrators are accessing and using support services.
An important secondary outcome of the research was that following the HERMES training there was an increase in the identification and documentation of female victims of DVA during the post-intervention period despite this not being the focus of the intervention training. With regard to the barriers reported in the wider literature, this pilot was able to address, through the training, some of the concerns reported by practitioners. These included feeling adequately prepared to address the issue; feeling more comfortable knowing they did not have to differentiate between victims and perpetrators (although still being concerned about potential aggression from perpetrators); having knowledge about referral agencies; and feeling more confident to raise the issue with male patients. This pilot study highlighted other areas where concerns still remained for practitioners, in particular concerns about how to manage abuse, particularly with respect to documentation, across patients in the same families.
There were a number of limitations to this pilot study that are important to recognise. As the study was a pilot, it included only a small number of general practices. The analysis of the PIM data tested across a wide range of outcomes that made it more likely for us to find statistically significant changes, although it should be noted that we found improvements in all areas targeted by the intervention. In relation to referral, our data were limited to the specialist agency named in the intervention, which meant that apart from the local services with whom we had contact, we were unable to measure if male patients had sought support elsewhere. Despite these limitations, however, the findings of the study support the hypothesis that participating clinicians benefited from the training.
Further research is needed to identify appropriate ways to document DVA for different members of the same family, and to ensure that attitudes towards sexuality and age do not prevent practitioners from engaging with all male patients who present with DVA related issues. We also need to better understand men’s help-seeking behaviour in order to improve uptake of appropriate support services.
Finally, the key message from practitioners who participated in this pilot study was the need to consolidate DVA training about responding to male patients into existing DVA training packages. This was primarily owing to time constraints related to competing training needs. This raises some concerns about whether there is capacity, in terms of time, within existing training programmes to include the material needed to address the needs of male patients who may be experiencing and/or perpetrating DVA, was and is a barrier that researchers need to address in future interventions.
Acknowledgements
This paper presents independent research commissioned by the National Institute for Health Research (NIHR) under its Programme Grants for Applied Research scheme (RP-PG-0108-10084). The views expressed in this publication are those of the author(s) and not necessarily those of the NHS, the NIHR, or the Department of Health. The study was approved by the South West Research Ethics Committee (reference 10/H0106/22).
Supplementary material
To view supplementary material for this article, please visit http://dx.doi.org/10.1017/S1463423614000358
