Aim, design and setting

The objective of this study was to address a gap in the available evidence about whether Australian health professional’s personal history of DV is associated with their clinical care of survivor patients. The research question at the outset of this project was: Is personal experience of DV associated with a health professional’s attitudes about DV survivor patients and the role of the health workplace; identification of survivor patients; comfort to discuss DV and clinical interventions with survivor patients? We hypothesised that, after adjusting for possible confounding background variables, compared with their non-abused peers, survivor health professionals would: (1) demonstrate more sensitive attitudes about survivors; (2) feel more comfortable discussing DV and sexual assault with their patients; (3) ask more patients about DV; (4) identify more survivors within a 6-month period and (5) provide more DV interventions to survivor patients, including DV referral. While not an initial focus of the study, the effect of training on clinical practice emerged as an interesting finding during the data analysis and was included in the results.

A full description of the study design, setting, participants and recruitment process has been reported previously in a paper about prevalence.1 In brief, we conducted an anonymous and voluntary cross-sectional survey of all health professionals in one Australian tertiary maternity hospital between 8 August and 31 December 2013. Participants were female health professionals (nurses, doctors and social workers) working with patients. An online survey link and encouragement to participate by the chief executive officer was distributed via email to all part-time/permanent clinical staff—nurse/midwives, doctors and allied health professionals. Staff were ineligible to participate if they were employed casually or did not work in a clinical capacity (ie, administration staff).

Data collection and measures

Exposure to DV encompassed 12 month and adult lifetime intimate partner violence and/or lifetime violence by a family member. Violence by an intimate partner was measured using the Composite Abuse Scale, a well-validated and widely used self-report measure of physically, sexually and emotionally abusive behaviours perpetrated by an intimate partner.16 This was measured by: scoring on the 12-month subscales, or two of the lifetime subscales, ‘Severe Combined Abuse’ or ‘Physical and Emotional Abuse’, or by scoring >7. Violence by a family member was measured by answering positively to either of two questions about lifetime physical, emotional and sexual abuse by a family member and witnessing parental abuse. Overall, 45.2% (212/471) of the female participants in this sample qualified for inclusion into the DV exposure group.1

The main predictor variable was exposure to DV. In a follow-up analysis, the predictor variables were DV training and demographics. The outcome variables were: attitudes (measured by Physician Readiness to Manage Intimate Partner Violence Survey,17 comfort discussing DV, DV inquiry and interventions after identifying a new DV case during a 6-month period (table 1). Adjustment for potential confounding variables was made a priori based on the literature, and included: age (40+ years),14 18 professional background (allied health),10 14 DV training (1+ days)9 10 15 and years of clinical experience (10+ years).14 15

Statistical analysis

Clinical interventions to identify and respond to DV were summarised using frequencies and percentages for categorical data and means and SD for ordinal data. Independent t-tests and X² tests of comparison were used to compare mean scores. Linear regression compared differences in mean scores across exposure for attitude scores, while logistic regression was used for comfort asking about DV and clinical intervention variables. ORs, 95% CIs and p values were used to assess the likely size of the association between each clinical action and DV.

Data were analysed with STATA V.13.1.20

Patient and public involvement

No patients or the public were involved in developing the research question or outcome measures. Health professionals were involved, however, and they were informed by their clinical work with survivor patients. Health professionals contributed to the research questions and overall design of the study. Results of the study will be disseminated to participants via workplace newsletter items and staff public speaking forums at the recruitment site.

Participant characteristics

The survey was sent to 1047 female health professional staff and 471 participated: 366 completed the survey electronically, while 105 returned a paper version, giving a response rate of 45.0%. Most participants were nurse/midwives, aged 30–60 years, had 10 or more years of experience, and were demographically representative of their non-participating peers (table 2). Survivor health professional participants (45.2%, 212/469) were significantly more likely to be aged 30–39 years and have an allied health background compared to participants who were not survivors.1

Training and preparedness

Survivor health professionals were more likely to have received one or more days of DV training (adj OR 1.9, 95% CI 1.2 to 3.2) and to report more sensitive attitudes about DV survivors (adj. coef. 0.2, 95% CI 0.1 to 0.4) compared with their colleagues who had not experienced DV. Survivor health professionals were no more likely than others to find it upsetting to talk about DV with their patients (adj OR 0.8, 95% CI 0.5 to 1.1) (table 3). Irrespective of whether a health professional had experienced DV, having undertaken at least 1 day of DV training was positively associated with good clinical care, including identifying survivor patients (adj OR 9.6, 95% CI 5.0 to 18.8), risk assessment (adj OR 4.6, 95% CI 2.2 to 9.5), safety planning (adj OR 4.3, 95% CI 2.1 to 8.9) and referral (adj OR 2.1, 95% CI 1.0 to 4.1). This finding occurred even after adjustment for possible confounders (table 4). Univariate analysis suggested a positive association between hours of DV training and asking patients about the issue. The analysis also suggested that allied health professional participants (ie, social workers) were more likely to have had 1+ days of DV training and to have safety planned and referred survivor patients than the other professional groups (table 4).

Identifying survivor patients

In the unadjusted analysis, being a survivor health professional was associated with asking patients about DV during the previous 6 months and motivation ‘to go the extra mile’ with them. However, in the adjusted analysis a between-group difference did not remain, although the significance level for asking patients about DV was approaching 0.05 (adj OR 1.5, 95% CI 1.0 to 2.3, p=0.07) (table 3).

Clinical care

Of the 193 participants who identified a survivor patient in the last 6 months, the unadjusted results indicated that survivor health professionals were more likely than others to have provided DV information to patients, conducted risk assessments, safety plans and made referrals to services (table 3). However, in the adjusted analysis, the only association that remained was accessing DV information for patients (adj OR 2.0, 95% CI 1.0 to 4.0).

Strengths and limitations

Strengths of this study include adjustment for potential confounders in regression,7 8 11 13 14 the inclusion of health professionals from all clinical backgrounds reflected in hospitals7 8 10–14 and the recruitment of primary DV survivors.9 11 13 14 Limitations of this study include self-report and social desirability, which may have led to under-reporting of abuse, and the single recruitment site that prevents generalisability of findings.21 22 It is possible that DV survivors were more motivated to participate in the project than the other people,21 and we acknowledge the possibility that non-respondents may have differed from respondents in a way that affected our conclusions. Considerable attempts were made to address selection bias by active recruitment and strong encouragement to participate; a 45.0% response rate was achieved. Despite the sample limitations, considering the work demands of our participants and the representational participation of nurses, doctors and allied health professionals, we argue that our response rate is acceptable and comparable to similar research.7 8

The study in the context of other studies

The finding of an association between a health professional’s history of DV and aspects of clinical care of survivor patients echoes other research.7 14 A possible interaction between DV training, personal experience and clinical care has been suggested previously.9 However, the finding in this study of a relationship between a health professional’s history of DV and their participation in training is critical and new. This finding was surprising; we did not posit a hypothesis about survivors accessing more hours of professional training. We suggest that survivor health professionals may be more likely to attend training because they understand the issue, resultant impact on health and the need for timely responses, and/or they are seeking information or validation about their own experience.

The association between being a survivor health professional, holding more sensitive attitudes about survivors and providing DV information to patients is consistent with one previous study.14 This small study examined nurses’ thoughts, feelings and proposed actions in response to identifying survivor patients, finding an association between being a survivor nurse and having more sensitive, empathetic responses to survivor patients.14 Our study extends these findings since that analysis did not adjust for potential confounders and the exposure group included health professional participants with secondary exposure to DV through friends/family. We postulate that survivor health professionals may hold more sensitive attitudes about survivors and fewer misconceptions about DV because of empathy stemming from a shared trauma experience. Additionally, they may be more likely to access DV information for their patients because they believe that DV awareness is an important intervention in itself.

Implications

Given the association between being a survivor health professional and attendance at DV training, this should be regarded when developing and delivering DV training for health professionals.7 Such training could incorporate reflection, safety information, emotional health psychoeducation, referral, workplace support and promoting a safe and supportive healthcare workplace.15 23 More broadly, these findings provide evidence that survivor health professionals are an asset to the organisations in which they work since among the participants in this study, they appear to be doing more of the work seen as better clinical care of survivor patients. This finding rebukes the misconception that women who have experienced DV are enduringly vulnerable, a distortion which can encourage women to remain silent, especially at work, for fear of how they might be regarded if they speak up.24 This study presents an opportunity for health services to explore how the lived experience of DV for both their patient and staff survivors could inform and improve their service. A past critique of health and other ‘mainstream’ DV response services has been that they have not meaningfully consulted survivors.25 Listening to the experiences and needs of survivor health professionals may enhance the support those health professionals feel from their employer, strengthening their personal and professional capacity as they care for patients. There is evidence that accessing support for DV can result in meaningful change in survivors’ lives, including in their employment.18 We argue the need for greater workplace supports aimed at promoting safety and recovery from violence and strengthening clinical practice with patients. This requires organisational leadership, evidence-based response guidelines and resourced individuals to whom a disclosure can be made and who can provide varied levels of support (resource information, clinical debriefing, longer term emotional support).1 Trauma-informed care may provide a useful framework to guide the response of hospitals towards better supporting staff and patient DV survivors.26 A trauma-informed system is one in which all components have been organised with the understanding that trauma is a centralising influence in survivor’s lives, and organisational, operational and clinical practice should prioritise safety, control and the recovery trajectory.27 More research is required to better understand the impact of DV workplace supports for DV on health professional women’s well-being and clinical care. This study sheds light on the survivor experience, especially for women at work.