Abstract
Background Domestic violence is frequently only disclosed when healthcare staff directly inquire. Healthcare staff worry that inquiry may offend.
Aim To identify the characteristics of women who find inquiry about domestic violence by healthcare staff unacceptable.
Design of study Anonymous interview based cross-sectional study.
Setting Three general practice surgeries, one antenatal clinic and one emergency department in Cambridge, England, with a total of 2306 women attending for health care.
Method Cross-sectional survey.
Results In total 1452 completed questionnaires were returned; response rate 63%. One hundred and twenty–two women (8.4%) indicated that they found inquiry by healthcare staff unacceptable. Women at the emergency department and GP surgeries were more likely to find inquiry unacceptable (odds ratio [OR] = 3.3, 95% confidence interval [CI] = 1.1 to 9.9) and (OR = 3.9, 95% CI = 1.3 to 11.5) respectively, than in the antenatal clinic. Women at the antenatal clinic reported lower rates of abuse within 1 year than at the emergency department or antenatal clinic. Abuse within 1 year was strongly associated with finding inquiry unacceptable (OR = 4.5, 95% CI = 1.8 to 11.4), but not lifetime abuse (OR = 0.9, 95% CI = 0.5 to 1.9).
Conclusions Inquiry about domestic violence by healthcare staff is acceptable to most women. Acceptability is highest in women who have not been abused in the last year and who are attending the antenatal clinic. Women who attend the antenatal clinic have lower rates of abuse within 1 year.
INTRODUCTION
Domestic violence towards women is frequently hidden and prolonged. Routine inquiry has been controversially advocated to try to detect cases earlier and offer referral and support.1–4 Recent Department of Health initiatives have advocated routine inquiry for all pregnant women.5 Clinicians commonly worry that they may cause offence by asking about domestic violence.6 We wanted to identify what factors were associated with finding inquiry for domestic violence unacceptable in three different healthcare environments. Wide variations in the proportions of women who find routine inquiry unacceptable have been reported.7,8 Previous abuse has been inconsistently associated with finding inquiry unacceptable;9 this is probably due to differing methodology. We also wanted to identify whether the sex of the healthcare professional influenced women finding inquiry unacceptable.
METHOD
We conducted an anonymous, confidential interview based cross-sectional survey in the waiting rooms of the antenatal clinic and the emergency department of Addenbrooke's Hospital, and three general practice surgeries in Cambridge. Women who participated in the study were interviewed alone in a private room. Abuse was inquired about with a single item scale, the Brief Conflict Tactics Scale, which has high sensitivity and moderate specificity.10 The question used was ‘Have you been hit, kicked, punched or otherwise hurt by a partner or ex-partner in the last year?’. Previous experience indicates that this is a better measure of current abuse than asking about abuse by a current partner, since many women continue to suffer abuse after a relationship has ended.
Women were then asked if they would mind being asked about violence, first by a doctor, or by a nurse and/or by a midwife. Education attainment and socioeconomic status were measured using National Statistics methodology.
Questions used to identify whether a subject found inquiry unacceptable
Women were asked to indicate whether they agreed or disagreed, strongly agreed or strongly disagreed or did not know with the following statements:
‘I would not mind if a doctor asked me about violence at home’;
‘I would not mind if a nurse asked me about violence at home’; and
‘I would not mind if a midwife asked me about violence at home’.
The women were then asked a series of questions about their opinions about inquiry about domestic violence, based on variables identified from the qualitative literature.11–13
A male and female interviewer conducted the interviews, with a randomly selected 250 completed by self-report. The allocation to either interview by male and female was at the times that were convenient to the researchers, to fit in with work commitments. The female researcher was unable to conduct interviews in the GP surgeries because of delays obtaining honourary contracts. Women were excluded if they were unable to communicate clearly in English, too ill, accompanied by another adult who refused to leave, personal contacts of the research team or accompanied by a child over 3 years of age. The sample size of 1536 was based on 5% 95% confidence interval width around five levels of socioeconomic status, anticipating that 20% of women would object to routine inquiry. Logistic regression modelling was used to identify predictors of finding inquiry unacceptable.
Variables that were not associated with finding inquiry unacceptable at bivariate level were discarded from the model. Stepwise regression techniques are inappropriate when there are multiple dichotomous variables. Women gave informed verbal consent to participate. Written consent would have made the anonymity of the study meaningless. Women who disclosed active abuse were offered information about local support organisations.
RESULTS
In total, 2306 women attended for healthcare in the sampling frames. Of these, 1744 women were approached and invited to participate. The number of completed questionnaires obtained was 1452, roughly equally divided between the three healthcare environments. The overall response rate was 63%. The findings and conclusions from parts of the study with low response rates (general practice, 50%) were similar to those with high response rates, (emergency department, 78%). The sample was regionally and nationally representative, with a bias towards higher socioeconomic status (Table 1).
How this fits in
Many women will not disclose domestic violence unless they are asked directly and many clinicians feel uncomfortable asking direct questions about domestic violence. Women receiving antenatal care find inquiry about domestic violence more acceptable than women attending an emergency department or primary care. Women who find inquiry unacceptable are more likely to have suffered abuse within 1 year. The rates of abuse within 1 year are lower in women attending antenatal care than other healthcare settings.
There was no significant difference in the rates of abuse or other findings disclosed to these three interview strategies. The women's experience of abuse was lower in the group of women attending the antenatal clinic than the two other healthcare environments (Table 2).
One hundred and twenty-two (8.4%) women indicated that they would mind being asked about domestic violence by a healthcare professional. Logistic regression modelling of potential predictors of finding inquiry unacceptable is reported in Table 3. Women attending the emergency department or GP surgeries and with reported violence in the previous year were more likely to find inquiry unacceptable.
Opinions as to the reasons for finding routine inquiry unacceptable were varied. There were many reported reasons why women found inquiry unacceptable, these have been ranked in order of how they discriminated between those that found inquiry acceptable and those that did not (Table 4 and Supplementary Table 1).
Annual prevalence rates of abuse did not differ very much between the three identification strategies; there were lower rates among women attending the antenatal clinics.
DISCUSSION
Summary of main findings
Routine inquiry about domestic violence is acceptable to most women. Acceptability is higher in women attending the antenatal clinic, where we found a lower rate of abuse within 1 year. Women who found inquiry unacceptable were much more likely to have suffered abuse within the previous year. Women who had suffered abuse at any time in their lives were no more likely to find inquiry unacceptable than the non-abused population. There was a practically unimportant trend to increasing unacceptability with declining educational status. Socioeconomic status, chief complaint, age and pregnancy were not associated with finding routine inquiry unacceptable at all.
We also found that there was no difference in rates of abuse disclosed to a male or female researcher or by self-completion. This implies, at least in a research setting, that when women are asked about violence, they tell the truth. There were many reasons why women found inquiry unacceptable. This is disappointing because a few reasons could be more easily addressed by training. Women who found inquiry unacceptable were more likely than other women to indicate that they preferred inquiry by a woman.
Strengths and limitations of the study
There are some limitations to this work. The response rate in the primary care arm was low, but findings from this arm were similar to those from the arms with high response rates. Women may give socially desirable answers in a research interview and introduce bias. However, the findings from the anonymous self-completion questionnaires were similar to that from direct interviews. Theoretical answers may differ from actual practice, there may be women in this survey who indicated that they found inquiry unacceptable who may find that they do not actually mind when asked. Conversely, there may be women who actually find inquiry unacceptable despite having indicated that they would find it acceptable. While patients are, in general, keen for doctors to ask about lifestyle matters, in the setting of a real, time-pressured consultation, they only want such questions when appropriate and when it does not take time away from their own concerns. Indeed, many women in this survey indicated that lack of time was a significant concern. However, many women in abusive relationships will not disclose violence unless asked directly and is top of their ‘hidden agenda’. Women who have recently been abused are much more likely to object and there are concerns that they may quickly learn to lie. This might compound the feelings of hopelessness that can accompany an abusive relationship. However, most abused women would not disclose unless asked directly and may find it a source of huge relief to be able to disclose in a confidential setting.
The measure of abuse that we used excludes sexual and emotional abuse. These may be areas that may be more upsetting to ask about, both for healthcare staff and patients and we may have found more women indicating concern about these important areas. Much perinatal care is delivered by community midwives and health visitors and this important patient population is not completely represented by this study. However, most pregnant women attend the antenatal clinic once. The benefit of interventions for domestic violence is still unproven, although this may represent absence of evidence rather than absence of effect.
Comparison with existing literature
Previous work has demonstrated that variable majorities of women find inquiry acceptable.7,9 This is the first study to consider acceptability across three distinct healthcare environments.
Our finding that women attending the antenatal clinic have lower rates of abuse is unexpected. Previous work shows that pregnancy is associated with a doubling of risk of violence, although this study did not completely adjust the confounding effect of socioeconomic status.14 Controlling for age indicates that the prevalence of violence is similar in pregnant and non-pregnant women.15 Women who have high-risk pregnancies tend to be older, more affluent and are more likely to be referred for hospital-based antenatal care. Increasing age and higher socioeconomic status are associated with lower rates of violence. The lower rate of abuse within 1 year persisted after stratification by socioeconomic status.
Implications for future research
Asking women about domestic violence is acceptable to the majority of women. The antenatal clinic may be the most acceptable place to do this, but the rates of abuse in this population are lower. Women who object are much more likely to have suffered recent abuse. The treating physician or nurse should consider offering referral to support agencies. The case for offering routine inquiry for domestic violence is not yet proven; further work needs to demonstrate effectiveness of interventions that are provided for women who are identified.
Supplementary Material
Acknowledgments
Dr Chris Palmer provided a sample size calculation.
Notes
Supplementary information
Additional information accompanies this article at http://www.rcgp.org.uk/Default.aspx?page=2482
Funding body
Adrian Boyle was supported by an NHS R&D (Eastern Region) Health Services Research Fellowship. Mary Cooke conducted some of the interviews and was supported by the Addenbrookes Emergency Department Research Fund
Ethics committee
Cambridge Local Ethics and Research Committee (LREC 2003/276)
Competing interests
The authors have stated that there are none
- Received April 4, 2005.
- Revision received November 1, 2005.
- Accepted January 31, 2006.
- © British Journal of General Practice, 2006.