Domestic violence in a genitourinary medicine setting – an anonymous prevalence study in women

INTRODUCTION

Domestic violence (DV), defined as ‘any incident of threatening behaviour, violence or abuse (psychological, physical, sexual, financial or emotional) between adults who are intimate partners or family members regardless of gender or sexuality’ ¹ affects around one in four women in the UK. ² For women aged 19–44 years, DV is the leading cause of morbidity, greater than cancer, war and motor vehicle accidents. ² DV is also known to be associated with sexual health problems. ^3–5 To the author's knowledge, there are no published UK prevalence studies of DV in genitourinary (GU) medicine clinical settings. Studies from sexual health clinics in Australia and the USA have reported rates of 8.8% in the preceding year ⁶ and 24–37.8% lifetime risk of physical violence, respectively. ^7,8

Our aim, therefore, was to determine the lifetime and preceding 12-month prevalence of DV in female patients attending an inner London GU medicine clinic, and to examine whether DV was associated with any demographic factors and health risk behaviours, in particular with sexually transmitted infections (STIs). There is existing evidence that women aged <25 years, ^2,9 who are cohabiting, single or separated, have children living at home with them, unemployed, ¹⁰ living in poverty ¹¹ and engaging in substance abuse ^7,8 are at increased risk of DV. However, we wanted to explore sociodemographic factors that were associated with DV in patients attending our particular clinical setting.

METHODS

The study was a cross-sectional survey of female attendees aged 16 and over, recruited from an inner-city GU medicine clinic between 1 July and 30 November 2006. The clinic is a sexual health clinic in a teaching hospital situated in a borough of London with high rates of STIs. ^12,13 The clinic operates as a daily walk-in service with a single queue for women, and doctors see patients sequentially as they arrive. The doctors work in varying weekly rotations and the recruiting doctor did two to three womens' clinics a week. Of the 3300 women who attended the walk-in clinic in the study period, 418 consecutive attendees seen by the recruiting doctor were given a self-administered anonymous questionnaire and information about the study. Interpreters were available on request for those who did not speak English. Women ticked a box in the questionnaire to indicate their consent. The questionnaire was completed in a private area and placed in a box to aid confidentiality. Those who requested help because of current or past DV were offered health adviser input in the clinic or a referral to MOZAIC, an independent DV advocacy service within the hospital. A contact card for MOZAIC and other local DV organizations was attached to the questionnaire for participants to take away if they needed help for DV.

DV was assessed using a variation of the Abuse Assessment Screen ¹⁴ (AAS) which included the following questions: ‘Have you ever been hit, slapped, kicked or otherwise physically hurt by a partner?’; ‘Have you ever been threatened with violence by a partner?’; ‘Have you ever been forced into sexual activities by a partner?’; and, ‘Have you ever felt afraid of your partner, even if he/she has never hit you?’. Any positive answer was followed by specific questions on whether the abuse was perpetrated by a current or former partner and whether the violence took place in the last 12 months.

All respondents who reported a history of STI, defined as a self-reported history of chlamydia, gonorrhoea, pelvic inflammatory disease, trichomoniasis, genital herpes, genital warts or syphilis, were also asked if they suffered any form of DV as a result of the STI: ‘Has your partner ever humiliated you or said cruel things to you about an STI that you have had?’; ‘Has a partner ever verbally threatened you with violence during or immediately after a discussion about an STI you have had?’; ‘Has a partner ever hit, slapped, kicked or otherwise physically hurt you during or immediately after a discussion about an STI you have had?’; and, ‘Have you ever not told a partner about an STI you have had because you were afraid of his/her response?’ Respondents were asked whether the abuse occurred in the last 12 months and if the perpetrator was a current partner.

Ethics approval for the study was obtained from Guy's and St Thomas' Hospital Research Ethics Committee.

ANALYSIS

Based on the literature, ^7,8 we estimated the lifetime prevalence of DV to be 23–50%. The level of accuracy for the study had been chosen to be 1.96 standard errors, so that a 95% confidence interval (CI) could then be calculated. We calculated the maximum sample size to be 385 for a prevalence estimate with an accuracy of ±0.05 and a confidence level of 95%.

Questionnaires were processed and analysed using SPSS statistical software. ¹⁵ Frequencies and descriptive statistics with CIs were used to calculate the lifetime and last 12-month prevalence of DV and the association between DV and history of STIs. Logistic regression analysis was undertaken to model social, health and behavioural predictors of DV in the last 12 months. The dependent variable for the multivariate analysis was whether or not DV in any form was reported in the last 12 months. The strength of the associations is reported as odds ratios (OR) with 95% CI.

RESULTS

Of the 3300 women attending the clinic during the study period, 418 were given the questionnaire. Of these, 393 were returned and 388 were completed giving a response rate of 92.6%. Sample demographics are presented in Table 1.

DISCUSSION

This first published UK study of the prevalence of DV in a GU medicine setting demonstrates a high prevalence of DV among female attendees. We found a higher lifetime prevalence of DV compared with studies conducted in other UK health settings, for example 41% in a primary care survey, ¹⁶ 34.9% in a family planning clinic, ⁹ 21% in a gynaecology clinic ¹⁷ and 23.5% in a maternity care setting. ¹⁸ However, the rate is comparable with studies carried out in inner-city sexual health settings in USA ⁷ and Australia. ⁶ We feel that further GU medicine clinic prevalence studies are warranted.

Our finding of 17.4% prevalence rate of DV in the preceding 12 months may reflect the relative youth of our sample. Data from the British Crime Survey indicate that women aged 25 and under are the most likely to suffer DV. ² In other settings where the AAS was tested in women of reproductive age, high rates of DV were recorded. Studies have reported prevalence rates of DV to be 14% in the last year in a family planning clinic, ⁹ 48% lifetime in college attendees in the USA ¹⁹ and 23.5% lifetime in a maternity care setting. ¹⁸

The high prevalence rate found makes DV an important issue for sexual health practitioners. Sexual abuse, sexual humiliation and rape are features of partner violence that have implications for women's health and choices regarding their sexual practices and behaviours. ^17,20,21 The 16.8% of women reporting sexual abuse in the context of DV in our study is comparable with the 16% reported in the 2005–2006 British Crime Survey ² but lower than a survey of primary care attendees in East London where 24% reported sexual violence in their lifetime. ²² DV is known to be associated with sexual health problems including chronic pelvic pain, ³ gynaecological problems, ⁴ unwanted pregnancy, ⁵ STIs ^7,8 and with increased risk behaviours for HIV infection such as unprotected sex and having multiple partners. ²³ An unwanted pregnancy or STI may result from sexual abuse or from the abusive partner's control of contraception. Abused women may fear negotiating condom use or refusing sexual intercourse because this may result in escalating violence. ²⁴

For victims of DV diagnosed with an STI, fear of disclosure of an STI to a partner may lead to failure of partner notification and result in repeated episodes of an STI. Following a discussion of an STI with a partner, DV (commonly threats of violence and humiliation) was reported in 13.1% of women in our sample who had an STI. This is consistent with studies of HIV positive patients ^25,26 and has implications for abused women who are encouraged to notify their partner of an STI. As UK GU medicine clinics struggle to achieve the 48-hour access target, ²⁷ many have attempted to reduce follow-up appointments by sending results by post or text message. Victims of DV may be placed at risk of further violence if their results are inadvertently made available to an abusive partner. Women may need to be assisted with safety strategies, including locating a safe place to stay, before informing an abusive partner about an STI. A high number of respondents in our study stated that they feared their partner's response and chose not to disclose an STI, including within their current relationship. In the context of a controlling and abusive relationship, a woman may be unable to abstain from sexual intercourse, adhere to prescribed treatment and attend follow-up appointments, all of which contribute to poor sexual health outcomes and higher rates of STIs.

In this study, women with a history of STI were more likely to experience DV in their lifetime. This association has been reported by other studies from sexual health clinics ^7,8 where having a main partner who had sex outside of the relationship was a factor associated with DV. ⁸ Our finding that a history of illicit drug use was significantly associated with DV in the preceding 12 months has also been found in studies of sexual health clinic attendees in San Francisco and Australia. ^7,8

The study has a number of limitations. First, not all patients seen in the clinic in the study period were offered the questionnaire. Lack of clinic space limited the feasibility of this, as we felt it was important that women filled in the questionnaire in a private area. We sought to minimize the potential selection bias introduced if different doctors recruited on an ad hoc basis which may have resulted in higher recruitment rates of patients with fewer clinical issues and when clinics were less busy. As it is a general walk-in service, we assumed the sampling of random clinics to be representative of our clinic population. Secondly, due to the cross-sectional design, it is not possible to comment on causality or temporal sequence of events. DV cannot be established as the antecedent factor in the reported associations, such as the association between a history of DV and STI, or the associations between DV in the last 12 months and illicit drug use and multiple sexual partners. However, the data does indicate that some respondents were abused as a result of discussing an STI with a partner. Prospective studies are needed to determine the precise nature of these relationships. Thirdly, data on socioeconomic status such as education and income levels were not collected. Given the higher proportion of ethnic minorities in lower socioeconomic groups it is important to control for socioeconomic status when examining associations between ethnicity and experiences of DV. Fourthly, the Abuse Assessment Screen was developed for assessing DV in pregnant women rather than young single women who make up the majority of our sample. Measurement against the revised Conflict Tactic Scale for physical violence during pregnancy in Brazil reported the sensitivity to be between 32 and 61% for minor and severe physical violence and specificity of above 97%. ²⁸ A similar trend was observed for a study in Hong Kong. ²⁹ The sensitivity improved for more severe cases. As a result, the Abuse Assessment Screen may underestimate moderate forms of abuse. Finally, the findings from this clinic sample may not be representative of GU medicine patients in the population.

Our results suggest that GU medicine clinic staff need to be aware of DV and the implications for women's sexual health. In order to support patients affected by DV, sexual health practitioners need training on DV. This should include information about prevalence, associated health outcomes, potential risk markers, strategies for asking women about partner violence, documenting the abuse confidentially and providing information on specialist DV services. It may be appropriate to opportunistically enquire about DV where it is suspected, for example in women with repeated STIs or who have not complied with treatment.

The Department of Health have produced a domestic abuse handbook for health professionals in which it states that ‘all Trusts should be working towards routine enquiry and providing women with information on domestic abuse support services. It is important to take the initiative and be proactive’. ³⁰ However, there has been no research on the acceptability of routine enquiry in GU medicine settings and little is known about the needs of lesbian, gay, bisexual and transgender communities and heterosexual men affected by partner violence. Further research about patient preferences regarding discussions about DV, including sexual violence and rape, and interventions in the GU medicine setting is needed to inform the development of culturally competent interventions.