A comparative study of the prevalence of domestic violence in women requesting a termination of pregnancy and those attending the antenatal clinic


  • TT Wokoma,

    Corresponding author
    1. Sexual and Reproductive Healthcare, Conifer at Wilberforce Centre, Hull, UK
    2. Department of Obstetrics and Gynaecology, Hull Royal Infirmary, Hull and East Yorkshire NHS Trust, Hull, UK
    • Correspondence: Dr TT Wokoma, Sexual and Reproductive Healthcare, Conifer at Wilberforce Centre, 6–10 Story Street, Hull, HU1 3SA, UK. Email tonye.wokoma@nhs.net

    Search for more papers by this author
  • M Jampala,

    1. Lupset Surgery, Wakefield, West Yorkshire, UK
    Search for more papers by this author
  • H Bexhell,

    1. Department of Obstetrics and Gynaecology, Hull Royal Infirmary, Hull and East Yorkshire NHS Trust, Hull, UK
    Search for more papers by this author
  • K Guthrie,

    1. Sexual and Reproductive Healthcare, Conifer at Wilberforce Centre, Hull, UK
    2. Department of Obstetrics and Gynaecology, Hull Royal Infirmary, Hull and East Yorkshire NHS Trust, Hull, UK
    Search for more papers by this author
  • S Lindow

    1. Department of Obstetrics and Gynaecology, Hull Royal Infirmary, Hull and East Yorkshire NHS Trust, Hull, UK
    Search for more papers by this author



To compare the prevalence of domestic violence (DV) in women requesting antenatal care (ANC) and termination of pregnancy (TOP) in North East England.


This was a cross-sectional comparative prevalence study using self-administered questionnaires, with women selected as opportunistic samples over a concurrent period.


The participants were screened anonymously and confidentially in the ANC and TOP clinics.


Pregnant women in the first trimester requesting a TOP or ANC.


The participants were screened for a history of DV using a modified version of the Abuse Assessment Screening tool.

Main outcome measures

Prevalence of DV between ANC and TOP populations, and any differences in the characteristics of the women, such as age, level of education, or marital status. We aimed to determine the reasons for requesting a TOP.


There were 507 respondents, with 233 attending ANC and 274 requesting a TOP. Of the ANC population, 219 completed the questionnaire. In the TOP population, all the questionnaires were fully or partially completed. Women requesting a TOP were six times as likely to suffer physical abuse in the current relationship (5.8 versus 0.9%; χ2 = 10.2 (2); P < 0.05), and were five times as likely to suffer emotional abuse (9.9 versus 1.8%; χ2 = 13.6 (2); P < 0.0001), than those attending ANC. Of the 274 women requesting a TOP, only ten mentioned DV as a contributing factor.


There is a higher prevalence of DV in the TOP population than in the ANC population, but very few women stated that DV influenced their request for a TOP.


Domestic violence (DV) and abuse is defined as ‘Any incident or pattern of incidents of controlling, coercive or threatening behaviour, violence or abuse (psychological, physical, sexual, financial, emotional) between those aged 16 or over who are or have been intimate partners or family members regardless of gender or sexuality’.[1] There was an estimated 1.2 million female victims of DV,[2] and the direct and indirect cost of DV alone in the UK is estimated to be £23 billion.[3]

Almost half of all adult women in England and Wales have experienced violence in one form.[3] It has been estimated that the lifetime prevalence of DV against women in the general UK population ranges from 13 to 31%, and that it is widespread amongst pregnant women.[4] Domestic violence is a matter of importance because of the many obstetric complications and psychological effects that may arise as a consequence.[5] It is a major risk factor for psychiatric disorders, chronic physical conditions and substance abuse.[6]

A study of pregnant women in antenatal care (ANC) in East London showed 15% were found to be experiencing DV, and that for 40% of these women DV began during pregnancy.[7] Women in ANC with a history of DV, and who have experienced DV in the last year, are likely to be current victims of DV.[5] In a study of 475 pregnant women in North England, the prevalence of DV was quoted as 17%, with the highest rates/frequency in women aged 26–30 years. In that study, 10% of the study population was coerced into sexual activity by their partners.[8] Another cross-sectional study showed abuse in pregnancy to be significantly associated with an increased risk of perinatal death and, among live births, of preterm low birthweight and term low birthweight. A greater abuse frequency was associated with increased risk.[9]

Few studies have specifically focused on comparing the prevalence in women in ANC with those requesting a termination of pregnancy (TOP), although there are several studies that have investigated the prevalence of DV in pregnant women.[6, 10-12] This present study, therefore, compares the prevalence of DV in women requesting antenatal or abortion care in the Hull and East Riding area to assess differences in their characteristics, such as age, level of education, marital status, and ethnicity, and to find out their reasons for the request for a TOP.


The study was conducted as a self-administered anonymous questionnaire survey of unaccompanied pregnant women attending the antenatal clinic or the TOP pre-assessment clinic in the first trimester of pregnancy at the Hull and East Riding NHS Trust in North East England. A convenience sample of all pregnant women attending the ANC and TOP clinics between January 2011 and November 2012 was selected. Exclusion criteria included young women below the age of 16 years, who are deemed to be minors, as the recognised definition of DV excludes them. Women not fluent in the English language were also excluded. We considered that the presence of an interpreter would be necessary for the clarification of details, and this was judged to be inappropriate because confidentiality and anonymity would be compromised.

For the purpose of the study, we enlisted the services of the hospital midwives, nurses, and healthcare assistants to recruit the participants and to hand out questionnaires. A brief discussion of the objectives of the study was undertaken with each volunteer healthcare staff member. Senior nursing staff members were informed and consent was obtained for the study to be conducted during routine duties and working hours. The participants were screened using a modified version of the Abuse Assessment Screening tool to suit the purpose of the research.[13] Screening took place in a regular clinic room that was identified for the purpose of the study. A pre-study assessment of the room for privacy and confidentiality was undertaken by a member of the research team. A secure box with an aperture in the top cover was also provided for forms to be confidentially placed after completion. The women were invited unaccompanied into the room to measure their weight, height, and to record their blood pressure. These are measurements that are routinely checked during regular clinic visits as part of a general medical assessment. Upon entering the room a brief explanation of the study objectives was provided by the clinician responsible for vital signs assessment, and thereafter the patient was invited to participate in the study.

Each patient was then offered a laminated Patient Information Leaflet that expanded on and reiterated the aims and objectives of the study. The initial section of the questionnaire reaffirms the invitation to the study and assures the patient regarding anonymity and confidentiality. Women who attended the clinic with friends, spouses, or relatives were invited into a private room for blood pressure measurement and were encouraged to attend alone. The invitation for lone consultation was offered only once. Repeated attempts to obtain a lone consultation were avoided, as we believe that this might imply coercion. Women who declined the lone invitation were not invited to participate. Accompanied women were not included in the study as the presence of a third party may limit full disclosure, invalidate anonymity, and may have untoward consequences, such as feelings of embarrassment, shame, or humiliation. Also, if the accompanying individual was the DV perpetrator, the knowledge of disclosure may also lead to a potential escalation and retaliation against the victim.

The women were allowed as much time as required to complete the questionnaire. They were then requested to place the completed (or uncompleted) form in a sealed box, which was emptied at the end of each day. The anonymity of data was achieved by the lack of personal identifiable characteristics on the questionnaire, and women were not under pressure to participate in the study. In addition, the women were assured full confidentiality, with no one being made aware of their decision to participate in the study or even the invitation to participate. The patient information leaflet explained that the information obtained from the study would help to improve services offered to women. Signed consent was not requested, as this was thought to discourage potential participants from taking part in the study; therefore, only verbal consent was sought.

The investigators were aware of the possibility that the questionnaires, about a sensitive or embarrassing issue, could provoke emotional distress in study participants, especially at a difficult time. However, previous studies in our unit have shown that our indigenous population did not appear to have demonstrated these tendencies, and did not report any distress in filling self-administered, confidential questionnaires on the topic of DV. Helpline numbers were printed at the bottom of each questionnaire, and advice and support were made available to the participants as required.

There were separate questionnaires for each clinic group, and they comprised closed questions to ensure that the participants did not find it overly tasking to complete. The questions on both questionnaires were the same, but the TOP questionnaire contained additional questions on whether the partner was aware of the TOP request, any previous TOP, and reason(s) for the TOP request.

For both groups, the first section was for the singular purpose of collecting demographic data, which included information about their educational background, number of children, and relationship status. Another section obtained data about whether women had or were currently experiencing physical or emotional DV in a relationship. The next section of the questionnaire sought to identify the perpetuators in the event of an affirmative answer for the question on whether or not the women experienced DV. In addition, the frequency of occurrence of DV was sought.

The second section of the questionnaire for both groups was meant for those who responded in the affirmative, and comprised questions on the perpetrators as well as the nature and extent of the abuse. A series of acts of abuse were listed, encompassing a range of actions of differing severity. These included physical and non-physical acts of violence, such as slapping, pushing and shoving, being beaten up, severe bruising, grievous bodily harm, burns, broken bones, head injuries, internal organ damage, permanent injuries and disabilities, and use of weapons, as well as verbal, emotional, and psychological violence. Further questions requested information on whether sexual coercion had been experienced. The questionnaire concluded by assessing whether the participant was living in fear of the persons responsible for the DV, and if previous attempts to address the problem had been made.

The study instrument described has been used successfully in previous studies of DV prevalence in our unit.[8] The study received ethics approval from the National Research Ethics Service (NRES) for Yorkshire and the Humber—Sheffield and from the local research ethics committee.

The anonymised data were entered into an excel spreadsheet and spss 19.0 was used for the analysis (IBM, Chicago, IL, USA). The descriptive statistics were calculated using percentages. A series of statistical tests were conducted to analyse the data. To explore significant associations with the control variables (age, number of children, etc.), a one-factor analysis of variance (anova) was conducted; to observe the relationship between DV during pregnancy and the rest of the variables (nominal), nonparametric correlation tests were used. Percentages were calculated from the total number of questionnaires; however, small numbers of women did not answer each question.


In total, 507 women agreed to participate in the study. Of these, 233 were recruited in ANC and 274 were recruited from the TOP clinic. In total, 219/233 (94%) of the ANC participants completed part or all of the form. Fourteen other questionnaires were also found in the collection box with no data recorded. The results are therefore based on the 219 respondents. The entire population of the 274 women recruited in the TOP clinic either completed the questionnaire forms in full or in part. Among the TOP clinic participants, 5.8% were victims of physical abuse in the current relationship, whereas 0.9% reported physical abuse in the ANC population, regardless of pregnancy status. The corresponding rates of physical abuse in pregnancy were 4.7 and 0.9%, respectively. Women in the TOP group also suffered a higher rate of emotional abuse than those in the ANC population (9.9 versus 1.8%, respectively; Table 1).

Table 1. Characteristics of women attending the antenatal clinic (n = 219) and those requesting a termination of pregnancy (n = 274)
 ANC n (%)TOP n (%) P
  1. Percentages refer to the whole cohort; however, small numbers of women did not answer each question, and some women answered secondary questions inappropriately. All statistical analysis were performed using chi-square tests.

Marital status
Single54 (24.7)163 (59.5)P < 0.0001
Married/cohabiting154 (70.3)79 (28.8)
Divorced/separated9 (4.1)24 (8.8)
White202 (92.2)256 (93.4)P < 0.201
Black2 (0.9)4 (1.5)
Asian3 (1.4)6 (2.2)
Mixed8 (3.7)8 (2.9)
GCSE77 (35.2)125 (45.6)P < 0.0001
A level31 (14.2)45 (16.4)
Degree49 (22.4)30 (10.9)
Other higher education45 (20.5)34 (12.4)
In a relationship
Yes195 (89.0)181 (66.1)P < 0.0001
No21 (9.6)92 (33.6)
Feelings about relationship
Happy178 (81.3)131 (70.4)P < 0.0001
Satisfactory13 (5.9)31 (16.7)
Unsatisfactory3 (1.4)9 (4.8)
Unstable2 (0.9)11 (5.9)
Emotional abuse
Yes4 (1.8)27 (9.9)P < 0.0001
No198 (90.4)224 (81.8)
Physical abuse
Yes2 (0.9)16 (5.8)P < 0.05
No199 (90.9)227 (82.8)
Physical abuse in pregnancy
Yes2 (0.9)13 (4.7)P < 0.033
No204 (93.2)240 (87.6)

The survey showed that 93.4% of women in the TOP study population were classed as white British, which is similar to the prevalence in the ANC study population, in which 92.2% were classed as white British. This is broadly reflective of the ethnic demographics of the background population.[14] Most of the women (95%) had three children or less, and 36% had no children at the time that the survey was conducted. Women recruited from the ANC population were more likely to be married or in a relationship compared with women from the TOP population (Table 1). In women requesting TOP, 85/274 (31%) had undergone a previous TOP. Of these women, 25% (21/85) of them had had more than one repeat termination; however, only 7% (6/85) of these women admitted to experiencing DV in the current pregnancy or relationship. A higher proportion of the total TOP population described their relationship as being unsatisfactory or unstable than the total ANC population [20/274 (7.3%) versus 5/219 (2.3%)]. Women requesting a TOP were less likely to have a degree in comparison with the ANC population, where 22% of those had obtained a degree, although the women requesting a TOP were also significantly younger, with a mean age of 24.4 years, than those in ANC who had a mean age of 28.8 years.

In the TOP group five women who suffered DV had been forced to engage in sexual activity, and the main perpetrators were either a boyfriend or an ex-boyfriend. Less likely perpetrators included ex-husbands, as reported by one study respondent. Most participants who admitted to DV were also forthright in identifying the perpetrator, with only one participant declining to respond. Most of the women had spoken to someone else about the problem, but this person was most likely to be a friend rather than their general practitioner (GP) or other authority figure. Two women had been physically hurt on four occasions, and one woman said she had been hurt on more than ten occasions. In the TOP group, the majority of the women who had been physically abused had been slapped or pushed, punched, kicked, beaten up, or had severe bruising and burns. Four women admitted to fear of a partner, or any of the listed categories of people (husband, ex-husband, boyfriend, ex-boyfriend, mother, father).

In the ANC population, the women who disclosed DV had been abused by a boyfriend, ex-boyfriend, or ex-husband. The different types of physical abuse experienced involved slaps, pushing, punches, kicks, with bruises and cuts, being beaten up, severe bruising, and burns. There was no sexual violence perpetrated in this group. Two women who disclosed DV were afraid of a partner or someone else. When asked if help had been sought, these women either declined to answer the question or stated that they had not sought help.

The most common reasons that women gave for requesting a TOP were financial worries and failure of contraceptives (Table 2). Only 2% (10/274) of the women recorded DV as one of the reasons for seeking a TOP.

Table 2. Reasons for requesting a termination of pregnancy in 274 women
  1. Women were allowed to give more than one reason; therefore, 527 reasons were recorded, and 16 women did not give a reason.

Money worries11121%
Contraception failure10820.5%
May affect education/career7714.6%
Family complete7113.4%
Not in a relationship/unmarried6111.5%
Unstable relationship377%
Do not think partner is suitable father173%
Never wanted children163%
Unsatisfactory relationship112%
Violent relationship102%


Main findings

Our study found a six times higher rate of DV in the TOP population compared with the ANC population. Women in this study who were requesting a TOP were more likely to be younger, in unstable relationships, and less likely to have higher education than the women attending ANC. There was little ethnic variation between the two groups, although the included numbers were small. This study agrees with other studies where emotional abuse was found to be more prevalent than physical abuse.[7, 15] Domestic violence was not the main reason given for requesting a TOP. Financial constraints and contraceptive failure were cited as the main reasons women requested a TOP.

Strengths and limitations

We believe that this is the first comparative study of DV in two cohorts of pregnant women that gives direct evidence of the higher prevalence of DV in women requesting a TOP. The study was conducted in a National Health Service (NHS) hospital. The NHS model is an all-inclusive model of health care, and this would make the results transferable to the general UK population. A limitation of the study was the difficulty in separating women from their partner, or other accompanying person, particularly in ANC, to be able to offer them the questionnaire in private. This may have underestimated the prevalence of DV. In addition, questionnaires were handed out by healthcare professionals in the clinics, and although the study did not investigate the attitude of healthcare professionals to DV screening specifically, we found that some staff members appeared more willing to recruit participants to the study and to give out questionnaires than others. We believe that the bias introduced as a result of this process is minimal, as staff allocation to the clinics is arbitrary.


The results of this study confirm findings from previous studies.[8, 15] It highlights concern pertaining to the TOP population, where there is a higher incidence of DV than in the general population. A US study confirmed that women requesting TOP experienced an above-average incidence of DV.[16] Women who have experienced DV are consistently found to have poor sexual and reproductive health when compared with non-abused women.[17] In our study, we found financial constraints were the most common reason for requesting a TOP. Many poor women requesting a TOP have had many disruptive events in their lives, such as unemployment issues and housing instability, leading to the decision to terminate their pregnancies. A study by Jones et al.[18] showed that women were more likely to be victims of DV emanating from disruptive conditions.[18]

The prevalence of DV in this study (0.9 and 5.8% for physical abuse, and 1.8 and 9.9% for emotional abuse, in ANC and TOP clinics, respectively) was less than in previous studies conducted in the same area that found 17% of women reported DV[8]; however, this study requested information on the current relationship and current pregnancy only, and not over a broader lifetime perspective.

Previous studies have found that psychological abuse was more prevalent, especially during pregnancy.[19, 20] We also found emotional abuse was more prevalent than physical abuse. Many abused women define the psychological effects of DV as having a more profound effect on their lives, even with life-threatening or disabling physical violence.[21] A pregnant woman's mental, emotional, and physical health bears a direct correlation with the presence of DV,[6] and in some cases is directly linked with postnatal depression.[22] This effect of DV on a woman's mental status has implications for women requesting a TOP, especially those with mental health problems, where there is an increased risk of adverse psychological sequelae.[23]

There was little ethnic variation between the two groups, in agreement with the British Crime Survey (2004) that showed little variation in the experience of interpersonal violence by ethnicity.[24] We observed a reluctance of some healthcare professionals in our unit in screening for DV. Some studies show that healthcare professionals did not agree to women being screened in a healthcare setting,[25] and there is a gap in the evidence base for screening for DV.[26] Barriers to screening by healthcare professionals included lack of training or experience of screening, fear of offending or endangering patients, lack of effective interventions, patients not disclosing or not complying with screening, and limited time.[25]

A limitation highlighted in our study was that it was sometimes difficult to see women alone in order to recruit them to the study or screen for DV. Men have for a long time been encouraged to participate in health interventions for their partners, and their attendance is now culturally acceptable in most developed countries, especially during antenatal visits. However, failure in being able to separate a woman from the accompanying person at a clinic visit may be a pointer to a coercive relationship. Reproductive control is known to be a factor in abusive relationships, where the intimate partners exhibit controlling behavior and determine the use or non-use of contraceptives, and invariably the outcome of the pregnancy.[17]

It is well known that women may sometimes withhold information for fear of the consequences. A study of mental health services in London enumerated the barriers that service users faced in the disclosure of DV to professionals. This included fear of the consequences (including fear of Social Services involvement and consequent child protection proceedings, fear that disclosure would not be believed, and fear that disclosure would lead to further violence), the hidden nature of the violence, actions of the perpetrator, and feelings of shame.[21]

Domestic violence remains a complex public health issue that may start or escalate with pregnancy and ultimately lead to the loss of fetal or maternal life. The tragic impact of DV in maternity is well documented in the confidential enquiries report ‘Saving Mothers' Lives’.[27] This is a small fraction of the true scale of the problem, and health professionals looking after pregnant women during the antenatal period or women requesting a TOP are well placed to help and support women.


The newly revised Royal College of Obstetricians and Gynaecologists guideline on women seeking induced abortion has now included the recommendation that services should identify the issues that make women particularly vulnerable, for example DV, and refer them to appropriate support services in a timely manner.[23] This should be implemented and robustly audited. For practitioners who provide services, there should be training and support for all relevant staff, standard definitions relating to DV service provision allowing all agencies to work towards a common purpose, good practice guidelines and standards, against which services can be assessed and compared.[28] Structured pathways offer women the opportunity to disclose to health professionals who sensitively and tactfully request the information, and for those who accept help to be referred to the appropriate services.

Our study has highlighted that the rate of DV was six times higher in the TOP population than in the population of women attending ANC. Even though DV was not given as a frequent reason for requesting a TOP, the women in the TOP population are at a higher risk of DV, and this may be related to other life issues. We recommend more research to determine effective and acceptable screening methods and interventions to help women, particularly in the TOP population. It is only when this has been completed that sufficient measures will be initiated to curb this growing menace that poses a significant health risk.

Disclosure of interests

No competing interests to disclose.

Contribution to authorship

TTW was involved in the conception and planning of the study, acquiring and analysing the data, and in drafting and giving final approval of the article. MJ was significantly involved in the acquisition of data, analysing data, and drafting the article, and gave final approval for the article. HB was involved in the conception and design of the study, in the acquisition of data, and in drafting the article, and gave final approval for the article. KG was involved in the acquisition of data, in revising the article critically for important intellectual content, and gave final approval for the article. SL was involved in the conception and design of the study, in the analysis and interpretation of the data, and in drafting the article, and give final approval for the article.

Details of ethics approval

The study received ethics approval from NRES Committee Yorkshire and the Humber—Sheffield (ref. no. 10/H1308/42; 12 August 2011). The sponsor was Hull and East Yorkshire NHS Trust.


No funding was received for the study.


We thank Gill Barnett for assistance with inputing data into the spreadsheet and Nkechi Emenike for editorial assistance.