Screening of women for intimate partner violence: a pilot intervention at an outpatient department in Tanzania.

UNLABELLED: Intimate partner violence (IPV) is a public health problem in Tanzania with limited health care interventions.
OBJECTIVES: To study the feasibility of using an abuse screening tool for women attending an outpatient department, and describe how health care workers perceived its benefits and challenges.
METHODS: Prior to screening, 39 health care workers attended training on gender-based violence and the suggested screening procedures. Seven health care workers were arranged to implement screening in 3 weeks, during March-April 2010. For screening evaluation, health care workers were observed for their interaction with clients. Thereafter, focus group discussions (FGDs) were conducted with 21 health care workers among those who had participated in the training and screening. Five health care workers wrote narratives. Women's responses to screening questions were analyzed with descriptive statistics, whereas qualitative content analysis guided analysis of qualitative data.
RESULTS: Of the 102 women screened, 78% had experienced emotional, physical, or sexual violence. Among them, 62% had experienced IPV, while 22% were subjected to violence by a relative, and 9.2% by a work mate. Two-thirds (64%) had been abused more than once; 14% several times. Almost one-quarter (23%) had experienced sexual violence. Six of the health care workers interacted well with clients but three had difficulties to follow counseling guidelines. FGDs and narratives generated three categories Just asking feels good implied a blessing of the tool; what next? indicated ethical dilemmas; and fear of becoming a 'women hospital' only indicated a concern that abused men would be neglected.
CONCLUSIONS: Screening for IPV is feasible. Overall, the health care workers perceived the tool to be advantageous. Training on gender-based violence and adjustment of the tool to suit local structures are important. Further studies are needed to explore the implications of including abuse against men and children in future screening.

Intimate partner violence (IPV) is one form of interpersonal violence defined as threats, attempts, or actual physical, sexual, or emotional abuse by a current or former partner (1). IPV is a known public health and human rights concern reported to be five times more commonly perpetrated by men to women than vice versa (2). The WHO multicountry study on women's health and domestic violence against women estimated a lifetime prevalence of physical or sexual violence ranging between 15 and 71% in the 15 sites studied (3). Despite the high figures, the hidden nature of IPV against women may still be an underestimation of its occurrence and impact and partly explain the variation in prevalence between countries and settings (4). In Tanzania, one of the countries included in the WHO study, the lifetime prevalence of physical or sexual violence was estimated to be 41% for the urban and 56% for the rural site (5).

Apart from physical injuries, IPV may cause other forms of serious health damages and decreased social well-being of the affected individuals as well as of their children and families (6). Mental health (7–9), gynecologic problems, adverse pregnancy outcomes (10), chronic pain, and changes in the endocrine and immune functions (6) are among the serious health consequences reported. Women may also suffer from reduced coping capacity that can cause alcohol and drug abuse, suicidal attempts, and homicides (11, 12). Sexual violence has been specifically linked to an increased risk of HIV and AIDS for exposed women (13–15). According to the WHO multicounty study, 30% of the Tanzanian women in the urban site, who had experienced physical violence, reported having been injured, and out of these 61% had needed health care for their injuries. Of the Tanzanian women, who had ever experienced physical or sexual violence, 11–12% reported having had suicidal thoughts (16).

Screening for IPV within the health care system is suggested to increase adequate care and support (17, 18). As women with IPV experiences may only not present with physical health symptoms, screening all women would ensure recognition of IPV as a health care concern (19). Shame and fear of retaliation surrounding IPV disclosure may also decrease with routine screening (6, 20). Screening is also seen as a means to increase early referral of women exposed to violence (21, 22). Asking about abuse is expected to promote communication with women clients and to reduce feelings of isolation and improve the self-esteem of those who have been exposed to violence. The attention given to women during screening sessions and the information given about available services are also believed to promote their help-seeking behavior (23–25). Arguments against screening have mainly focused on the lack of evidence that screening actually improves the health outcomes for the affected women and results in a reduction of IPV incidents (26, 27).

In 2006, we performed a qualitative study regarding Tanzanian health care workers’ (HCWs) experiences of meeting IPV clients. IPV was found to be a ‘hidden agenda’ because both HCWs and clients had difficulties to bring up the issue of violence. For the HCWs, this became a challenge even if they clearly indicated a desire to make a difference and identified the need for training and guidelines (28).

The overall aim of this study was to explore the feasibility of introducing a screening routine for IPV against women at a district hospital outpatient department (OPD) in urban Tanzania. The specific objectives were to assess how the screening tool captured abuse experienced by women, to observe how HCWs guided, counseled, and referred the affected women, and to discuss how the HCWs perceived the benefits and challenges to introduce routine screening for IPV within the Tanzanian health care system.

Methods

Study setting

We performed the study at the OPD at Temeke District Hospital (TDH), Temeke district, in the city of Dar es Salaam, Tanzania. The site was chosen because the study was built on previous experiences of studying violence against women and children in the same district (28–30). In the 2002 census, Temeke had 813,667 inhabitants living in an area of 656 km2. About 90% of the population lives in periurban wards that utilize TDH as the main health service provider. About 18% of the population lives below poverty line, with an under-five children mortality rate of 134 and an infant mortality rate of 84 per 1,000 live births (31). The population is of mixed tribal origin and represents the social, cultural, and economic diversity of the country.

At the time of the study, 285 HCWs were employed at TDH. The OPD had a manager, 2 medical officers, 5 assistant medical officers, 12 clinical officers, 6 nursing officers, 13 nurse midwives, and 19 medical attendants who attended about 1,000–1,500 outpatient visitors per day. The majority (62%) of clients who attend the OPD were women.

Study design

The study design included a training workshop, the introduction of a screening tool in a clinical setting as well as an evaluation of the process and outcome of the intervention. The evaluation consisted of a quantitative analysis of women's responses to the screening tool, observations of the interaction between the HCWs and their clients. Focus group discussions (FGDs) with HCWs and their short-written narratives included to capture HCWs perception about benefits and challenges of using the screening tool. There was no baseline information collected apart from the study of HCWs experiences of meeting women exposed to violence (28). Our benchmarking for change was the health care workers’ ability to use the screening tool to detect IPV among screened women. As we have mentioned earlier, the lifetime prevalence of sexual or physical violence in urban Tanzania is high, estimated at 41% (5).

Preparation for screening

The training workshop

Meetings with the OPD manager and four senior staff were held to discuss logistics and to arrange for the training workshop. With the assistance of the OPD manager, 40 HCWs, representing different professions and departments, were selected and invited for the workshop. Out of these, 39 accepted and participated in the training focused on gender-based violence, particularly IPV and the use of a screening tool. The medical power and control wheel was used to illustrate how the HCWs may influence the way IPV is discussed within the health care setting (32). Motivational interviewing techniques were introduced to raise their counseling skills and to discuss how these techniques may benefit the use of the abuse screening tool (33). One session included a discussion about the ethical issues relating to IPV care and support.

The sampling of HCWs

After the training workshop, the OPD manager helped to select four clinical officers, one medical officer, and two nursing officers (5 women and 2 men) from the OPD who had taken part in the training. They were purposively selected on the basis of being likely to meet all women entering the hospital and willing to implement the abuse-screening tool within their daily duties.

The screening tool

We modified the Abuse Assessment Screen tool developed by McFarlane et al. (32) that was used to assess violence on pregnant women. Our tool included five questions focusing on the experience of emotional, physical, and sexual violence (Table 1). The original tool also had questions that needed ranking of abuse episodes and questions related to pregnancy. These were excluded in our tool because the ranking questions demanded further training.

Table 1

Abuse screening tool to women attending the outpatient department,Temeke District Hospital, 29 March 2010 to 17 April 2010

	Questions
1.	Have you ever been emotionally or physically hurt by anyone in your lifetime (Yes/No)?
2.	Within the last year, have you been hit, slapped, kicked, or otherwise physically hurt by someone (Yes/No)?
	If yes, by whom? (Relationship, not name)………………………………………………
	Total number of times…………………………………………………………………….
3.	Within the last year has anyone forced you to have sexual activities (Yes/No)?
	If yes, by whom (Relationship, not name)………………………………………………..
	Total number of times……………………………………………………………………..
4.	Are you afraid of anyone of the people you mentioned above (Yes/No)?
5.	Please tell me any complementary information regarding the violence you have been subjected to. Is there something that you would like to tell me?..............................................................................................................................

Implementation of the screening

Between 29 March and 17 April 2010, seven HCWs at the OPD invited their first three patients (women>18 years) each day to be screened. With support from the hospital management, the normal routine work schedules were slightly modified to assure confidentiality during the study period. The HCW was assigned to attend a single client in one room at a time instead of the normal routine, where two clinicians attend two clients in the same room. The HCWs could choose either the English or the Kiswahili version of the screening form. The first author (RL) collected and checked the filled forms twice a week.

According to the plan, we expected 315 women to be screened. However, logistic challenges such as overcrowding of patients and severe health condition of some of the women made it difficult for the HCWs to meet the plan. Few women who were invited did not consent to be asked question about their abuse experience. In total, the HCWs screened 104 women but as two forms were incomplete, 102 remained for analysis.

Evaluation of the screening

The observations

All seven HCWs who used the screening tool were observed at least once during the intervention period on their communication and interactions with the clients. All observations were performed by the main author (RL) who was involved in ticking action(s) that the HCWs took during the meeting with the women clients (See supplemental file 1/Appendix 1). When a HCW identified a woman with abuse experiences, it was expected of him/her to give advice to help or refer the client to Muhimbili National Hospital, located in the nearby district of Ilala. The average time spent with a client during screening was also recorded.

Focus group discussions

Focus group discussions were chosen to explore HCWs’ experiences and perceptions of performing the screening. We regarded the group interaction in FGDs important for exploring the variation in attitudes toward this type of intervention (34). All HCWs who took part in the training were invited to participate in FGDs. The HCWs who had implemented the screening tool were expected to share their experiences of the tool and the potential dilemmas faced. The others were expected to give their general views and perceptions regarding the introduction of routine abuse screening for women. The first author (RL) moderated all the discussions and one of the nurse managers facilitated the logistics. A thematic guide was used with flexibility. The guide consisted of central topics covered during the workshop, their perceived roles and responsibilities, confidentiality issues, preventive measures, legal assistance, policies and guidelines, and suggestions for organizational change (see Appendix 2). The FGDs lasted 60–80 min, were audiotaped, and transcribed verbatim into Kiswahili to enable preliminary analysis. Later, they were translated into English to facilitate peer-debriefing sessions and joint analysis within the research group. All seven HCWs who had used the screening tool and another 13 HCWs who had not used the tool participated in the three FGDs. The first two groups were mixed in terms of professions, sex, and with respect to whether the participants had used the tool or not. The last group consisted of nurses and midwives who had not used the screening tool (Table 2).

Table 2

Basic characteristics of informants in the three focus groups

	Sex		Profession
Group number	Women	Men	Clinicians	Nurses and midwives	Age range (years)
1	3	4	4	3	29–49
2	2	5	5	2	32–48
3	4	2		6	30–54
Total	9	11	9	11	29–54

Written narratives

Five of the seven HCWs who had used the screening tool also wrote short narratives about their experiences and views regarding the screening. They were asked to retrospectively reflect on the implementation of the tool. This included not only giving their general opinions about the usefulness of the tool but also commenting on the challenges they faced or foresee for future abuse screening.

Data analysis

The variables included in the quantitative analysis of women's responses to the screening tool are related to their experiences of emotional, physical, and sexual violence. Women were asked whether they were abused or not. Those who responded ‘yes’ were further asked about their relationship with the perpetrator, their experience of abuse during the last year, the frequency of the abuse, and whether they were afraid of the abuser (Table 1). This information as well as the observation checklists was computerized using Excel and analyzed using SPSS. The information from the FGDs and the narratives were merged and subjected to qualitative content analysis (35). This methodology aims at creating a picture of a given phenomenon, in this case the screening, embedded within a particular context. The unit of analysis was the FGD. After reading the transcribed FGDs several times, the meaning units were summarized into condensed meaning units, still preserving the core meaning (36). These units were entered into the Open Code program (37) for systematic data organization to facilitate the open coding process. We developed six subcategories with three categories that constitute our overall interpretation of the manifest meaning of the text (Table 3).

Table 3

Condensed meaning units (text), codes, subcategories, and categories constituting the manifest content

Condensed meaning units (‘text’)	Codes	Sub categories	Categories constituting the manifest content
It's easy, needed to sit and talk to a woman following questions	Our job made easy	A sign of hope	Just asking feels good
	Sit, talk, tick, follow tool
We benefit from this our role is to learn how to talk to women in polite language	Communication skills improved	Work satisfaction
After we recognize woman having the problem what next?	Structures limitation	An ethical dilemma	What's next?
We have limited structures.
I know we refer if needed but to whom? We studied together	Same level of skills	Reinforce organizational change
Men source of problem be included and counseled	Men be involved	Neglecting other types of violence	Fear of becoming a ‘women hospital’ only
Women should be satisfied	Women supported

Ethical issues

This study was part of a larger study on IPV against women and children in Temeke District performed under the Reproductive Health Project of Muhimbili University for Health and Allied Sciences (MUHAS). MUHAS Ethical Committee granted ethical approval. Official permission to conduct the study was obtained from the Temeke district council administrative officer and from the District medical officer at the study area. We followed the ethical guidelines of research on violence against women approved by WHO/CIOMS (38). This implied asking for informed verbal consent by the women exposed to the screening tool as well as ensuring confidentially in the meeting between the HCWs and the women. Training of the HCWs included sessions to assure proper counseling and referral of women exposed to violence. However, as talking about physical and sexual violence may be emotionally sensitive also for the HCWs, the research team arranged for a counselor to support them if needed.

Results

Responses to the screening tool

Out of 102 women at the OPD who were asked about their abuse experience, 78% reported to have been emotionally or physically hurt during their lifetime. Out of these, 64% had been slapped, kicked, or otherwise hurt during the last year. Two out of three (62%) had been abused by their current or ex- husband/boyfriend, 22% by a relative, 9.2% by somebody at work, 4.6% by strangers, 1.5% by neighbors, and 1.5% by a friend/girlfriend. Two-thirds of the women (64%) had been physically hurt more than once during last year and 14% had been physically hurt several times. Almost one out of four (23%) women had been forced to participate in sexual activities during the last year. Only one woman who had been physically hurt by a current or former husband/boyfriend reported to be afraid of him.

Communication, guidance, and counseling

Generally, the HCWs had good communication with their clients and the affected clients were guided on what to do or where to go. Out of the seven HCWs who were observed, five greeted and interacted with women in a polite and respectful manner, four had to ask women additional questions, five listened to what the women had to say and probed her adequately, six clearly stated that the information would be treated confidentially, and six gave appropriate guidance. However, six of the HCWs also forwarded their personal opinions without following the procedures in the study protocol and the techniques for motivational interviewing presented during the training. Among the four HCWs who conducted counseling, three of them did not follow the basic steps in counseling. During screening, three of the clients seemed to be troubled to discuss about their experiences of abuse with the HCWs. The average length of the screening session was 15 min.

Focus groups discussions and short reflective narratives

The analysis of the FGDs and the written narratives resulted in three main categories reflecting different attitudinal sets toward the screening. Just asking feels good indicates a positive position toward routine screening. What's next? illustrates the frustration felt for not having the resources needed for proper support and referral, and Fear of becoming a ‘women hospital’ only. shows a concern that to focus only on men's violence against women in intimate relationships may mean neglecting other types of violence. The three categories (bolded as headings) and their related subcategories are incorporated in the following description. Quotes from the informants are given to illustrate how the interpretation is grounded in the text from both FGDs and narratives. The analysis is summarized in Table 3.

Just asking feels good

The informants gave many explanations for the perceived advantages of using the tool. They mentioned how the tools provided them with adequate information for appropriate diagnoses, as opposed to the existing situation where they often had to struggle for information to confirm abuse. They also mentioned the advantages of having information on the number of women at their hospital who had been exposed to IPV. Some HCWs found the intervention to be a sign of hope as the tool created a window of opportunity for them to support abused women. They recommended the tool to be used because it was seen to make their job easier, and the short and clear form with few questions to answer made it worth the effort. They realized that using a few minutes allowed them to learn about suspected abuse and that a few more minutes were enough to counsel or refer the exposed women.

D4: The form is good. The only problem is that it needs more time to spend with one client. Anyway it will be a good entry point for us to start talking about IPV and may be a relief for women in future. I feel good to ask them. We only need to sit and talk to a woman following those questions. (FGD1 Male Clinician)

The subcategory work satisfaction came up when the HCWs were talking about the intervention giving them a workable tool to reach abused women who attended their hospital. The HCWs also discussed how the tool allowed the HCWs to gain skills in terms of practicing therapeutic communication.

D3: One of our roles is to learn how to talk to women politely, because these women have been injured; they are in pain even if not physical injuries. They are affected psychologically. Polite language will make them tell us the information we need to help them…Yes and I agree with the other speaker that this work (intervention) will be useful to us and a relief to many women. (FGD1 Female Nurse)

What next?

This category represents a more cautious attitude among the informants indicating a worry that the use of the tool could cause an ethical dilemma, if the help offered to the abused women is not adequate. In the existing referral system, the abused women would risk being handled by untrained colleagues with limited competence in providing IPV care and other medical support. The informants also emphasized the risks involved in not being able to handle emotional reactions due to lack of time and having too many clients. Even if representatives for this type of attitudes suggested the screening to be part of routine history, the risks involved made them suggest that the tool should only be allowed when the hospital was ready to employ more staff and could ensure confidentiality. Otherwise, the HCWs felt they had not much to offer to the women who were experiencing IPV. This category thus represents an uncertainty as to whether the health care system is ready for routine screening for IPV and suggests a need for reinforced organizational change. The identified gaps were seen as challenges resulting from an inadequate work environment in terms of structures and adequate skills to attend clients identified with IPV experiences.

D3: It is also difficult to examine a patient in front of another one even if we use curtains. There is one examination bed for two of us and when you ask questions about STD patients feel embarrassed. Although we try to use low voice, people like to listen to others’ conversations. (FGD1 Female Clinician)

The lack of capacity to handle clients who experience IPV also limited HCWs’ motivation for taking on the extra work to implement routine screening. The HCWs agreed that primary prevention was important, but it was to their opinion that health care policy makers should get involved.

D5: This is true but our role is to treat while primary prevention is necessary but others should do it. We should also inform the MOHSW about these cases and what we have been doing. (FGD3 Male Nurse)

Thus, even if this theme indicates an overall positive attitude toward introducing routine IPV screening, the prerequisite is clear that there have to be adequate resources.

D1: You know I have nothing much to say but would like to do the screening -the resources are my dilemma. Many women are poor ‘wanyonge’ and are not strong enough to fight with their husbands but maybe this would be their good start. They will be happier later in future. (FGD 3 Female Clinician)

Fear of becoming a ‘women hospital’ only

The most negative attitudes about using the screening tool were aired by the male HCWs. To some of them, the screening tool could be seen as neglecting other types of violence that affect children and men. They were concerned with low budgets, shortages of staff, and inadequate infrastructure and perceived it to be unfair to conduct IPV screening to women alone, while other groups were also affected by violence.

D2: The hospital being a place for both sexes. We should not over emphasize for services to women alone in this intervention. We should take care of them just as we do to all other clients. (FGD3 Male Nurse)

Women were seen as privileged, having access to many different kinds of free services such as reproductive, maternal health care, and even provision of free insecticide bed nets. There was a fear that TDH would turn into a women's hospital only. Men had to contribute to all consultation fees and treatments. For them, to include men in IPV screening than women alone with other services in the future would be more cost effective. To their opinion, IPV-affected women should be treated just like others and the same routines should be followed. The HCWs reflected on the existing heavy workload and viewed as unfair if the management decides for the HCWs to allocate their time and their few resources to care for only one group of people, meaning the women.

D7: In one shift we normally attend up to 60 plus in a room for the two clinicians. Sometimes we reach up to 100 clients when it is a busy day, but if we are to attend one client at a time then it will be only 15 clients per day in a room. Where will others go? To show that we are not working.well at TDH, they will blow the news on ‘Nipashe.’ (A local newspaper) (FDG 2 Male Clinician)

However, eventhe claim that women were well-supported proponents of these attitudes acknowledged the need to help affected women.

D6: On the other hand if we manage to detect women with such problems we may solve most of the frequent complaints from women. When their husbands learn that their wives are asked about abuse, some may stop their abusive behavior because of shame to be known by the hospital staff, good to continue with this move. Those men who are rude may not bother but a certain percentage of them may stop abuse. (FGD 1 Female Clinician)

Other HCWs (mainly women) suggested including men in the screening but mainly due to their role as perpetrators in need of treatment.

D5: My views are to men who are the source of the problem, so to say in this discussion. Can't they also be included for counselling? They (men) say that women should be satisfied with the present care. Women experience more illhealth than men. Couple counselling services may be in future. Men perpetrators also need help, informed about the services even through radios. It may work well for everybody later. (FDG 2 Female Nurse)

A few of the male HCWs blamed women who experienced IPV. They were seen as having failed to adhere to marriage norms. These HCWs seemed to accept wife beating and regarded screening for IPV as a waste of time and resources.

D2: We may have skills for counselling but we have no time with such clients because of pressure of work. Some of these patients are themselves to be blamed. You know some women don't want to be polite to their husbands and adhere to the norms in their marriages, that is why they are beaten. It takes time, need to be more patient and expertize to screen.which we miss. It may be too costly for training. (FGD3 Male Nurse)

Discussion

Our findings confirmed that IPV is highly prevalent in the study setting and that screening for abuse experience is feasible. Among the women who reported to have been emotionally or physically hurt during their lifetime, 62% had experienced IPV. Violence from other family members and at the workplace was also surprisingly high (22%). The HCWs had good communication with their clients and mostly guided them on what to do next in a respectful and appropriate manner. Still, some of the women seemed hesitant and troubled with the conversations. The HCWs attitudes toward the use of a screening tool could be categorized into three groups. ‘Just asking feels good’ represented a positive view where screening was seen as an opportunity to improve the quality of care and routine assessments of women clients. Using the tool, justified the HCWs to spend more time with clients for consultations. ‘What next?’ indicated the challenges posed by inadequate resources and the ethical dilemmas in having inadequate number of consultation rooms and concerns about the competence of the staff at the referral points. ‘Fear of becoming a “women hospital” only’ questioned the whole idea of screening only women for IPV. The claim was that it would be unfair and an underutilization of scarce resources. The available services should be offered to both men and women.

Screening in this study was defined as a step by step procedure where HCWs asked all clients about their experience of violence. This type of screening is encouraged to improve IPV detection, provide thorough clinical assessments, and provide prompt medical support to those who experienced IPV (39). In our study, HCWs reported to feel good to ask women about abuse using a tool because it made their work easier to elicit information from women. We indicated in our category ‘Just asking feels good’ that the HCWs expressed to be satisfied with their work when they managed to identify women who experienced IPV with less effort. The women who experienced IPV were guided, treated, and referred in a better manner and quicker than before. Studies in Sweden and USA conducted with midwives and health care workers have also indicated that women accept being screened for abuse and that they feel comfortable talking about violence because screening also is seen to reduce women's isolation (40, 41).

A review of screening studies conducted by Waalen and coworkers from the USA has reported specific barriers to screening among HCWs. These include lack of education and limited time to screen, fear of offending clients, and lack of effective interventions after identification of IPV (42). Another study among midwives in Northern Sweden showed that without training on IPV, they will not ask pregnant women direct questions about abuse unless they had strong reasons to do so (43). Other studies from the USA and Sweden have emphasized the need for training, provision of standard protocols, and specific services for women exposed to violence (44, 45). Our category ‘what next’ indicates the same challenges of being confronted with lack of skills and guideline, no adequate rooms for screening sessions, inadequate time, and heavy work load.

The role of training was discussed by Fawole et al. (46) in a cross-sectional study among primary HCWs in Nigeria. This study reported that the HCWs who had previous training on IPV were three times more likely to screen competently than those without. However, training alone may not necessarily make HCWs effectively screen without other resources. In our study, the HCWs’ attitudes toward screening of women alone was a concern, which could also hinder HCWs to screen women. A qualitative study among certified nurse midwives in the USA has indicated that despite specific training, midwives found it difficult or were resistant to screen consistently following the screening protocols (47).

A challenging argument against routine screening for intimate partner abuse includes the lack of evidence that screening actually reduces the risk of IPV against women. A trial in Canada that evaluated women's health outcomes from screening in 12 primary care sites concluded that despite the fact that the screening cannot be deemed harmful, there is not enough evidence on reduced risks to women to warrant for advocacy on routine abuse screening for all patients (48). In our study, we found a different argument against screening, where some HCWs viewed routine screening to be unfair because of neglecting other types of violence and suggested both men and children to be included in the screening process in the future. Women were seen as privileged and already well supported. Some male HCWs even blamed the women themselves for being exposed to violence. Such blaming attitudes have also been reported from a cross-sectional study on HCWs’ readiness to screen for IPV in northern Nigeria (49). In this study, male HCWs were found to have stronger blaming attitudes than their women counterparts.

In our study, few women who responded to questions about abuse were seen to be hesitant and troubled to disclose violence experience. Spangaro and coworkers (50) have reported similar types of adverse effects. In their study, 6.0% of the 119 women screened reported not to be helped by screening and five women had feelings of sadness after recalling the abusive stories during a screening session.

Our study adds to the literature on screening by showing that a simple screening tool can detect alarming figures about the extent to which women experience violence. IPV against women was the most common type of violence; but in this setting, there is a cause for concern about other types of abuse from relatives, at the workplace, and violence perpetrated against men and children.

Despite challenges for screening women in patriarchal low-income societies, where the prevalence of violence is known to be high, yet socially hidden, screening for abuse in the health care setting may open a gate for other interventions. Zapien et al. (51) claimed that determining the magnitude of violence is a first step for change that may lead to other interventions to improve the health of women exposed to violence.

Trustworthiness

Our study is hospital –based, and the sample of women being screened is not representative of the general population. In addition, our sample size is small and limited to those women available during the early morning sessions and in a health condition that allowed them to answer the screening question. These factors may have biased our results in different directions. The fact that only women seeking care were screened may have caused an overestimation of the prevalence, whereas the actual selection procedures may have caused an underestimation of the prevalence of violence if compared to the general population. However, despite this, we could show that the screening tool was able to identify violence exposure, even if we cannot make any conclusions about IPV prevalence.

The study area is characterized by a patriarchal system where women are culturally considered to have a subordinate status and minimum influence on decision making and the HCWs were not immune to this social system. Although this may limit the application of the findings to other contexts, there are many settings with similar gender norms that may recognize the perceptions and concerns regarding screening that we have described. The discussions in the FGDs represent different types of health care workers and both men and women. Repeated visits to the study site, discussions within the research team (peer debriefing),and member checks with the few respondents increased the study credibility. Open coding categorizing and looking at the relevance of the themes to the research questions were tested by constant oscillation between the text, codes, and subcategories and by posing specific attention to outliers or negative cases.

Conclusions and recommendations

IPV screening can be conducted by HCWs also in resource-poor settings. The pilot intervention facilitated a better working environment that enhanced HCWs’ and women's ability to talk about abuse. The training gave HCWs an opportunity to update their knowledge about gender issues and increased their diagnostic assessment skills. The high prevalence of IPV may not only motivate screening also for other types of violence in the future but also support multisectoral efforts to change the current normative acceptance of violence among individuals, families, and communities. However, conducting screening on violence experience in resource-poor settings poses ethical dilemmas as HCWs may not be able to fulfill even the necessary requirement of privacy during screeningwhere the women may be left without getting adequate support. This suggests that for screening for violence to be ethically defendable, these minimal resources have to be secured.

No	Actions by the trained healthcare worker	Responses
		To a great extent	To some extent	Not very much	Not at all	Comments
1	Greets the woman in a polite and respectful manner
2	Is clear in asking the woman for consent to ask some additional questions
3	Interacts with the woman in a polite and respectful manner
4	Listens to what the woman has to say and probes adequately
5	Gives her own ideas and recommendations
6	Counsel the woman according to the training curricula
7	Give correct information about referral possibilities
8	Informs about what it implies that the information will be treated confidentially
9	Woman observed to be happy with the conversation
10	Ends the discussion with appropriate guidance for the woman