Acceptance of screening for Intimate Partner Violence, actual screening and satisfaction with care amongst female clients visiting a health facility in Kano, Nigeria

Abstract Background Healthcare providers have advocated for the screening and management of Intimate Partner Violence (IPV) against women and its consequences. Unfortunately, data from high income countries suggest that women may have varied preferences for being screened for IPV in healthcare. Although women's preference for screening in sub-Saharan countries has not been well researched, IPV remains an accepted societal norm in many of these countries, including Nigeria. Objective The objective of the study was to assess women's acceptance of screening for IPV in healthcare, the extent to which inquiry about IPV was carried out in healthcare and whether such inquiry impacted on satisfaction with care. Method Data on these variables were gathered through structured interviews from a sample of 507 women at a regional hospital in Kano, Nigeria. The study design was cross-sectional. Results The results found acceptance for screening in the sample to be high (76%), but few women (7%) had actually been probed about violence in their contact with care providers. Acceptance for screening was associated with being married and being employed. Actual screening was associated with ethnicity and religion, where ethnic and religious majorities were more likely to be screened. Finally, being screened for IPV seemed to improve satisfaction with care. Conclusion The findings demonstrate the need for adaptation of a screening protocol that is also sensitive to detect IPV amongst all ethnic and religious groups. The findings also have implications for further education of socio-economically disadvantaged women on the benefits of screening.


Introduction Setting
A recent development in the prevention and control of Intimate Partner Violence (IPV) has been the involvement of healthcare professionals through screening for the practice amongst their female clients. 1,2 Yet, data from high income countries suggest that women may have varied preferences for being screened for IPV in healthcare. 3,4,5,6 Although women's preference for screening in sub-Saharan countries has not been well-researched, IPV remains an accepted societal norm in many of these countries including Nigeria. 7,8,9,10,11,12,13,14,15 In fact, women in these settings appear to endorse IPV to a higher degree than the men, 9,16 raising questions as to whether female clients in such settings may accept being screened for IPV. Moreover, institutionalisation of IPV seems apparent in many parts of sub-Saharan Africa. In Northern Nigeria for example, the penal code (Section 55d) permits a husband to 'correct' his wife provided no grievous hurt is inflicted in the process. 17 Such norms and laws have prompted women to blame themselves for abuse 18 and may also demotivate them from disclosing abuse. 19 In other settings outside Africa, women's preferences for screening have been attributed to a number of factors including, shame, fear of retaliation from the husband's relatives and an increased risk for divorce in case the husband finds out. 20,21,22,23,24 Furthermore, the protection of family honour, and the possible economic and emotional dependence of women on their husbands may further impact on disclosure and thereby on preference for screening amongst women. 20,25,26 Although the role of some sociodemographic variables (e.g. ethnicity, religion, education and literacy) for women's preferences for screening remain elusive, these variables have consistently been reported to affect women's attitudes towards IPV, 27,28,29 suggesting that they might also be related to women's willingness to be screened. This study will undertake to understand the extent to which women in Kano, Nigeria, are willing to be screened for IPV and whether this will vary depending on sociodemographic characteristics. Findings from this study could be useful in identifying demographic barriers to screening amongst clients and inform interventions to improve screening in the study setting.
Unlike in many developed countries, screening for IPV in healthcare is not yet routine practice in most middle and low income countries. 30,31 Still, an assessment of provider screening at own initiative may give some idea of readiness for routine screening in healthcare in low income countries. This study aims to assess the extent to which healthcare providers undertake screening on their own initiative in Kano, Nigeria.
Whilst screening for IPV has received significant advocacy during the past two decade, 2,32 assessment of the outcome and benefits of screening has not received equivocal attention. One such outcome measure could be satisfaction with care. Satisfaction with care has recently emerged as an indicator for good medical practice that includes screening for IPV. 31 Whether or not screening for IPV improves women's satisfaction with different aspects of care deserves further scrutiny. This paper aims to investigate this notion.

Significance of the study
Based on this background, the current study will assess women's acceptance for screening for IPV in healthcare, the extent to which inquiry about IPV is carried out in healthcare and whether such inquiry impacts on satisfaction with care. More concretely, answers to the following research questions will be sought: • What are women's preferences with regard to screening for IPV in Kano, Nigeria and do these preferences depend on sociodemographic factors? • How prevalent are inquiries about IPV in healthcare in Kano, Nigeria and does inquiries vary depending on the client's social demographic characteristics? • Does inquiry about IPV improve a client's satisfaction with care?

Ethical considerations
Ethical approval for this study was obtained from the Nigerian Institute of Medical Research, Lagos, Nigeria and permission was granted by the authorities of Aminu Kano Teaching Hospital, Kano. Participants' identification information (e.g. names and addresses) were not requested. The aims and relevance of the study were explained to participants when their consent was requested in a separate document accompanying the structured interview format (questionnaire). Voluntary participation was emphasised, privacy guaranteed and informed consent obtained.

Setting and materials
Data for this study were collected at the Aminu Kano Teaching Hospital in Kano, Nigeria. Screening was not routinely practised at this centre at the time of our study; therefore if screening was practised at all, it was on individual initiative of the health care providers. Aminu Kano Teaching Hospital is a fully equipped (tertiary) health institution that should provide tertiary health care services; however, the hospital provides primary and secondary care services in some instances. The Departments of Psychiatry and the Social Welfare have experts and personnel to manage victims of Intimate Partner Violence whenever they are referred to them. The study was based on structured interviews with women attending the General Outpatient Department, and the maternal and child health clinics at the hospital. Five female and one male assistant were recruited to assist in the interviews. All the female assistants were nurses or midwives of different grades. The male assistant was a final year medical student. Two training sessions on the study, that is, its aims, questionnaire administration and ethical consideration, were conducted for the assistants.

Design
The study design was cross-sectional. One in three women was politely approached by one of the assistants as they left the consultation room. The project was explained briefly to the patient and if she consented, she was escorted to an adjoining room where the second assistant was stationed. The procedure and project was explained in more detail and the participant was assured of the confidentiality of the study, as well as reminded that she had the right to withdraw at any stage without any personal repercussion. Every time after an interview was completed, the next woman visiting the clinic would be approached until responses from 507 women were achieved. This sample size was based on a power analysis assuming a binomial distribution with a prevalence of IPV in developing countries of 0.12 estimated from previous studies, 34,35 a statistical significance level (alpha = 0.05) and a power of 80%. Forty women declined to participate in the study.

Procedure
Each interview was performed by a pair of interviewers using structured questionnaires. Participants' preferences for screening for IPV were assessed by asking participants to describe how they would feel if asked about abuse in their intimate relationship in a healthcare situation. The concept of abuse was explained to each participant that consented to be interviewed. An Acceptable response implied that the participant expressed a clear and positive attitude to being questioned about IPV.
A Neither acceptable nor unacceptable response implied that the client expressed no sign of being offended by the question, nor any overtly acceptable expression. A Both acceptable and unacceptable response implied that the client expressed a positive attitude to being asked about IPV, but also expressed some disagreement to being asked such questions, that is, a mixed response. An Unacceptable response by the participants implied offence or discomfort expressed when questioned on IPV. These categories have been determined previously in qualitative studies and responses validated. 3 Actual screening was assessed by asking the interviewed women whether they had been questioned by their health care provider during the present contact about the possibility of IPV. Response alternatives were 'yes' or 'no'.
Clients' satisfaction with care was probed with the pyramid patient questionnaire, a previously validated instrument with three subscales. 36 Nurses' competence and skills were assessed by the subscale 'Nursing Staff' (this subscale consists of three items. Cronbach's alpha for the current sample was 0.87). Contact with the staff was assessed by the subscale 'Contact' (this subscale consists of three items; Cronbach's alpha for the current sample was 0.83). Support of the Client was assessed by the subscale 'Social Support' (this subscale consists of four items; Cronbach's alpha for the current sample was 0.83). For each item under these subscales, the response alternatives were scored using the Likert's score of 1-4 (1 equalled 'Not at all'; 2 was 'No'; 3 was 'Yes'; and 4 equalled 'Strong Yes'). High scores, therefore, reflected higher satisfaction both in specific items and subscales.

Sociodemographic
Indicators including age, marital status, and the number of children, profession, religion, ethnicity, literacy level, educational level and employment status were recorded and the response alternatives for these variables were noted (Table 1).

Analysing
The Chi-square test was used to assess associations between participants' preferences for screening and sociodemographic variables. There were only a few participants in two categories of the variable measuring acceptability for screening (i.e. 'Neither acceptable nor unacceptable' and 'Both acceptable and unacceptable') and consequently a dichotomous variable was formed with the other two options, that is 'Acceptable' and 'Unacceptable'. Associations between actual screenings for IPV and the different satisfaction with care subscales or total satisfaction was assessed using the student t-test. Statistical significance was assumed at p < 0.05 and SPSS (Statistical Package for the Social Sciences) version 16.0 for Windows was used for all the analyses.

Demographic characteristics of participants
The majority of participants were married, Muslims, of Hausa or Fulani ethnic belonging, unemployed, literate and with at least primary school education ( Table 1). The average age of the sample was 29 years with a Standard Deviation of eight years.

Acceptance for Intimate Partner Violence inquiry and actual Intimate Partner Violence inquiry
The majority (76%, n = 355) of the participants found it acceptable to be probed about IPV in healthcare and almost 20% regarded such inquiries as unacceptable ( Table 2). The healthcare provider had probed 7% of the interviewed women on the possibility of IPV during their latest visit.

Association between acceptance of Intimate Partner Violence, actual Intimate Partner Violence inquiry and sociodemographic factors
The marital status (c 2 [2] = 9, 49, p < 0.01) and employment status (c 2 [1] = 4, 4, p < 0.05) are indicated in Table 3. There was no significant association between acceptance for screening and the other demographic variables.
Ethnicity (c 2 [1] = 5,6, p < 0.05) was associated with being probed about IPV in healthcare (Table 3); in other words, participants of Hausa or Fulani ethnic group were more often probed about the possibility of IPV in healthcare than the other ethnic groups together (i.e. migrant ethnic groups in the region). A trend was observed regarding the association between probing about IPV in healthcare and religion (c 2 [1] = 2.8, p < 0.09), where Muslim participants seemed more likely to have been probed on IPV than other religions combined. There was no statistical association between being probed about IPV in healthcare and other demographic factors.

Association between actual Intimate Partner Violence inquiry and satisfaction with care
There was an association (  19; p < 0.001). Participants who had been probed on IPV in their latest contact expressed on average higher satisfaction in these regards than peers who had not been probed on IPV.

Discussion
The current paper studied women's preferences for screening, actual inquiry of Intimate Partner Violence (IPV) in Kano, Nigeria. The majority of women (76%) in our sample opted for IPV inquiry in healthcare. These figures are comparable to those observed in high income settings. In their Swedish sample, Stenson et al., 3 found that 80% of women in their study sample to opt for IPV inquiry. In a sample from New Zealand, up to 97% of clients opted for screening. 37 Although encouraging, the findings were rather surprising.
Studies from the region have abundantly found women to be reluctant to disclose abuse. 18,24,25,29,38 Thus an important implication of these results is that women find inquiry of IPV acceptable, but for some reason may not be in position yet to disclose if asked. Societal norms and laws that are gender restrictive may account for the failure to disclose. Further analysis of this discrepancy is warranted to disclose where, for some reasons, women decline disclosure of violence and hence did not accept inquiry on abuse.
Married women were more willing to accept IPV screening than were women with other civil status. Married participants are classified as 'currently in a relationship' and are consequently more likely to be affected by IPV and to find such inquiry warranted. These findings could be a reflection of such circumstances. Unemployment was also significantly associated with unacceptable preference for screening for IPV, a finding that concurs with previous research. 11,26 Women at the lower bracket of the socio-economic hierarchy are more likely to justify violence and may not readily disclose abuse. Their unemployed status may hinder such women from accepting screening because they are more likely to be economically dependent on their abusive partners. 8 Our findings revealed that only 7% of the clients were probed for the possibility of IPV by care providers, corroborating studies in high income countries where barely 10% of HCPs screen for IPV. 2,39,40 This has important implications for the initiation of a screening campaign in the hospital. Education and supervision of healthcare providers on routine screening for IPV has been shown to improve actual screening in some settings. 33,41 Our results demonstrated further that ethnic and religious majorities (Hausa or Fulani and Muslims) were more likely to be screened, suggesting demographic inequalities in the access to care (in this case access to screening). Recent studies from the same hospital suggest that care providers from ethnic and religious majorities were more likely to screen. 42 These findings together suggest that the migrant populations in the region may be denied care to a higher degree by their indigenous peers.
Our results revealed that clients who had been probed about IPV, exhibited higher satisfaction with care than peers who had not been probed, corroborating recent work in the field. 43 Screening may offer easy access to support services and management of immediate and later trauma of patients, thereby enhancing satisfaction. 37,44 Conclusion In summary, this study provided new insight on the acceptance of screening, actual screening and satisfaction with care. Acceptance for screening was exceptionally high in a societal context where women's acceptance of IPV is high. 16 As expected, few women had actually been probed about violence in their contact with care providers. Acceptance for screening was associated with being married and being employed. Actual screening was associated with ethnicity and religion, with ethnic and religious majorities more likely to be screened. Finally, screening for IPV seemed to improve satisfaction with care. The findings demonstrate the need for the adaptation of a screening protocol at the hospital, sensitive however to detecting IPV amongst all ethnic and religious groups. The findings also have implications for further education of socio-economically disadvantaged women on the benefits of screening.