Family violence among mothers seen at the University of Ilorin Teaching Hospital , Ilorin , Nigeria

Family violence (FV) is a term used to describe any form of violence against any member of a household or family by one of its members. Usually women and children are the victims and men the perpetrators of FV. Violence in the family is a complex global problem that results from biological, social and cultural factors, as well as having psychological and social consequences. It cuts across age, race, religion and socio-economic boundaries.

correlates and effects of FV among mothers attending a primary care facility in Nigeria, with the aim of gaining an understanding of whether screening for FV in the primary care setting in Nigeria would be beneficial.

Methodology.
A cross-sectional study of FV among 250 mothers attending the General Outpatient Department of the University of Ilorin Teaching Hospital was undertaken over a 5-month period.Data on the mothers' sociodemographic characteristics, and experience of FV and its psychosocial correlates and effects were collected using a semi-structured questionnaire and a 20-item Self-Reporting Questionnaire (SRQ) as instruments.
Data analysis.EPI Info version 6 was used to analyse the data.
Results.Sixty-nine mothers (28%) had experienced FV at the hands of their husbands.Of these women, 49 (71%) indicated occurrences within the preceding 2 years; in 17 (25%), the violence was severe enough to warrant a hospital visit or treatment.Mothers who experienced FV were significantly more likely to have had previous experiences of violence by an in-law; to have reported child cruelty by a husband; to have children with difficult behaviour; and to have reported that they were neglected by their husbands and not enjoying their marriages.They were also significantly more likely to have a high score on the SRQ and be identified as probable cases with psychological problems (SRQ score ≥5).The medical effects of FV are many, and frequent examples include behavioural disorders (anxiety, depression, diagnosable conditions such as acute stress disorders and post-traumatic stress disorders), physical injuries or even death. 1,2,7,11,16Children of women affected by violence may also suffer serious psychosocial or physical abuse. 9,17,18obally, there is consensus that FV is still under-recognised and under-reported within the medical community. 2,9,19,20The reason for this has been adduced to be the negative and neglectful attitude of the medical community towards FV, and to inadequate training in how to recognise and manage it. 19,20Other reasons that have been suggested are societal misconceptions about FV as a private or criminal matter (rather than medical) and the high tolerance threshold of women towards FV. 2,19,20ny women attending primary care facilities in Nigeria are known to have psychological problems from diverse causes, among which adverse life events are prominent; 21,22 the contribution of FV to this is as yet unknown.Clinical experience also indicates that many women afflicted by FV do not volunteer information without prompting.A simple, short, local screening instrument for eliciting information about FV in a busy outpatient situation will be useful.But the first need is to establish whether the opportunity provided by the primary care setting would be rewarding by ascertaining the extent of the problem, its nature and impact.
Few studies have been done in Nigeria to address FV in the primary care setting, and none to our knowledge in Ilorin.
Our present study therefore adds to the limited literature on reporting the frequency of FV, its correlates and effects.Most importantly, this study could provide a basis for planning initial intervention strategies and for conducting a long-term epidemiological study from which appropriate screening or diagnostic instruments and intervention methods can be developed.

Methods
This study was a cross-sectional investigation carried out at clinics in the General Outpatient Department (GOPD) of the University of Ilorin Teaching Hospital (UITH) over a 5-month period.The GOPD is essentially a primary care unit since all patients visiting the hospital are seen there first, except for accident and emergency cases.
The study population consisted of 250 mothers from a nonprobability sampling of consecutive married women attending the above clinics who met the inclusion criteria and gave verbal consent to participate.The inclusion criteria for the study were: (i) being currently married and (ii) having at least 1 child ≥6 years old and in school (this factor was used as an inclusion criterion because some questions enquired information on performance of the child at home and at school).Women who were widowed or currently divorced were excluded from recruitment for the sake of uniformity and ease of statistical analysis.
A questionnaire was designed for collecting the sociodemographic data of the mothers and their characteristics in relation to FV.It included questions intended to establish the following: (i) socio-economic characteristics of women and their spouses; (ii) the occurrence of previous physical assault, its frequency and the social context in which it occurred; and (iii) some of its physical and psychosocial effects on the victim and her children.
The SRQ was used to assess psychological disorder in the subjects.The SRQ by Harding et al. 23 was designed for use in the World Health Organization (WHO) study on strategies for extending mental health care.It was especially designed for screening of psychiatric disturbances in the primary care setting. 23The SRQ-1 (or 20-item SRQ) containing nonpsychotic questions was used.Each item in the questionnaire has a score of 0 or 1.
SRQ-1 was validated in a primary care setting in rural south-western Nigeria and found to effectively discriminate between patients with and without psychiatric morbidity. 24his was best done at a cut-off point of 5, which has the optimal sensitivity of 98.8% and specificity of 90.9%. 24The cut-off point of 5 was also adapted by another study 22 for assessing the prevalence of psychiatric morbidity among attendees at 5 primary health care centres in 5 towns of a rural local government area in south-western Nigeria.
The Yoruba version of SRQ-1 used by one of the authors (MF) in an earlier study in the same hospital, was used.The two instruments were pilot-tested using 25 mothers at another clinic in Ilorin.The pilot testing revealed that most items on the measures were comprehensible and easy to complete and could be administered within 20 -25 minutes.Only slight modifications were needed in the semi-structured questionnaire.
Although the instruments were to be self-administered, because of the low level of education among the group, and the need for uniformity, they were read out in private to each of the mothers in the appropriate language by one of the investigators (BA).

Results
A total of 280 mothers who met the inclusion criteria were approached to participate, but only 250 of them gave consent and were assessed; those who declined gave reasons such as lack of time and interest.
Of the 250 mothers, 193 (77%) were not ill but accompanied their children to the hospital; 29 (12%) had an acute illness, of whom 18 (7.2%)had malaria fever, and 28 (11%) had chronic illnesses such as hypertension and diabetes mellitus, with onset of illness <2 months and considered to be stable with no clinical complications.

Demographic characteristics (Table I)
In terms of age, 210 (84%) mothers were between 20 and 39 years, while 113 (45%) husbands were in the same range.
The mean age of husbands was 40.7 years (± standard deviation (SD) 7.5), which was significantly higher than that of mothers (32.3±6.0)(p=0.000).The majority of mothers (83%) and their husbands (90%) had no, or primary school, education and were engaged in unskilled or semiskilled work (71% of mothers v. 58% of husbands), indicating a not particularly literate population.Occupation classifications were as follows: • skilled -professionals with or without university degrees, and senior and intermediate civil servants • semi-skilled -artisans, junior civil servants and drivers • unskilled -labourers, petty traders, uneducated, unemployed and housewives.
The Yoruba ethnic group constituted about 83% of mothers and husbands (as expected by the study's location in a Yoruba-speaking town).Christianity was practised by 43% of the sample, while 57% practised Islam.

Family violence characteristics
Sixty-nine (28%) mothers reported having been beaten by their husbands; 49 (71% − 49/69) of these women reported that the beating occurred within 2 years of the study, and 17 (51% − 17/49) indicated a frequency of more than twice within the same period.Among the 69 mothers who had experienced FV, 17 (25% − 17/49) indicated that the injuries received were severe enough to warrant a hospital visit and treatment.

Correlates of FV (Tables I and II)
Mothers who had been beaten previously by their husbands were compared with those who had never been.The group that had experienced FV previously was found to be significantly more likely to report the following: (i) to have husbands of low educational (≤ primary education) and occupational (≤ semi-skilled) levels; (ii) to be in marriages ≥10

Impact of FV (Tables II and III)
Mothers who had experienced FV previously were significantly more likely to score higher on the SRQ and be identified as probable cases with psychological problems (SRQ score ≥5).Of the 69 mothers who experienced FV, 28 (41% − 28/69) had an SRQ score ≥5, compared with the 30 without FV experience (17% − 30/181) (p<0.000).These figures suggest that FV probably contributed to psychological problems in these mothers.Also, the mean SRQ score of mothers (4.2±2.8)who had experienced FV previously was significantly higher than those who had never experienced it (2.2±8.0)(p<0.000).
In addition, victims of FV were significantly more likely to have children who were under-achieving at school, and with difficult behaviour or bed-wetting after 7 years of age.FV also adversely affected marriages: of the 17 women who indicated that they were not enjoying their marriages, 10

Discussion
With a prevalence rate of 28% in this study (that was restricted to physical violence and a specific group of women), family violence is probably more common in primary care settings in Nigeria than is realised.Primary care practice provides a setting that offers better scope for women to discuss issues such as FV.It is time to take advantage of this opportunity, by way of primary care providers carrying out assessment, intervention and/or appropriate referral of women in violent relationships.
Compared with two other studies in Nigeria 9,13 and some in Africa, 10,12 the prevalence rate of FV in this study is relatively low.The reason may lie in the fact that this study was restricted (as described above), as well as its hospital and urban base, with possible cultural differences possibly also contributing.
N = total number of mothers in the study population = 250 N 1 = total number of mothers who reported at least one incidence of family violence N 2 = total number of mothers who reported no incidence of family violence.n 1 and n 2 are the percentages of N 1 and N 2 respectively to the nearest whole number.

Sign. level (p-value) articles
In terms of correlates, socio-economic status as measured by mothers' educational and occupational levels did not influence the occurrence of FV in this study.This is consistent with the pattern observed in Sierra Leone 12 but contradicts some other studies. 2,9However, husbands with low educational (≤ primary education) and occupational (≤ semi-skilled) levels were more likely to beat their wives.This suggests an inverse relationship between these factors and FV, which is consistent with the pattern in a previous study in Nigeria. 9Husbands indicted as neglectful, behaviourally disturbed or childabusing in this study were also likely to be 'wife-beaters'.This pattern probably suggests progressively dysfunctional family situations.
Family structure and formation factors were found to influence FV in this study.For example, victims of FV were more likely to: (i) have ≥5 children, (ii) have experienced in-lawrelated violence; or (iii) be living with an in-law.Based on the prevailing sociocultural environment, it is reasonable to expect this type of family structure or formation in lower socio-economic groups, where it is possible for husbands to be 'culture respecting' enough to permit family members to physically abuse their wives.
Despite strong evidence in many studies 1,9,14 linking substance abuse (especially alcohol) to FV, it was found not to be influential in this study.The small size of our study was probably a major factor; a larger study would be necessary to fully explore the relationship of this variable with FV in the primary care setting.
The consistent global pattern of adverse physical and psychosocial effects of FV on mothers 2,9,13,15,16 is also supported by this study.A substantial number of mothers (25% − 17/69) suffered injuries severe enough to seek medical attention.In addition, a significant number of mothers who had experienced FV were identified as probably having psychological problems by the SRQ instrument (41% − 28/69).When this figure is compared with that of 28% found in mothers of children with sickle cell disease by the same instrument in the same setting as this one, 25 the severity of the psychological distress of FV can be better recognised.
Furthermore, as part of a complex family disorder, the children of victims also suffered.They were more likely to be underachieving at school, to have behavioural problems and be bedwetting after 7 years of age.In order to identify, manage or prevent FV, a holistic or 'biopsycho-social' approach to health care 26 by primary care practitioners is crucial.While public education, aimed at men, on the harmful effects of FV is a good strategy, husbands of victims should be invited for counselling and education in every case of FV.Children of victims should also be assessed for physical and psychosocial problems. 27e of the major limitations of this study was its restriction to only physical violence -to the exclusion of emotional, verbal and sexual violence.It was additionally restricted to currently married mothers with children ≥6 years attending the outpatient department only, excluding mothers from the emergency and other departments.All this put a limit on its generalisability and applicability.Considering that it is a preliminary study, opportunity still exists in the future to improve these defects.
The tertiary hospital and urban-based nature of the study, although carried out in a primary care setting, may also restrict the application of its findings to the wider community -especially rural areas -where the level of health care and socio-economic status are likely to be very different.

Conclusion
This study has provided evidence that women at primary care level in Nigeria not only experience FV but that it also negatively affects their physical and psychosocial well-being and that of their children.Unless primary care providers take interest in the psychosocial care of these victims, the majority of them will continue to suffer in silence.
Moreover, based on the findings of this study, it can be reasonably concluded that screening for FV in primary care settings in Nigeria would be rewarding, with potential for improving women's health.We therefore recommend that primary care physicians increase their interest, knowledge, skill and research in the practicable and affordable identification and management of family violence.

Volume 14 No. 3
August 2008 -SAJP Conclusion.In view of these findings, screening for FV in the primary care setting would be beneficial.Primary care physicians should therefore increase their interest, improve their skill, and carry out more research in the identification and management of FV.Volume 14 No. 3 August 2008 -SAJP

articles Volume 14
No. 3 August 2008 -SAJP Data analysis was carried out using EPI-info version 6. Simple frequency tables were obtained and chi-square tests were performed to determine significant differences between various variables and correlates.Statistical significance was set at p ≤5.
years' duration; (iii) to have ≥5 children; (iv) to have husbands with a previous history of divorce; (v) that their husbands had behavioural problems or were neglectful; (vi) episodes of child cruelty by husbands; (vii) previous experiences of FV caused by an in-law; or (viii) having an in-law living with them.There was, however, no correlation between FV and the following: (i) age of the mother or her husband; (ii) mother's educational and occupational levels; (iii) type of family or marriage; (iv) absence of a male child; and (v) alcohol use by husband.

Table I . Sociodemographics, family characteristics of the mothers and correlates of FV
(59%) had experienced FV. articles

Table III . Positive items score on SRQ by mothers with and without FV report
is the total number of people in the study population = 250.N 1 is the total number of mothers who reported at least one incidence of family violence.N 2 is the total number of mothers who reported no incidence of family violence.n 1 and n 2 are the percentages of N 1 and N 2 respectively to the nearest whole number.