In this study almost two fifths of the respondents had interpersonal conflict with their partner since their marriage or relationship time, out of which one third had a history of physical violence. The lifetime prevalence rate of intimate partner physical violence was 31.0% which was similarly reported in the study done in India 30% [19
], and Gondar (32.2%) [21
], and lower than the study reported from Palestinian refugees (42.5%) and the Jordanian refugee camps (44.7%) [22
], even lower than the study done among married women in rural Ethiopia which was 49.5% [16
This differences could be due to the fact that their marital relationship is arranged by themselves without family involvement shoed there is statistical significant difference between self arranged relation and experiencing of physical violence than family supported which was observed to be statistically significant. The differences could also ascribe to underreporting and cultural make up of family formation patterns. This variation can also be attributed not only to the differences in the levels of violence between settings, but also to differences in research methods, definitions of violence, sampling techniques, interviewer training and skills, and cultural differences that affect respondent's willingness to reveal intimate experiences. For these reasons, it is difficult to make direct comparisons or to make judgments between cultures or countries about in which society intimate partner violence is worst [23
The prevalence of intimate partner physical violence in the last 12 months found in the present study was consistent with the study done in Leon, Nicaragua (27%) and lower when compared to study conducted in the Republic of Korea (38%), and Palestinian women in the West Bank and Gaza Strip (52%) [24
]. Apart from methodological issues, the difference in magnitude of physical violence reported in different literatures and this study could be explained by socio cultural and societal perspectives and contexts of the population under study that differ between the nations.
The chance of experiencing physical violence in life time in this study was significantly associated with injury and multiple logistic regression showed that those women who are victims of partner violence were 37 times more likely to suffer injury in the last 12 months (AOR = 37.4, [95% C.I 14.4, 97.4]), which is consistence with other one study conducted elsewhere [25
], and also consistence with studies in Canada and the United States, which showed that female victims of partner violence are three times more likely to suffer injury, five times more likely to receive medical attention [26
]. Physical violence in the last 12 months was 2 times higher among women that the current relationship with their partner was arranged by couples agreement than those family supported relation (OR = 2.0 [95%C.I:1.2, 3.2]). After adjusted the risk is increased among women who know that their current partner has other wife than those that their current partner did not have other wife (AOR = 1.9[95%C.I: 1.0, 3.6]). This is consistence with study done by James K.et al. in Uganda the husband having another partner (OR = 2.4, 95% CI 1.0, 5.7) were associated with higher risk of intimate partner violence.
In multiple logistic regressions analysis level there is higher likelihood of experiencing physical violence in the lifetime among Muslim than those Orthodox Christian followers women (AOR = 2.4 [95%C.I: 1.1, 5.5]) Which is consistence with study done in Arsi showed significant difference between Muslim and Orthodox Christian [27
], the difference could be due the presence of most Saho ethnic and Kunama ethnics are Muslim in addition to this they were from rural origin and came from low social status. After controlling for other variables being experiencing physical violence in the last 12 months was 2 times higher among women whose current partners 'drink alcohol than those women's partner don't drink alcohol (AOR = 2.1[95%C.I: 1.0, 4.5]) this is consistence with study conducted around Gondar that physical violence was about five times more likely to occur among women whose male partners consume alcohol frequently and study in Aris zone, use of alcohol were associated with intimate partner physical violence [21
]. A Cross sectional study in Albania indicated risk of violence reduced with decreasing employment status of women: women in blue collar work, housekeepers, and unemployed women were at lower risk than those in white collar occupations [28
] which is different from the finding from this study that farmer women were more likely to experience intimate physical violence as compared to those housewife (OR = 5.6 [95%C.I: 2.3, 13.2]). This might be due to the difference in study population, sampling proportion, techniques to collect the sample and the socio cultural context where dominated by patriarchal belief for male dominance.
In this study it was less likely to experience physical violence in the last 12 months among women that can read and write, completed, primary and secondary school as compared to women in higher education level. It is less by 88.5% times among women whose partner was above grade 12 than women having illiterate partner (OR = 0.1[95%C.I: 0.0, 0.9]) and less by 68% times among the women whose partner was employee than those women whose partner was farmer (OR = 0.3[95%C.I: 0.1, 0.8]). This is similar with study done in Nicaraguan the risk increased with increasing educational level for women, with decreasing educational level for men [29
] and study in Bangladesh in 2005 husband's education beyond the 10th grade decreased the risk of violence [30
Those farmer women were 13 times more likely to experience physical violence in lifetime as compared to housewife (AOR = 13.0 [95%C.I: 3.7, 45.5]) and those women whose partner was employee were 93% time less experience of physical violence in the last 12 months as compare to those women whose current partner is farmer (AOR = 0.07[95%C.I 0.1, 0.4]). This finding was consistent with study done in South Africa factors include social and demographic characteristics of the men and women, their economic circumstances influence violence against women [31
Similarly study in south Asia in 2003 showed financial dependence on a partner or unemployment were the risk factors that contribute to this phenomenon [32
]. Similarly study in South Asia in 2003 shown financial dependence on a partner or unemployment were the risk factors that contribute to this phenomenon [33
]. After controlling for other variables being experiencing physical violence in the last 12 months was 2 times more higher among women whose current partners 'drink alcohol than those women's partner do not drink alcohol (AOR = 2.19[95%C.I: 1.0, 4..5]) and this is consistent with study conducted in Gondar, Northwest Ethiopia in that physical violence was about five times more likely to occur among women whose male partners consume alcohol frequently and as similarly reported from India, the commonest reasons for violence in household were husband's alcohol addiction [21
The strength of this study is the use of World Health Organization (WHO) instruments for physical violence which is developed enabled to make the comparison of findings with other national and international literatures to be valid. Before conducting this study enumeration of all household was done to identify married women with intimate partners. As limitation outcome was assessed only by the report of respondents. Therefore, under reporting of physical violence was inevitable due to unwillingness of respondents to disclose their personal information. Since this study was a cross sectional it is difficult to establish causes and effect relationships among outcome of interest and explanatory variable. There are also unusual large odds ratio and wide confidence interval observed in this study which might affect the precision. In addition there are also some variables that were not significantly associated with the outcome of interest. This might be due the small sample size to justify the relationships between the explanatory variables and outcome of interest. The observed counts also being small in some of the variables made the odds ratio so large and wide. Therefore any interpretation of this finding within these variables shall take into account the degree of precision. And also all variables associated with physical violence might not be exhaustively investigated. Beside this, since there is limited data in similar set up we are highly obliged to compare some figures with the study with population based survey which might have different circumstance and social, economic and cultural background. Therefore, readers who read this as a reference shall take in to account this as a limitation to the findings.