The study was a superiority parallel single-blind randomized controlled clinical trial with two groups of intervention and control in pregnant women at risk of domestic violence.
Setting and participants
The inclusion criteria comprised minor and medium levels of domestic violence in physical, psychological, and sexual subscales of the CTS-2 in mothers, being over 18 years old, being less than or equivalent to 27 weeks of pregnancy based on Last Menstrual Period (LMP) or ultrasound results, literacy, living in the city of Zanjan, being married for at least a year, having no participation in any other classes or counseling courses simultaneously, having cell phones, absence of underlying illnesses, willingness to participate in counseling sessions and living together as a couple. The exclusion criteria included known psychological illness, psychoactive drug consumption, signs of addiction to drugs in pregnant women and their spouses, and being absent in more than one counseling sessions.
The recruitment and the intervention were done in childbirth preparation class in a comprehensive urban health care center in Zanjan, Iran. At the time of the study the center had the only governmental childbirth preparation classes in the entire city. Pregnant women learn about pregnancy, childbirth and breastfeeding, and do exercise related to pregnancy in these classes. In the intervention group, counseling sessions based on solution-focused approach were held weekly for a period of six weeks in the form of 90- minute. Counseling sessions were conducted individually and were held by the researcher. The themes in each session were as follows: the aims of the first counseling session include highlighting general principles of solution-focused counseling and providing proper definitions of problems to clients. The second session focused on the familiarity of participants with the concept of quality of life and solution-focused approach. During the third session, clients learnt that there are different interpretations for an event and that they can develop the best interpretation in their minds while in the fourth session, clients were encouraged to discover exceptional opportunities of living as a couple. In the fifth session, with the help of miracle questions, participants were able to recognize their destructive behavior patterns. One of the main elements in this approach is the miracle questioning. This question actually encourages the person not to think about how and the chances of achieving the goals. Instead, the question arises as to what would occur to them if miraculous things happen, such as a miracle. This question helps people to have a very positive and different attitude towards their lives. With this question, one's mind shifts its focus from the cause of what is happening. Instead, the mind goes in a direction where the person has nothing as a problem. In the sixth session, a conclusion was made from the whole previous sessions to help the clients replace and experience their former thoughts and behaviors with the new ones. The control group received no intervention. Since the exchange of information related to counseling sessions between the intervention and control groups could distort the results, the sessions for the participants of the intervention group were held in a separate place (second floor of the building) from the sampling location and the intervention group was urged not to talk about the material presented to other people. Six weeks after the last counseling session, both groups were invited to complete the study tool for the second time. At this time the session was conducted in the presence of an interviewer who was completely unaware of the grouping procedure. In order to comply with ethical issues, after the questionnaires were completed at the follow-up period, the sessions were held for individuals in the control group who were willing to participate in these sessions. Ethical principles of privacy, anonymity and confidentiality were applied to the research project.
The outcomes of the present study included domestic violence and the quality of life in pregnant women who had been exposed to domestic violence which were measured and evaluated between the two intervention and control groups at six weeks follow ups.
Based on previous study , the sample size was calculated for each group of 45 people with an error of 5% and a power of 80%.
Participants were selected through a convenience sampling method and divided into two intervention (A) and control (B) groups using, quadruple random blocks. At first, all sequences of participants in each block was considered and then 23 blocks sequence was selected from random table number. The overall number of 90 pregnant women entered into the study from which 45 individuals were allocated to intervention and 45 individuals to the control group.
Data collection tools
Research tools for collecting data included demographic and reproductive checklist, CTS-2, and health-related quality of life questionnaire (SF-36); which were completed by qualified women. Demographic data included couple's age, couple's employment, couple's residential address and education status, marriage longevity, economic status, having cell phones, number of marriages and having children from the previous marriage. Reproductive data related to pregnancy status included pregnancy duration, frequency of pregnancy and delivery, intended and unintended pregnancies, and infertility history. In order to identify whether pregnant women were at risk of domestic violence; CTS-2 was applied. The questionnaire is the newly revised version of the conflict tactics scale in addressing marital conflicts which was developed by Straus et el. (1996) with the confirmed Cranach's Alpha of 79% for physical violence and 86% for psychological violence .
The applied questionnaire in the present study was validated by Ardabily et el. (2010) with the Cronbach's Alpha of 80%  which comprised 36 questions studying the conflict tactics on different dimensions of negotiation, physical, psychological, and sexual violence that leads to injuries. Although, the applied criteria in measuring the frequency and severity of violence in the CTS-2 is considered to be from last year, according to the questionnaire designer, one year is predetermined and could be generalized to the desired time expected by the researchers . Thus, in the questionnaire associated with the present study, instead measuring the previous year, three months' period was accounted for measurement. The SF-36 was applied for measuring the quality of life of participants, which included 36 questions and its validity in Iran has been verified by Montazeri et el. (2010) with the Cronbach's Alpha of 77% to 90%. This tool evaluates the quality of life in two dimensions of physical and psychological health. Each dimension contains four subscales. Physical subscales include physical function, role limitation-physical health, bodily pain and general health and psychological dimension include vitality, psychological health, role limitation-emotional health, and social functioning .
The collected data were analyzed using SPSS V.16 and R statistical software. In order to study the normal distribution of variables, Kolmogorov– Smirnov test was applied. Qualitative data were analyzed using Chi-square and Fisher tests. In order to compare conflict tactics scale and the quality of life subscales between the two groups of intervention and control at baseline, independent t-test and Mann- Whitney U test were used. Since the residual distribution of variance between the two groups was not normal and there were no parametric covariance analysis (ANCOVA) assumptions, nonparametric ANCOVA was used to compare the variables between the two groups after the intervention. In this test, the outcomes are compared between the groups adjusted for the amount of baseline variables. Also, for comparing the conflict tactics scale, the quality of life, and their subscales between different phases of study, Wilcoxon and paired t-test were applied. The significance level of the aforementioned tests was based on P value <0.05. The economic class was calculated considering factors such as house foundation, housing, and transportation means. In addition to that, for calculating social and economic levels, factors such as job ranking, education, and economic class were taken into consideration.