Sample and instrument
We conducted a cross-sectional survey at three clinics that provide Mother and Child Health Care in the West Bank. The clinics were located in the three largest cities in the West Bank: Jenin in the north, Ramallah in the center, and Hebron in the south. They were selected from a list of clinics that was provided by the Ministry of Health (MOH) based on the following criteria: (1) the clinics should provide most of the reproductive health services (antenatal care including high-risk pregnancy, postnatal care, immunization of pregnant women against tetanus, baby immunization, and family planning); (2) they should be the referral clinic from the surrounding villages and camps; and (3) they should be sufficiently staffed and equipped (as identified by the MOH). The purpose of criteria (1) and (2) was to ensure the availability of a large number of women with diverse socio-economic backgrounds coming for a range of services from the city and surrounding areas; and the purpose of criterion (3) was to facilitate the conduct of the study as we did not want to burden an understaffed clinic with our research study. A total of 84 clinics in the West Bank met our criteria . One clinic was in Jenin, two were in Ramallah, and four were in Hebron. In Ramallah and Hebron, we chose the clinics that saw the highest numbers of patients.
A questionnaire was developed on the basis of a literature review and adopted questions from previous studies including the Demographic and Health Surveys (DHS) and the national survey of Sultanate of Oman, some of which had been conducted in Arabic language [26, 27]. It assessed issues related to reproductive health: postnatal care, attitudes towards domestic violence, use of family planning methods and the understanding of reproductive rights. The questionnaire was first composed in English and then translated by two independent translators into Arabic. Cases of disagreement were discussed among both translators together with the first author (ED). It was pilot tested to assess acceptability and comprehension  among 30 women in the Ramallah clinic. Data from the pilot tests are not included in the analyses conducted for this paper. Based on results from the pilot test, the questionnaire was slightly revised and shortened from 35 to 25–30 minutes by deleting some questions and rephrasing others. Copies of the questionnaire in English or Arabic are available from the first author upon request.
The three clinics were visited 6 days a week from Saturday through Thursday in Spring 2006. Two female data collectors with medical or social science backgrounds, who had experience with collecting data in health care centers, assisted the first author (ED) in the fieldwork. Data collectors attended a three-day workshop conducted by ED to ensure a common methodology at all three clinics. Women were approached after they had received their health care in the clinic. Given the mean duration clinic visit of 5 minutes and the duration of both informing the women about the study and the interview (which took between 25–30 minutes), approximately every 5th, or 6th eligible woman was interviewed. Only four women refused to participate, resulting in a response rate of 99%. The interviews took place in a private room in each clinic. About 92% of women completed the interview without any company, while 8% were accompanied by some other person(s) mainly a sister, a mother and in rare cases a mother in law. A total of 450 women (186 pregnant and 264 postpartum) completed the interview during a 12 week period. Initially, a sample size of 100 women at each clinic was planned, assuming a 75% response rate. A total sample of 300 respondents would allow estimating proportions with +/-5% accuracy for the 95% confidence interval. Given the excellent response rate in the pilot phase, the available resources were used to aim for a sample of 150 responses at each clinic, which was accomplished. This sample size provides 80% power to detect differences of 16% or more between cities. The analysis in this paper is restricted to postpartum women only.
All postpartum women who had delivered a baby within the past 15 months (N = 264) were asked whether they had obtained postnatal care any time during the first six weeks after delivery. Women who had not obtained postnatal care were asked about the reasons for that. The question was open-ended and women were able to provide multiple reasons. In the latter part of the questionnaire women were also asked about their attitudes towards postnatal care: "In your opinion is postnatal care necessary for a woman's health?"
The questionnaire included information about several socio-demographic variables: woman's current employment status (unemployed or employed), woman's and husband's highest level of education (< secondary or ≥ secondary school), woman's age and age at first marriage, total number of living children, and woman's self-assessed economic situation (in three categories: high, middle, low). Education of both partners was recoded into a single variable with four categories: both < secondary, wife ≥ secondary + husband < secondary, wife < secondary + husband ≥ secondary, both ≥ secondary school.
Several additional variables related to medical care were collected: delivery place, having had problems during last delivery, number of antenatal visits during the last pregnancy, and whether the woman was informed about danger signs to be monitored after delivery related to her and her baby's health before discharge from the hospital. Women who had received postnatal care were asked whether they had received advices on family planning and breast feeding.
SPSS 12 statistical software was used to enter and analyze the data. Cross tabulation and Pearson chi-square test were used for descriptive and bivariate analyses. For the multivariable regression model the following variables were selected: location of the clinic, woman's age (in years), joint education variable, number of living children (to assess the extent of previous experiences with childbearing), number of antenatal visits during the most recent pregnancy (to assess general use of medical care), woman's age at first marriage (below or above 20 years, as a marker of more or less traditional upbringing), problems during delivery (since women who experienced problems may be more likely to receive postnatal care), delivery place (public versus private hospital, which may indicate different information received at discharge) and whether the woman was informed about danger signs for her and her baby's health before discharge. We did not include women's attitude regarding the importance of postnatal care, because attitude might be influenced by the actual behavior [29, 30]. To limit the number of variables in the regression analysis, we included education only as a proxy for the socio-economic status of the women. Furthermore, we did not include receipt of information about danger signs related to the mother's or the baby's health before discharge from the hospital, because these two variables were significantly associated with place of delivery: Women who delivered at a private hospital were more likely to be informed. Reasons for not obtaining postnatal care were analyzed by grouping similar answers.
A steering committee comprised of representatives from the Palestine Ministry of Health, the United Nations Relief and Works Agency for Palestine Refugees in the Near East (UNRWA), and Juzoor Foundation, one of the largest local health non-governmental organizations, reviewed and approved the study protocol and questionnaire prior to data collection. Permission for data collection at each site was given by the Ministry of Health. Prior to the interview, each woman was asked to read and sign a consent form, which stated the purpose of the study, that participation was voluntary, and that women's responses were kept confidential.