A cross-sectional study was conducted between 15 July 2003 and 15 August 2003 among women residing in the north of Jordan. Considering the population size, geographic location, and the number of families, according to the National Department of Statistics, the study population consisted of 20 clusters. The study sampled 5 clusters out of these 20 (i.e. one fourth of the total number of clusters). The included clusters represented the whole 20 clusters of the Irbid Governorate, which are mainly semi-urban communities. The 5 selected clusters were chosen using systematic sampling technique. The proportion of ever-married women in the age group 15–49, as estimated by the Ministry of Health in 2000, was 55% . For a sample size of 344 married women who had at least one child aged between 6 months and 3 years, the calculated power exceeded 80%, with 95% confidence that the true proportion of mothers that reported full breastfeeding is within ± 10% of the expected 50% proportion. Sample size was calculated using EPI-Info, 2002.
It was calculated that 625 households were needed to interview 344 married women. Having assumed that in every two households with married women one would have a targeted child, the number of households to be visited was doubled to 1250. Using systematic sampling technique (every 5th cluster), primary sampling units of 5 clusters were selected. The households to be visited in each cluster were calculated using the "proportionate to size" method.
The data were collected by trained fourth year medical students from the Jordan University for Science and Technology. They were instructed on interview techniques and were familiar with the geography of these clusters. Oral consents were obtained from all participants and all approached women agreed to participate. The study was approved by the Scientific Research Board of the Jordan University of Science and Technology.
A structured questionnaire with 32 items was used. It covered demographic variables that included mother's age, mother's education, father's education, mother's employment status, total family income, family's size, mode of delivery, gender of last child, and history of neonatal hospitalization. Breastfeeding was divided into three types: full breastfeeding which included exclusive and almost exclusive breastfeeding for a duration of 6 months (exclusive breastfeeding when no other liquid or solid from any other source entered the infant's mouth, while almost exclusive breastfeeding when occasional tastes of other liquids, traditional foods, vitamins, medicines were allowed), mixed breastfeeding when infants received both artificial formula and breastfeeding, and exclusive bottlefeeding when infants received only artificial formula after 2 weeks of life. All mothers were asked questions regarding their main reasons for either breastfeeding or bottlefeeding.
Questions with five-point Likert rating scale, from strongly disagree to strongly agree were used to assess women's knowledge and attitude to breastfeeding. The original Likert rating scale was converted to 100% scale by multiplying the corresponding coefficients (i.e. 1, 2, 3, 4, and 5) by 20. A higher score indicated a higher participant's agreement with the item tested. Items testing knowledge included recommended breastfeeding duration, its benefit in decreasing the risk of acquiring diarrhea and its role in contraception. Questions that evaluated mothers' attitude included mother's comfort with breastfeeding, cost, effect on care of other family members and effect on marital relationship. Items that tested community's attitude toward breastfeeding included feeling shy of breastfeeding in public places, role of community nurses and medical staff in encouraging breastfeeding, duration of maternity leave and facilitation of breastfeeding at work places.
The Statistical Package for Social Sciences software (SPSS, version 11.5, Chicago. Inc) and Microsoft Office Excel 2003 were used for data processing and statistical analysis. Variables were described using frequency distribution for categorical variables and mean and standard deviation for continuous variables. The distribution of infant feeding practices by independent variables was tested using the Chi-square test. Factors associated with not practicing full breastfeeding were identified using univariate and multivariate logistic regression. Crude odds ratios and adjusted odds ratios and their 95% confidence intervals were reported. A p-value of less than 0.05 was considered significant.