Study area and period
The study was conducted in the west Shewa zone of Oromia regional state, Ethiopia, from November 2017 to the end of January 2018. West Shewa zone has 24 woredas and the woredas are sub-classified into urban and rural kebeles (the smallest administrative unit). According to the information obtained from the zonal health office in 2017/2018, the west Shewa zone has a total population of 2,058,676, of whom 1,028,501 are men and 1,030,175 women. Out of this, the total women in the reproductive age group were 447042. All reproductive age women in a west Shewa zone where source population and all reproductive age group women who are married, living in union, fecund pregnant women and who lives in the zones for more than six months were included.
Sample size and sampling procedures
The sample size was calculated with Epinfo version 7.1 stat calc for a cross-sectional study design using the assumption [Zα/2= 1.96, a margin of error 5% P= 28%; Women’s knowledge and associated factors in PCC (9), design effect of 2]. By adding a 10% non-response rate, the final sample size becomes 680. A multistage, stratified sampling procedure was employed. In the first stage, 8 woredas from the 24woredas in the zone were selected using a lottery method. In the second stage, one urban and one rural kebeles from each woreda were randomly selected. In the third stage, from those selected kebeles, households which reproductive-age women were live in were selected randomly from the sampling frame obtained from kebele health office and health extension workers. The sample size for each kebeles was determined proportionally to the number of women’s reproductive age groups within each kebeles. In the case of more than one eligible woman were encountered in the selected household, a lottery method was used to determine which woman would be interviewed.
Data collection tool, quality control and measurement
A structured, interview administered questionnaire was used to collect data. The questionnaire was prepared in English(Additionalfile 1) and translated into local language, Afan Oromo by the translator, and then translated back to English by a third person to check for consistency. The tool adapted from previous literature in different parts of the world and modified according to the local context (11-14 and 16). Eight nurses were recruited as data collectors and Assistant professors with a background of health professionals were hired as supervisors. In addition, the data collectors were trained for one day on the techniques of data collection and the purpose of the study for study participants before the start of data collection. Pre-test was done on 5% of the total study participant and necessary adjustment was made. The data was collected house to house using an interview questionare. Data completeness and consistency were checked, cleaned and compiled by the supervisors on a daily basis. Incomplete data were removed from the study.
Measurements
The knowledge level of the study participants was determined using a dichotomous scale. Eleven knowledge related items were used to assess women’s knowledge on PCC and the question was scored out of twenty points. With a 50% cut of point women’s knowledge was divided into two. Those participants who have scored 10-20 of correct responses to PCC knowledge questions were considered as having good knowledge while those who scored less than 10 of correct responses considered poor knowledge (11, 13).
The uptake of PCC was determined, if the women received PCC at least once types of intervention either advice or treatment, and lifestyle modification care (screened for any disease and get treatment, take folic acid, take the vaccine, get counseling, modify diet, cessation of alcohol, cessation of cigarette smoking, stop taking illegal drugs, free from, create healthy environment) before being pregnant (11, 14).
Data management and analysis
Data were entered into Epi-Data Version 3.1 and exported to SPSS version 22 for analysis. Factors were tested using the bivariable analysis, and p-value≤0.2 was a candidate for the multivariable logistic regression analysis. To descriptive statistics; frequencies and percentages were used. Binary logistic regression analysis to examine the crude association of predictors on the desire to use PCC and knowledge about PCC, then multiple logistic regressions to see the effect of predictors on the desire to use PCC and knowledge about PCC and Odds ratio, 95% CI and P-value 0.05 were used.