Factors associated with birth preparedness and complication readiness in Southern Ethiopia: a community based cross-sectional study

BACKGROUND: Birth preparedness and complication readiness (BP/CR) is a strategy to promote use of skilled maternal and neonatal care so that they can get timely skilled care, particularly during child birth. There is minimal evidence on the factors associated with BP/CR among pregnant women in Ethiopia. Hence, this study aimed to assess the factors influencing BP/CR among pregnant women in Southern Ethiopia for the purpose of improving utilization of skilled attendant at birth. The study was conducted in March 2015 in Arba Minch Zuria Woreda, one of the woredas in Gammo Goffa Zone, South Ethiopia. The woreda consists of 29 rural kebeles (smallest administrative unit in the Ethiopian government administration). According to the Woreda‘s health office report, the total population of the woreda projected for the year 2007 E.C (2014/2015 G.C) was 205,204. Pregnant women in the woreda were estimated to be about 7101. The Woreda has 6 health centers and 37 health posts. The Woreda‘s main town is Arba Minch which is the capital town of Gammo Goffa zone. It is located 505kms far to the South of Addis Ababa, capital city of Ethiopia. Farming is the predominant occupation of residents in the woreda. Low land areas of the woreda are the major producers of banana supplying the markets in big cities in Ethiopia. Two of the Rift Valley lakes (Abaya and Chamo) are also situated at the Eastern border of the Woreda. It is also one of the tourist destinations in Ethiopia having Nech Sar National Park on the plain at the back of a mountain bridging the two lakes, namely `Ye Egzier Dildiy’ meaning the bridge of God. The forty springs, a spring in the forest at the Eastern border of Arba Minch town is also one of the attraction areas in the woreda [25]. A community- based cross-sectional study was conducted among pregnant women with a self-reported pregnancy of 3 months and above who were randomly selected from 9 kebeles in the woreda. During data collection, women who were seriously sick or unable to give information and who lived in the study area for less than 6 months were excluded. For the quantitative survey, the sample size was calculated using single population proportion formula assuming 17% proportion of birth preparedness among pregnant women in the zone [15], 95% confidence, and a margin of error of 4%. The sample size calculated with this consideration was 339. After applying finite population correction formula and 10% non-response rate and design effect of 2, the final sample size was 713.For the qualitative data, a total of six focus group discussions (FGDs) were formed in groups of husbands of pregnant women, traditional birth attendants (TBAs) and health professionals. Each of the groups consisted 8 to 12 members. There are a total of 29 rural kebeles in Arba Minch Zuria Woreda [25]. Pregnant women in the woreda were estimated to be about 7101 which were 3.46% of the total population (3.46% is a conversion factor for estimated number of pregnancies for SNNPR in Ethiopia). Hence, the estimated number of pregnant women for each kebele was obtained by multiplying the total population of the kebele by 3.46%. Multi-stage sampling was used to select the study subjects. In the first stage, nine kebeles were selected from the 29 kebeles randomly. The sample was proportionally allocated to each kebele with consideration of the estimated number of pregnant women per each kebele. Then using family folder (a folder containing different health and health related information of a family) in the health posts of each kebele, pregnant women were listed in each kebele. From the list, the required number of pregnant women in each kebele was selected randomly by using IBM SPSS statistics 20. For the qualitative data, non-probability (purposive) sampling technique was used to obtain homogenous groups for each category (husbands, TBAs, and health professionals). A total of 7 public health nurses and 2 health officers were recruited for the quantitative data collection and supervision of the collection process, respectively. Both data collectors and supervisors were given a day long intensive training on the data collection methods. A pretested, structured, interviewer administered questionnaire was used. The questionnaire was partly adapted from the survey tools developed by JHPIEGO Maternal and Neonatal Health program [5] and household food insecurity status of women was assessed using the standard Household Food Insecurity Access Scale (HFIAS) questionnaire by Food And Nutrition Technical Assistance (FANTA III) [31]. During the data collection period, the data collectors and supervisors were guided by health development army (HDA) leaders in each kebele so that they can easily access the houses of each sampled pregnant woman. The data collectors were given the list of the pregnant women to be interviewed in their respective kebeles in advance. For the qualitative data, unstructured open ended and non-directive FGD guide was designed in order to triangulate responses obtained from the quantitative survey. The principal investigator moderated the discussion of health professionals, while the TBAs and the husband‘s group were moderated by an experienced health officer working in the woreda in the presence of note takers and technical assistants. A day long training and practical exercise was carried out before the commencement of the data collection. Group discussions with their respective discussants were conducted in a quiet kebele and health center halls. Each discussion was tape recorded not to miss all issues discussed. Data was entered in to Epidata version (3.1) and analyzed using IBM SPSS statistics 20. Descriptive statistics using frequencies, percentages, mean, and standard deviations was used to describe findings. Bivariate analysis using logistic regression was done and all explanatory variables which have association with the outcome variable at p- value of less than 0.25 were selected as candidates for multivariable analysis. Multi-collinearity between the candidate variables was checked. Then multivariable analysis using back ward stepwise selection method was done to control for possible confounding variable and to determine presence of statistically significant association between explanatory variables and the outcome variable. Level of statistical significance was declared at a p-value of <0.05 and OR with 95% CI was used to measure the degree of association between independent variables and the outcome variable. Model fitness was checked using Hosmer and Lemeshow goodness of fitness test. Hence, a total of 13 variables with a P-value less than 0.25 in the bivariate logistic regression analysis (age of women, educational status, marital status, occupation, ever listened to the radio, history of still birth, frequency of ANC, advice during ANC, place of delivery for the last birth, knowledge on key pregnancy, labor and delivery and post- partum danger signs, household food security status) were entered in to a multivariable logistic regression model. Factor analysis was conducted to set wealth /economic status of pregnant women. A total 13 items on household assets were analyzed using principal component analysis method after checking the fulfillment of assumptions using Kaiser-Meyer-Olkin measure of sampling adequacy and Bartlett’s Test of Sphericity. The qualitative data was analyzed manually by the principal investigator following the steps of frame work analysis [32].However, all authors read the material and reflected on each of the transcripts, codes, subcategories, categories and themes until agreement was reached. Prior to analyzing the data, all FGDs were transcribed in Gamotho and then translated into English by the first author and a research assistant (moderator of the husband’s and TBAs group). Preliminary coding of transcripts was done and consistent codes were condensed in to categories and subsequently in to themes. Finally, the categories and themes emerged from the data were identified and were presented together with powerful quotes in narrative ways by triangulating with quantitative data. Various activities were performed to assure the quality of data. The English version questionnaire was translated to the local language (Gammotho) by a person having knowledge of both of the languages. Then another individual who had very good knowledge of both English and Gammotho language translated the Gammotho version back to English to check for its original meaning. The questionnaire was pre-tested on 36 respondents (5% of sample size) in Chano Dorga kebele that had similar characteristics with the study population other than the sampled cluster in the study area. Findings were discussed among data collectors, supervisors and the investigator in order to ensure better understanding to the data collection process. Based on the pretest, questions were revised, edited, and those found to be unclear or confusing were modified. Finally, a structured Gammugna version questionnaire was used for data collection. The principal investigator and supervisors made supervision on the data collection process. Every day, 10% of the completed questionnaires were reviewed and checked for completeness and consistency by the supervisors and principal investigator and the necessary feedback offered to data collectors in the next morning before the data collection begins. Each woman was interviewed in a separate private place to avoid social desirability bias. For the qualitative data, the kebele administrators, health administrators, and the supervisors, with principal investigator identified eligible discussants which were supposed to be informative from each group. Kebele: refers to the smallest administrative unit in the Ethiopian context. Woreda: refers to an administrative unit above kebele. Family folder: It is a household centered service delivery tool or registration book that contains information on household characteristics and other health and health related information of members of the household which is available at the health posts. A woman was considered as prepared for birth and its complication if she reported that she saved money (planned to),identified health facility(planned to) for delivery (health center/hospital),, identified skilled provider at birth(planned to) and identified a means of transport to place of childbirth or for the time of obstetric emergencies(planned to) ahead of childbirth. Those mothers who reported that they have followed (planned to) at least three of the four BP/CR components were considered as ―well prepared for birth and its complication. The remaining women were considered as ―not well prepared for birth and its complications. A woman was considered knowledgeable on key danger signs of pregnancy, if she can mention at least two of the three key danger signs for pregnancy (vaginal bleeding, swollen hands/face and blurred vision) spontaneously [14]. A woman was considered knowledgeable on key danger signs of labor/childbirth, if she can mention at least three of the key four danger signs for labor/childbirth (severe vaginal bleeding, prolonged labor (>12 h), convulsion and retained placenta) spontaneously [14]. A woman was considered knowledgeable on key danger signs of postpartum, if she can mention at least two of the three key danger signs for postpartum (severe vaginal bleeding, foul-smelling vaginal discharge and high fever) spontaneously [14]. A mildly food insecure household worries about not having enough food sometimes or often, and/or is unable to eat preferred foods, and/or eats a more monotonous diet than desired and/or some foods considered undesirable, but only rarely [31]. A moderately food insecure household: eats a monotonous diet or undesirable foods sometimes or often, and/or has started to cut back on quantity by reducing the size of meals or number of meals, rarely or sometimes [31]. A severely food insecure household: Any household that experiences one of these three conditions(running out of food, going to bed hungry, or going a whole day and night without eating) even once in the last 4 weeks (30 days) [31] . A household was considered as food insecure when it has mild, moderate or severe food insecure conditions whereas a food secure household experiences none of the food insecurity conditions or just experiences worry, but rarely [31]. Ethical clearance was obtained from Jimma University Ethical Review Committee (Ref.No. PFHD/055/677/07). A formal letter of permission to conduct the study was obtained from Gammo Goffa zone health desk and subsequently from Arba Minch Zuria Woreda Health office (Ref.No.  + 5/2799/2). Verbal consent was obtained from the study subjects and they were informed that the data will be kept confidential.