Effects of health education on spousal knowledge and participation in birth preparedness in Farafenni Regional Hospital, The Gambia: a randomized trial

Background: The Gambia is a male-dominant society in which the cultural norms empower husbands to decide when and where their wives seek care, yet they are not always involved in maternal health care services. Therefore, the purpose of this study was to design and measure the effects of antenatal health education on spousal participation in birth preparedness in Farafenni and satellite villages. Methods: The study used a quasi-experimental design, and the participants were 300 spouses of pregnant women attending their antenatal care booking at Farafenni Hospital. A multistage sampling method was used to select the study participants who were then equally distributed to the intervention and comparison groups. Pre-test data were collected from both groups. Thereafter, the intervention group was exposed to two health education sessions on obstetric danger signs and birth preparedness. The post-test data were collected immediately before discharge of the participants’ wives after institutional delivery or within 2 weeks post-delivery for those who did not accompany their wives to the health care institution, or whose wives delivered at home. IBM SPSS version 21 software was used to analyze the data. Results: The differences between the demographic characteristics of participants in the intervention and comparison groups were not statistically significant except for the highest level of education achieved. After controlling for the demographic variables, the health education administered to the intervention group effectively increased knowledge on birth preparedness among them (F (1, 255) = 376.108, p <.001). Every unit increase in the intervention led to a unit increase in the spouses’ knowledge on birth preparedness (β = 0.789, p < 0.001). Furthermore, the participants in the intervention group had higher mean score (M = 4.4; SD = 0.8) on participation in birth preparedness than those in the comparison group (M = 0.9; SD = 0.8). The spouses in the intervention group were four times more likely to be prepared for the delivery of their wives after being exposed to the health education than those in the comparison group (F (1, 255) = 522.414, p <.001). Conclusion: The study provides evidence that educating men on maternal health care can improve their level of participation in birth preparedness. Trial registration: Name of Registry: Pan African Clinical Trial Registry (www.pactr.org). Registry Number:PACTR202004752273171. Date of Registration: 19th April 2020. Retrospectively Registered. This study used a randomized control trial with pre-test and post-test comparison groups design to evaluate the effect of a health education intervention programme on spousal participation in birth preparedness in Farafenni and satellite villages. This study design was guided by the action points identified during baseline research done among the target group [16]. The study design is simplified in Fig. 1. Quasi-experimental design with pre-test and post-test comparison groups The study population was spouses of pregnant women attending antenatal care at the Maternal and Child Clinic of Farafenni during the data collection period. The participants’ inclusion criteria for this study were as follow: The exclusion criteria were: This study was conducted at the Antenatal Clinic of the Farafenni Regional Hospital for easy identification and access to the target group. Farafenni is located in the North Bank East Region of The Gambia and it has both urban and rural characteristics, as well as diverse ethnic distribution. Therefore, the findings obtained from this study can give a picture of both rural and urban communities in The Gambia. The health education sessions were conducted in two separate rooms, i.e., the Antenatal Examination and the Weighing Rooms, as these rooms were not in use during the intervention days. This study was a clinic-based intervention. There was no antenatal care (ANC) on Fridays in Farafenni Antenatal Clinic, so these days were chosen as research days. The health education intervention and comparison placebo were conducted on alternate Fridays. This had helped to prevent the two groups from meeting in the clinic, thereby reducing the risk of sharing the interventional messages with the comparison group. Thirty (30) spouses were invited each Friday and were requested to come without their wives but with their antenatal cards. The duration of the intervention lasted for six (6) months. There was no prior antenatal health education programme for spouses of pregnant women at Farafenni. Therefore, the health education curriculum was designed by the researcher using the indicators of monitoring birth preparedness and complication readiness for maternal and newborn health developed by Jhpiego [12] and the research objectives as guides. The health education curriculum was divided into two modules. Module one included information on birth preparedness (including danger signs of pregnancy and childbirth) while module two covered normal signs of labour and institutional delivery care. Besides, detailed health education posters that reviewed the main messages of the modules were used as visual aids to enhance learning during the intervention sessions. Posters with danger signs of pregnancy and childbirth and birth preparedness indicators were adopted and used in the health education sessions. Physical samples of the materials needed for a clean delivery were also made available during the sessions. With the assistance of persons knowledgeable in the local culture, the health education modules and visual aids were designed to reflect the community’s characteristics and tradition. The health education sessions were conducted by two research assistants (male and female) and the researchers using Wollof, Mandinka, or Fula local languages depending on the participants’ preferences. The health education was delivered individually (one-to-one) to all the members of the intervention group. There were two health education sessions for each of the spouses in the intervention group as outlined below and summarized in Table 1. Health education on birth preparedness and institutional delivery for spouses in the intervention group One hundred and fifty (150) spouses came for the first health education sessions, and each received 25-min one-to-one health education on obstetric danger signs and birth preparedness. Posters with obstetric danger signs and birth preparedness indicators were also shown to the participants during the health talk. Counseling on blood donation was also conducted during the first health education contact with each spouse as blood donation is part of the birth preparedness process. The spouses who consented to donate blood were referred to donate blood to save for emergency during the pregnancy or delivery periods of their wives. Participants were made to understand that the blood donated may be used for other patients if their wives did not need it to avoid expiration, but the Blood Bank Unit would ensure that blood is made available to them when it is needed. The spouses were also informed of the expected dates of delivery of their wives so that they could make plans for it. After evaluating their comprehension and reviewing the main messages, each participant was given a typed list of all the materials required for a clean delivery and A-4 size paper photocopies of the posters with the obstetric danger signs and birth preparedness indicator to take home. Each spouse was also informed on the date of his second session, which was at 36 weeks gestation of his wife. Participants’ contact details and expected dates of delivery of their wives were recorded. These records were reviewed at the end of each intervention day and the participants whose follow-up dates (second health education sessions) were due in the next scheduled date for the intervention group were reminded through phone calls or through letters given to their wives during their normal ANC visits to deliver to them. A record of those who donated blood, and that of their wives were kept in the Blood Donation Unit. A total of 150 participants came for the second session and each was given another 25-min health education on signs of normal labour and the importance of institutional delivery care. These topics constituted module two of the health education package. These sessions started with reviewing the main messages of the first health education session. Thereafter, spouses were taught signs of normal labour to equip them with the required knowledge needed for them to make early decisions for their wives to seek skilled care when they were in labour. Inquiries were made regarding their level of preparation for the deliveries of their wives and blood donation. Those who reported that they did not donate blood were counseled again and referred to the blood donation unit. They were also informed that the post-test assessments would be conducted immediately after the deliveries of their wives. The health educators summarized the health information on each of the topics co2vered into talking points as follows: The treatment of the participants in the comparison group was similar to that of the intervention group except that they did not receive the health education messages from the intervention package but were given group counseling on nutrition during pregnancy in the first session and exclusive breastfeeding in the second (duration: 45 min each). To avoid confusion, this group of participants was informed before the pre-test that the questions that were asked in both the pre-test and post-test were not related to the topics that were discussed in the counseling sessions. The participants from this group were not counseled or sent for blood donation. They were informed that their post-test data would be collected immediately after the deliveries of their wives (Fig. 2). Study Intervention Two buy-in meetings regarding the aims and objectives of the study were undertaken with the Management of Farafenni Regional Hospital in September 2017 before the commencement of the study. Two types of training were also conducted for the staff of the antenatal clinic, labour ward, and blood bank before the starting of the study intervention. The whole site training was given to all the staff including support staff such as general assistants and clerks. The training aimed to familiarize staff about the project, address any concerns, and identify interested staff to serve as research assistants. This training lasted for 45 min. The second component of the training was more intensive and aimed at the research assistants only. Six Community Health Nurses and Midwives were selected as research assistants. Four of the research assistants were from the antenatal clinic whilst the remaining two were from the labour ward. They were given a 90-min refresher training on a variety of topics including obstetric danger signs, preparation for delivery, involving men in maternal care, basic teaching and counseling techniques, and on the structured intervention protocol. The primary outcomes were spousal knowledge of and participation in the birth preparedness of their wives. The Jhpiego prototype questionnaire for monitoring birth preparedness and complication readiness for safe motherhood [12] was adapted and modified to suit the research objectives and target group (Appendix A: Interview Guide). The questionnaire measured knowledge of birth preparedness by asking the participants questions about the danger signs of pregnancy and childbirth, and resources needed when preparing for childbirth. There were seven items under the sections measuring knowledge. A correct response for each item was scored 1 and the wrong response 0. The percentage of the total score for each participant was calculated. A total percentage score between 0 and 30 was defined as low, 31–61 was defined as moderate and 62–100 was defined as a high level of knowledge of birth preparedness. Spousal level of participation in birth preparedness was measured based on the number of arrangements a spouse had made, including (1) acquired the required materials for clean delivery (included, two clean gowns for mother, two clean pants for mother, a new packet of a pad for mother, two clean buckets with lids, surgical spirit for cleaning baby’s the umbilical stump and a clean wrapper for the baby), (2) saved money for the delivery, (3) donated blood to the blood bank for an emergency, (4) arranged for transportation to the delivery center, (5) made an emergency plan and (6) identified a health facility. Spouses were considered ‘highly’ prepared if more than 3 of these arrangements were reported, moderately prepared if 2–3, and if 0–1 the arrangement was defined as poorly prepared for the delivery of wife [12]. Data on the primary outcomes were collected using a research assistant-administered questionnaire method. The study questionnaire was translated from English to three of the major local languages of the community (Wollof, Mandinka, and Fulla), with the help of an expert in language translation. To ensure quality, before the data collection, six community health nurses and midwives fluent in the three major local languages were selected as research assistants and trained in the administration of the questionnaire. Each data collector was given a sheet containing the basic field protocol. There were two teams of data collectors; one was at the antenatal clinic (with four research assistants) and the other in the labour ward with two research assistants. In each team, one person from the data collectors served as a supervisor. The principal investigators monitored and supervised the overall study, to ensure that the research team adhered to the research procedures. The pre-test data were collected from the participants immediately after signing/thumb printing the voluntary informed consent forms. The same questionnaire was used to collect both pre-test and post-test data. The participants’ contact details, expected date of delivery of wife, and follow-up appointment dates were noted. The first follow-up of the participants was at 36 weeks of pregnancy of their wives during which no data was collected. The second follow-ups were used to collect the post-test data. These follow-ups were conducted in the labour and postnatal wards, immediately before discharge of the participants’ wives after delivery by the two research assistants working in the labour ward. The post-test data of those who did not accompany their wives to the health facility for delivery, or those whose wives delivered at home, were collected within two to 3 weeks after the delivery of their wives. To minimize loss to follow-up, participants who did not turn up after an invitation were contacted through phone calls or home visits. The completed forms from the field were reviewed daily and on-the-spot feedback was provided, with follow-up/callback undertaken, where needed. The data of pre-tested participants who did not participate in the post-test were not included in the analysis to eliminate attrition bias. The estimated sample size was 147 participants in each group (making a total of 294 participants). This was based on the assumption of detecting a 15% effect from the proportion of men accompanying their spouses to antenatal care 20% (13), considering the power of 80% with 5% significant level, a design effect of 1.5, and a non-response rate of 10%. However, this was increased to 150 participants in each group, making a total of 300 participants in this study. A multistage sampling method was employed to select a representative sample as shown: Farafenni Regional Hospital’s Maternal and Child Health Care Catchment Area was divided into rural and urban areas. The rural area was further divided into 12 villages, namely, Macca Farafenni, Yallal Ba, Dutabullul, Gigimarr, Jerri Kaw, Sagab, Kerr Sulay, Kunjo, Elliyasa, Kerr Madi, Kohen Bereto, and Sabach Nyen. The names of these 12 villages were written on pieces of paper, folded, put in a container which was closed and shaken several times to ensure a good mix or randomization of the pieces of paper. Six of these pieces of paper with the names of different villages were randomly withdrawn from the container using the balloting by replacement method. These six selected villages were Macca Farafenni, Dutabullul, Yallal Ba, Kunjo, Kohen Bereto, and Sabach Nyen. The urban area, which is Farafenni, was divided into four wards and two of them, namely, Farafenni Mauritani, and Farafenni Ballangharr, were selected using random sampling as in the selection of the villages. Selected villages and wards were divided into intervention and comparison groups’ residential areas using random sampling with replacement method to ensure that there were an urban resident and three villages for each of the study groups. This was done to ensure that the participants in the intervention and comparison groups came from the similar environment and had similar characteristics. To ensure that the intervention and comparison groups’ residents were not too near to each other (to reduce the risk of social contacts between the members of the two groups), the balloting was done in such a way that there was always a non-study resident between an intervention and a comparison groups’ residential areas. Farafenni Mauritani, Yalla Ba, Kunjo, and Sabach Nyen were the intervention group’s residents whilst Farafenn Ballangharr, Macca Farafenni, Dutabullul, and Kohen Bereto were that of the comparison group. The recruitment period started from 1st October 2017 to 31st January 2018. Women from the selected six villages and Farafenni town wards meeting the gestational age criteria (they were sent for a pelvic ultrasound to confirm their gestational ages) were identified and informed about the purpose of the study during antenatal booking. The verbal consent of these women to share their gestational age information with their husbands and to allow them to participate in their obstetric care was sought before sending participation invitation letters to their husbands. This gave the women freedom to choose whether they wanted their husbands to participate in their health care or not. Five hundred and four (504) spouses of the women were assessed for the eligibility criteria but 61 of them were excluded (29 did not meet the inclusion criteria and 32 of the women did not agree with their spouses to participate). The remaining 443 spouses were sent invitation letters but 26 of them rejected the invitation and were replaced randomly. The participants’ enrolment was conducted at the antenatal clinic of Farafenni Hospital. Three hundred (300) of them were randomly selected as the study participants using the balloting with replacement method. The selected spouses from the villages and Farafenni ward identified as the intervention group’s residents were enrolled as the intervention group, while those from the comparison’s residents were the comparison group. All the selected spouses attended the first sessions but 21 (12 from comparison and 9 from intervention groups) of them did not return for the second sessions and were randomly replaced until the required sample size was achieved in both groups (see Fig. 3 for detail). The selected participants were given a detailed description of the activities, benefits, and risks involved in participating in the study. Enrollment Process The Principal Investigator was responsible for the randomization but the research assistants enrolled the study participants and assigned them to the intervention and comparison groups. It was only the study participants who were blinded. Blue stickers with the study identification number of the participants were pasted on the antenatal cards of the wives of spouses in the comparison, whilst green coloured stickers were for those in the intervention group. However, they were not informed of the group that they belong and participants from the intervention and comparison groups were not living together, that is, they came from different villages and Farafenni town wards. In addition, participants of the intervention and comparison groups were invited to participate in the study on different days so that they do not meet to prevent contamination of the intervention. All the completed questionnaires were reviewed for completeness and incompletely filled questionnaires either of the pre-test, post-test, or both were excluded from the study. The data from the completed questionnaires were then entered into an Excel spreadsheet and scanned for errors before exporting to Statistical Package for Social Sciences (SPSS). The data was cleaned by running frequencies of all variables to check for incorrect coding. After double-checking with raw data, needed corrections were made before the analysis. IBM Statistical Package for Social Sciences (SPSS) version 21 software was used for data analysis. Descriptive statistics were used to summarize the socio-demographic and outcome variables. The differences in the socio-demographic variables between those in the intervention and comparison groups were examined using either independent-samples t-test (for continuous variables), or chi-square (for nominal/grouped variables). The proportion of participants that gave a correct answer to each item on the obstetric danger signs and birth preparedness indicators during the pre-and post-tests were calculated by representing the sum of the correct answers as a percentage of the total. The mean knowledge score differences between the pre-tests and post-tests of both groups (i.e., between-groups differences) were tested using the independent sample t-test and at a statistical significance level of p <  0.05). Spousal level of participation in birth preparedness was analyzed by assigning a score of 1 to each indicator of the birth preparedness items that each participant reported participating in or 0 if he did not participate. The percentage of participants that reported participating in each item during the pre- and post-tests was calculated. The percentage differences in performing each item between the pre-tests and post-tests of both groups (i.e., between-groups differences) were tested using Chi-square and at a statistical significance level of p <  0.05). The effect of the health education intervention on spousal knowledge and participation in birth preparedness was tested using analyses of covariance (ANCOVA) and hierarchical linear regression (to control for the demographic variables). Due to the clustering in the multistage sampling used in the study, analysis of covariance (ANCOVA) was used to adjust for any preexisting between-group differences not controlled for at the study’s onset. ANCOVA, using the pre-test mean scores as a covariate, would help to determine whether the adjusted post-test mean scores between the two sample groupings were significantly different from another (F-value; confidence level of p <  0.05). Levene’s test was used to test the ANCOVA assumption of homogeneity of variance of the dependent variable. The categorical independent variables (health education and demographic variables) used as predictors of the outcome variables were transformed to dummy variables before conducting the regression analysis. The assumptions of hierarchal linear regression analysis of linearity between independent and outcome variables, normal distribution of outcome variables, and multicollinearity between the independent variables were checked and met. Principal Component Analysis was conducted to check for variability between the independent variables. Screeplot was developed and Kaiser Criterion was used to select the independent variables with Eigenvalue > 1 as predictors of the outcome variable in the hierarchical regression analysis. The statistical significance level was set at p <  0.05.