Male involvement in the maternal health care system: implication towards decreasing the high burden of maternal mortality

A cross-sectional descriptive community survey was conducted in North Gondar Zone, Northwest Ethiopia from May 3, 2016 to July 5, 2016. This zone has 19 districts; the zone has a total population of 2,903,165, of this 14.1% being urban residents.

The sample size was calculated by using single population proportion formula by taking 5% desired precision, 95% confidence interval, assuming 42% men awareness of danger signs [21], design effect of 2 for cluster sampling and 10% non response rate, which results in sample size of 710. A multi stage cluster sampling procedure was employed to select four districts out of 19 districts of North Gondar zone. Firstly four districts, three rural districts and one urban district were randomly selected. All households within the catchment areas having men whose wife had given birth within the last 2 years were invited to interview. A total of 856 men were invited to participate in the study.

A self administered questionnaire was adopted for our setting from Johns Hopkins Program for International Education in Gynecology and Obstetrics (JHPIEGO) [13]. The questionnaire has three sections: The first section is regarding socio demographic characteristics of the participants. The second section assesses participant’s knowledge of obstetric danger signs during pregnancy, labor and post partum periods. The final section assesses birth preparedness and complication readiness among participants during the birth of their last baby. The data collection process had taken 2 months (from May 3, 2016 to July 5, 2016). Data was collected by five properly trained research assistants with previous experiences in survey data collection. So as to maintain the quality of the data the researcher reviewed collected data daily and sent feedbacks to the data collectors continuously. BP/CR practice was considered “well prepared” for BP/CR when at least three of the following six practices were reported to have been made in the birth of the last baby: prepared birth kit, identified a skilled attendant, saved money, knew where to go in case of emergency, contacted a blood donor in advance and prepared transportation in advance. Men who reported less than half of the above stated steps were considered to be less prepared. If at least three steps were reported as having been made, the respondent was labeled as well prepared. The three out six was chosen because previous studies have used 50% and above to determine who was well prepared [14]. The questionnaire, originally written in English, was translated to local language (Amharic) and back to English in order to ensure that the translated version gives the proper meaning. The content validity of the questionnaire was confirmed by local experts, including reproductive health experts. (Additional file 1) The questionnaire was pretested on 5% of the sample size prior to the real data collection, which was excluded from the final study. Considering the high burden of puerperal sepsis in this community, which is common in the first 10 days of postpartum period and accounts for 13% of maternal deaths in Ethiopia [15], The time frame of first 2 days in the questionnaire used by Johns Hopkins Program for International Education in Gynecology and Obstetrics (JHPIEGO) was amended to cover maternal health problems in the first 10 days of the post partum period. Once data were collected, each questionnaire was checked for completeness.

The final data collection tool was checked for completeness, and responses were entered into and analyzed by the Statistical Package for the Social Sciences software Version 21.0 for Windows. Participants’ socio-demographic characteristics, knowledge of obstetric danger signs, and level of BP/CR were described using frequencies and percentages in tables. Bivariate logistic regression was conducted to explore factors associated with BP/CR and factors that were found to have a p value of less than 0.2 were entered in to multivariable logistic regression. To declare significant association P-value 0.05 was used as a cut-off point and results expressed by using odds ratio (OR) with 95% confidence interval (CI).

This study was approved by the Institutional review board of University of Gondar. Informed written consent was obtained from each participant before starting the interview and participants were also told their right to stop the interview at anytime. The participants were also told that participation is based on their willingness. The obtained information was kept anonymous and recorded in such a way that the respondent could never be known.

From 856 men who were invited for participation, 824 men agreed to the interview giving a response rate of 96.2%. A total of 721(87.5%) of the participants were found to be married and half of them have 3–4 children 403(48.9%). One third of the respondents 288(34.9%) did not go to school or did not complete their primary education. Participants from the rural areas accounts for the majority of respondents 637(77.3%). Participants who had escorted their wives to ANC follow up in previous pregnancy accounted for 334(40.5%) of the total participants and a quarter 201(24.4%) of the participants’ spouses had delivered in a health facility. A total of 118(14.3%) participants’ spouses had obstetric complication in their last pregnancy. (Table 1).

Half of the participants had stated at least one danger sign during pregnancy 407(49.4%); 271(32.9%) of the participants stated at least one danger sign during delivery 271(32.9%) and fewer, 213(25.8%), of the participants stated at least one danger sign during the postpartum period. The commonest mentioned danger sign during pregnancy was high fever 105(12.7%) and excessive vaginal bleeding accounts for the majority of responses during childbirth as well as during the post partum period. 85(10.3%) participants mentioned prolonged labor as a danger sign. There were also other signs mentioned which were not stated in the table such as anemia, dizziness, palpitation that accounted for 10.1% during pregnancy, 3.2% during delivery and 2.7% during postpartum period. (Table 2).

As illustrated in table three, the most common preparations made during the birth of their last child were getting birth kit 309(37.5%) followed by saving money 218(26.5%). (Table 3) Furthermore, 91(11%) of the participants had arranged transportation, 25(3%) had identified which health facility to visit in case of emergency, 67(8.1%) had identified a skilled birth attendant, 3(0.4%) prepared possible blood donor during the birth of their last baby. Two hundred and fifty six men (31.1%) had not done any preparations; 363(44.1%) had made one of the six birth preparing steps assessed; 116(14.1%) made two preparations; 82(9.9%) made three preparations; 5(0.6%) made four preparations; 2(0.24%) had made five preparations; furthermore, no one made all the six preparations assessed in the current study. (Fig. 1).

Place of residency, current marital status, escorting wife to antenatal care in previous pregnancy, number of children, level of education, knowledge of at least one danger sign during pregnancy, knowledge of at least one danger sign during delivery and knowledge of at least one danger sign during the postpartum period were associated with BP/CR in the bivariate logistic regression analysis. Further analysis by using multivariable logistic regression demonstrates that completing college education have a significant positive association with the steps taken for birth preparation and complication readiness (AOR = 3.2, 95% CI: 2.8, 4.6). Being married, escorting wife to antenatal care, and urban residence were also found to be significantly associated with the steps taken for birth preparation and complication readiness. Furthermore, the steps made towards birth preparation and complication readiness were found to have significant associations with knowledge of at least one danger sign during pregnancy (AOR = 3.3, 95% CI: 3.1, 3.9), during delivery (AOR = 2.2, 95% CI: 1.1, 2.8) and post partum period (AOR = 1.8, 95% CI: 1.1, 2.4). (Table 4).

As this is a cross sectional study, it is not possible to provide a causal relationship between being well-prepared and other variables. Moreover, the participants were required to remember things happened in the past 2 years so there is a possibility of recall bias. Although, all the necessary precautions were followed to maintain the quality of the data, it is not possible to avoid the effect of data collector’s attitude on the quality of the data. These limitations mean that caution is required when generalizing the results to other settings. Despite these limitations our study demonstrates a lot of strengths. One of the strengths of this study was the large number of participate in the study and men’s knowledge of danger signs and BP/CR were not assessed in this area previously.