Close to half (47.8%) of the respondents had 1 or 2 children, 38.7% had 3 to 4 children, and 13.5% had 5 or more children. Nearly 78% of the respondents had ANC check-ups during their recent pregnancy. Seventy percent of the respondents rated themselves as having better knowledge of obstetric complications.
Respondents were also asked about their participation in household decisions such as health care, large purchases, small purchases, and visits to relatives. More than 93% reported that they had participation in deciding on the concerns assessed in the study. Only 6.2% reported that they had limited autonomy to decide on those household concerns.
A greater proportion of women living in urban areas delivered at health facilities than their rural counterparts (69.1% vs. 42.2%). Compared with women who delivered at home, those who delivered at a health facility were more likely to have literate husbands (59.2% vs. 40.8%), were more likely to have ANC follow-ups during their recent pregnancy (54.5% vs. 45.5%), and had somewhat better knowledge about obstetric complications (53.6% vs 46.4%).
About 54% of respondents reported strong tie networks and 45.6% reported weaker tie networks. Among respondents embedded in weaker tie networks, 47.2% delivered at home and 52.8% delivered at a health facility. Of those respondents embedded in strong tie networks, 54.4% delivered at home and 45.6% delivered at a health facility. Women who delivered at home had stronger tie network scores than women who delivered at health facilities (54.4% vs. 45.6%). Nearly 60% of the respondents had a homogeneous network and 41% had a less homogeneous network. Among women embedded in homogeneous networks, 47.8% delivered at home and 52.2% delivered at a health facility.
Furthermore, 48.5% of women were embedded within low SBA endorsement networks and 51.5% were embedded within high SBA endorsement networks. Of those women who were embedded within low SBA endorsement networks, 71.4% delivered at home and 28.6% delivered at a health facility. Among respondents with high SBA endorsement network, 31.9% delivered at home and 68.1% delivered at a health facility. Thus, women who delivered at a health facility had larger SBA endorsement scores than those who delivered at home.
The influence of social networks on facility delivery
The main intent of the research was to understand whether the women’s social networks or their individual attributes were more important in predicting the use of SBAs during uncomplicated delivery of their recent-born children. The -2log likelihood in model I with the individual attributes added was (313.836). The -2log likelihood in model II after the network variables were added was (245.685), indicating a reduction in error associated with the inclusion of the network variables in predicting facility delivery service utilization. The chi-square analysis in model II showed a significant difference between the -2log likelihood ratio at model I and -2log likelihood ratio at model II (χ2 = 134.116, df = 15, N = 274, p < 0.001). The overall correct classification of respondents into home delivery and facility delivery improved from 69.3% model I (that included women’s individual attributes) to 77.4% model II (that included the individual attributes and the network variables).
The results indicate that there was a statistically significant improvement in predicting facility delivery use with the network variables i.e., the network variables were important predictors for distinguishing between respondents who delivered at home and those who delivered at a health facility after controlling for women’s individual attributes. Women’s social network variables better predicted facility delivery service utilization for their recent-born children than their individual attributes.
The odds of facility delivery was 1.29 (95% CI: 1.16–1.45) times higher for every one member increase in network size. The odds of women who were embedded within a homogeneous network members to deliver in a health facility was estimated to be 2.53 (95% CI: 1.26–5.06) times higher than women embedded within less homogeneous network members. Women who were embedded within high SBA endorsement networks were 7.97 (95% CI: 4.07–12.16) times more likely than women who were embedded within low SBA to deliver in a health facility.
As indicated in Table
3, women’s individual attributes (educational level, residence, and knowledge of obstetric complications) significantly predicted facility delivery. The odds of women who had high school and above education level to deliver at a health facility was 8.07 (95% CI: 1.36–17.64) times higher than for illiterate women. The odds of women who had good knowledge of obstetric complications to deliver in a health facility was 3.01 (95% CI: 1.46–6.18) times higher than women who had limited knowledge about the complications. The odds of urban women to deliver at a health facility was 3.32(95^CI: 1.37–8.05) times higher than their rural counterparts.
Table 3
Hierarchical Logistic Regression analysis of women’s individual attributes and their social network variables predicting facility delivery (N = 274)
Age of respondents | .92(.82–1.14) | .94(.86–1.03) |
Educational level |
No education® | 1 | 1 |
Primary (1–8) | .95(.29–3.12) | .54(.25–1.19) |
High school and above | 3.79 (2.31–23.04)* | 8.07(1.36–17.64)* |
Husband’s educational status |
Illiterate® | 1 | 1 |
Literate | 2.20(.62–5.05) | 1.70((.84–3.43) |
Residence |
Rural® | 1 | 1 |
Urban | 2.75(.86–8.77) | 3.32(1.37–8.05)** |
Parity |
1–2 children® | 1 | 1 |
3–4 children | .22(.04–1.09) | .85(.38–1.92) |
> = 5 children | 1.91(.08–4.36) | 2.35(.57–9.73) |
Use of ANC |
No® | 1 | 1 |
Yes | 4.33(.85–2.13) | 1.86(.83–4.17) |
Knowledge of obstetric complications |
Limited ® | 1 | 1 |
Better knowledge | 2.31(.70–7.63) | 3.01(1.46–6.18)** |
Household autonomy | .63(.21–1.92) | 1.47(.61–3.52) |
Network size | – | 1.29(1.16–1.45)*** |
Network tie strength |
Weak® | | 1 |
Strong | – | .73(.35–1.55) |
Network homogeneity |
Less homogeneous ® | | 1 |
Homogeneous | – | 2.53(1.26–5.06)** |
SBA endorsement |
Low ® | | 1 |
High | – | 7.97(4.07–12.16)*** |
Network neighborhood |
Non-neighbor® | | 1 |
Neighbor | – | 1.65(.85–3.36) |
Model G2 (−2log likelihood) | 313.836 | 245.685 |
Degree of freedom | 10 | 15 |
Changed Chi-square | 95.731*** | 134.116*** |
The results also revealed the relative importance of network variables in predicting facility delivery service utilization. Network size and SBA endorsement had the smallest p –values (p < 0.001) followed by the homogeneity of network members (p < .01).Thus, network size and SBA endorsement by network members were more important in predicting health facility delivery service utilization than other network variables.
Informants in the qualitative study described that social networks have great influence during pregnancy, labor, and delivery. Their social networks often give them emotional support and assisted them in household chores such as fetching water, grinding cereals, preparing ‘tella’, and shopping during pregnancy. During labor and delivery, members of their social networks were praying-mariam, mariam, to shorten the duration of labor and to make the delivery smooth.
Some informants shared that their social networks facilitated transport to the health facility for delivery. For example, Tiringo shared: “we are living away from both relatives. When labor started, I told my husband. He immediately called on a transport and we went to the health facility.” Tatu similarly reported:
It was my first child. I was new for labor pain. When I felt unusual pain, I told my husband. He told my mother what I felt. She came …. She guessed that it was labor. Of course, it was…. After a while, other relatives heard the event. There was praying. The labor was long. … My mother wanted me to deliver at home. However, my husband called an ambulance and we went to a health facility for delivery (Tatu, parity 1).
Some informants shared that their place of delivery for their recent-born children was suggested by their network members. Tatu delivered her child at a health facility though her mother preferred that she deliver at home:
It was my first birth. I stayed long on labor, about 12 h. People were praying all the day expecting that I would deliver soon. Around sunset, my husband proposed that I should go to a health facility. He expected there might be a complication. Mother asked, ‘why?’ She said, ‘I delivered all my children at home…. There is no need to go to health facility.’ Other relatives supported his idea…. Thus, I delivered at a health facility.
The informant’s narrative demonstrates that she was not the primary decision-maker about where to deliver her child. Some informants shared that they planned to deliver at a health facility because they had information about the importance of facility delivery from different sources. For example, Zinash stated:
I knew the importance of facility delivery even before pregnancy. Sometimes, when someone delivered recently, neighbors and relatives come together and chat about her delivery. In those occasions, we used to chat about what she said, what her husband said, and on the overall birthing process. I think I learned a lot from such discussions. During recent pregnancy, just a month or so before, my friend delivered at a health facility. I accompanied her to the health facility. I observed …. Even, she joked at me that the next will be my turn. When labor started, I went to health facility. My husband and a few other people accompanied me (Zinash, parity 1).
Zenebu and Shita both shared that they delivered at a health facility because they worried about the absence of network members who would assist them at home during delivery. Zenebu shared that:
We are living in new residence away from both relatives. Around the date of delivery, my husband was not at home. I was living with my sister. We did not have relatives in the new residence. I worried a lot. I was thinking who will assist me during the pain. I knew health facilities provide delivery services. During ANC follow-up, the HEWs advised me to deliver at a health facility. Thus, when labor started, I went to the health center for delivery (Zenebu, parity 2).
The informant’s narrative indicated that she believes that social networks (relatives and neighbors) are important to support/assist women during labor and delivery. Because of the absence of relatives and neighbors who could assist her during labor and delivery, she went to the health facility.
Some informants delivered at home because the majority of their network members preferred them to deliver at home. As Aster shared:
When labor started early in the morning, relatives and neighbors who heard the event came to our home. Everybody was praying-mariam, mariam, mariam. My mother became more concerned about the event. … she suggested that I should go to health facility for delivery. However, other people did not support her. They argued, it was not long time since labor was started. There was no sign of any difficulty. The doctors will do nothing in the normal condition; she will lie on the back and they will say push. It is St. Mary who will bring positive outcome. Let us pray to St. Mary together. In the afternoon, I delivered at home (Aster, parity 4).