Sociodemographic characteristics
A total of 1281 (response rate 95.8%) reproductive aged women (15–49 years) participated in the household survey. As Table 2 shows, the majority of respondents were married 1208(94.6%), Orthodox Christian followers 1230(96.0%), and Amhara in ethnicity, 1249(97.5%).
Over three-quarters, 978(76.3%) of the women were rural residents and almost similar proportion, 993(77.5%) were multi-parous. A high number of grand multiparty, 332(25.9%) was also noted. The age distribution of the participants showed that more than half 674(52.6%) of them were between 25–34 years, and the mean (+SD) age was 30.3(+6.0) years.
With regard to the distribution of the respondents’ educational status, more than half of the women, 660 (51.5 %) had never been to school, while only a far smaller proportion, 50 (3.9 %) of them attained tertiary education. The results regarding the distribution of the women’s household wealth index indicated that almost half, 631(49.3%) of them belonged to the lower two wealth quintiles (Table1, annex1).
Descriptive presentation of key elements of the continuum of maternal care
Antenatal Care (ANC) follow up
While 898(70.1%) (95% CI: 67.5–72.6) of the women received antenatal services from skilled health professionals (doctor, health officer, nurse, or midwife in Ethiopian context) at least once, 39.9% (95% CI: 37.2–42.6) of them continued for the WHO recommended four or more ANC visits during their most recent pregnancies. Of the mothers who had at least one ANC visit, only134 (14.9%) of them had their first ANC visit during their first trimester, while 615 (68.5%) of them during the second trimester, and the rest 149(16.6%) started during the third trimester.
Furthermore, the findings showed that most of the women did not receive the key ANC service components recommended by the WHO; and only 418 (46.2%) of them received all items of the antenatal care contents. For instance, of the nine key ANC components considered in this study, a smaller percentage of women had urine 519(57.8%) and blood samples taken 556(61.9%). On the other hand, the coverage of the blood pressure measurement at least once was much better than other antenatal service,857 (95.4%).
Skilled Birth attendant (SBA) use
Over all, less than less than half, 483(47.5%) of the most recent births were assisted by skilled birth attendants either at hospitals or health centers. For those mothers who gave birth outside the health institutions sudden onset of labor 395(58.8%) and lack of transportation for getting to health facilities 323(48.1%) were the main barriers mentioned for not seeking care. This is further confirmed by the fact that only 175(28.5%) of the respondents got an ambulance service to travel to the health facilities when labor started.
The finding further revealed that 178(13.9%) of the women were encountering at least one complication during or immediately after delivery. In this aspect, severe vaginal bleeding 30(16.9%) and prolonged labor 109(61.2%) were the most frequently mentioned problems.
Postnatal care use
For the postnatal care, despite 562(43.0%) of the respondents received at least one PNC within 6 weeks after-delivery, only a third, 192 (14.9%) of the mothers reported a health check within the first 48 hours of birth. Women who reported post-partum care were asked about the content of care they received; and the most frequently received service was counseling on breastfeeding, reported by 467( 83.1%). On the contrary, counseling about follow ups 154 (27.4%) and postpartum family planning 221(39.3%) were among the least frequently received services.
Pre-pregnancy and Postpartum family planning use
Contraceptive uses, both pre-pregnancy and postpartum, were more commonly reported than any other components of the continuum of care. Of the respondents, 942(73.5%) reported that they were using modern contraceptive before the index pregnancy. The proportion of women who received modern contraceptive after delivery was 762(59.5%). The contraceptive method mix among postpartum women is dominated by injectable 545 (71.5%) followed by implants, 164(21.5%).
For women who failed to use any modern contraceptive after delivery issues related to postpartum amenorrhea 137(26.4%), fear of side effects of hormonal contraceptives 97(18.7%), and infrequent sexual intercourse 119(22.9%) were the topmost reasons cited for non-use of modern contraception.
Retention on the continuum of the maternal care pathway
Figure 1 illustrates the flow of services within the continuum of care to indicate the proportions of women who transit from one maternal service to the next, and the points along the continuum where women drop off from the journey.
As could be seen in the figure 1, 70.1%(67.5-72.6%) of the women received ANC services at least once in their pregnancy but a substantial of them( 30.2%) did not continue on the pathway to receive 4 or more ANC visits, which was the highest drop-off in the continuum. As a result, only 39.9% (37.2-42.6%) made four or more visits. Across the continuum, the smallest relative drop-off (8.8%) was observed between ANC 4+ and facility delivery; 31.1% (28.5-33.7%) of women who received at least four ANC continued giving birth in health care facilities. Furthermore, after facility delivery about one in five (19.0%) did not go on to receive a PNC health check within 48 hours of birth (Figure 1).
The percentages of women who received the various possible combinations of maternal health services within the continuum of care is shown table 2; and this combination helps to point out the closely aligned elements of the continuum of care.
The finding indicated that 321 (25.1%) of women interviewed reported not having any contact at any points of the continuum during their pregnancy, and only 155 (12.1%) accomplished all stages of the continuum of care (4+ ANC visits, SBA, and PNC visit within 48 hours). The result also highlighted that only a few (or none at all) women attended facility delivery or postnatal care within 48 hours or both without first having received focused antenatal care, showing positive linkages among different maternal services. Accordingly, facility delivery without having received four or more antenatal care was uncommon at less than one percent. Furthermore, postnatal care was uncommon unless it was combined delivery with an SBA and at least four ANC visits (Table 2).
Multilevel analysis
As stated in the methods section, three multilevel logit regression models were fitted to identify factors influencing utilization of maternal health services along the continuum of care.
The first step in the multilevel model analysis was to consider if our data justified the decision to assess random effects at kebele (cluster) level. In this regard, the result of the empty model revealed that there was a considerable amount of variation for each indicator of maternal health service utilization across the clusters (P < .05), which justifies developing a multilevel model.
As shown by the intra-community correlation coefficient (ICC) values, 24.9%, 20.1%, and 21.2% of the total variances in the use of at least four antenatal care (ANC4+), facility delivery, and postnatal care within 48 hours were attributable to unobserved differences across clusters, respectively (Table 3). When adjusted for all the study covariates, the kebele(Cluster) level variances in the use of ANC 4+, skilled attendance at delivery, and retention in PNC reduced to 22.4%, 10.1% and 11.8%, respectively (p>0.05).
As shown in table 4, the full model showed that there are some variations in the predictors of the specific indicators of maternal service use; hence, we present the findings separately for each maternal indicators (Table 4).
Predictors of ANC 4+ visits
Model I analyzed the predictors of four or more antenatal care visits by women who took at least one antenatal care service. Results of the fully adjusted model showed that women’s age at the last birth, their education, and their knowledge on pregnancy-related complications were relatively the strongest level one predictors. The odds of retention in the antenatal care were higher among mothers aged between 18-24 years (AOR = 6.15, 95% CI = 2.41–16.48); women attended at least secondary education (AOR = 3.02, 95% CI = 1.16–7.83); and those who knew at least two pregnancy danger signs of pregnancy (AOR=3.31; 95%CI; 2.02, 5.41). Moreover, mothers who belonged to the richest wealth tertile (AOR = 2.21, 95% CI = 1.12–4.34) and those who had intended pregnancy (AOR = 2.24, 95% CI = 1.13–4.43) were other level one predictors of this model.
Of level two covariates considered in this study, the highest readiness score of the nearby health care facility (in providing antenatal care) had a positive association with the use of four or more antenatal follow-up (AOR=2.21, 95%CI; 1.12, 4.34) than women residing near to facilities of low readiness score.
The ICC (rho) in the use of at least four antenatal care show a marginal reduction from 24.9% (p< 0.05) in the empty model to 22.4% (still appreciably large) in the full model. In other words, the proportion of reduction in variance at the kebele level due to the covariates is only 12.8% ((1.092- 0.952)/1.092), indicating continued clustering of antenatal care utilization even after controlling for both level 1 and level 2 covariates(Table 4).
Retention in skilled birth attendance (ANC4 &SBA)
Model II analyzed the factors associated with the continuation of care from pregnancy to having skilled birth attendance among women who received at least four ANC visits.
In the fully adjusted model, the odds of retention in SBA were higher among ANC clients whose pregnancy was intended (AOR= 6.28, 95% CI 1.56, 25.31); who experienced any kind of pregnancy-related complications (AOR= 2.73, 95% CI; 1.44, 5.19); and those who had better knowledge of danger signs (AOR = 9.71, 95% CI; 4.56, 20.68) than their counterparts. In addition, women who had attended secondary and above education (AOR = 5.93, 95% CI = 1.92–18.34) and partnered with at least primary education (AOR = 1.63, 95% CI = 1.01–2.64) had a positive association with retention in skilled birth attendance. On the other hand, the age of the women showed an inverse relation with safe delivery; the odds of retention in SBA were higher among ANC clients aged between 15–24 years (AOR = 6.96, 95% CI; 3.33, 14.54) than 35 years and above.
As a level 2 predictor, mothers who lived in urban areas (AOR = 4.08, 95% CI = 2.63–6.31) were found to be positively associated with receiving skilled attendance at birth. Similarly, women’s odds of giving birth to health facilities was 3.07 (AOR= 3.07, 95% CI; 1.27, 7.41) times higher for women resided near to health facilities that were more ready to deliver antenatal interventions than the women around less ready facilities to antenatal service (Table 4).
Retention in PNC within 48 hours after birth
Model III estimates the effects of predictors on the continuation of care from delivery to early post-delivery period among women who received both at least four antenatal care and skilled birth attendance (i.e. completion of the continuum of care).
Retention in the entire continuum was higher among mothers who initiated ANC within the first 16 weeks (AOR = 7.53, 95% CI; 2.94, 19.29) than those booked lately. In addition, women who received the recommended contents during ANC consultation, which are indications of ANC quality, were virtually three folds [AOR= 3.31, 95% CI; 1.08,10.16) more likely to complete the continuum of maternal care as compared to women received poor quality of antenatal care.
The birth history of mothers (such as abortion, stillbirth, LBW) and mode of delivery were also important predictors in influencing the utilization of maternal health services. Women who had a history of poor fetal outcome before the last birth were 70 percent (AOR= 0.30, 95%CI; 0.12, 0.79) less likely to use maternal services compared to their counterparts who had normal pregnancy outcomes. Similarly, the odds of retention in early PNC attendance decreases by 64% for women who delivered spontaneously than assisted deliveries, (AOR=0.36, 95%CI; 0.20, 0.68).
Women married to better-educated husbands were more likely to use facility-based delivery; the odds of reporting in the use of postnatal care among women whose husbands were attending secondary or higher education were about three (AOR=2.48, 95%CI;1.07,5.73) times higher than those married to non-educated husbands. Likewise, housewives were 3.07 times (AOR=3.07, 95%CI; 0.38, 24.55) more likely to use maternal services compared to professionally employed women.
Women who mentioned at least two pregnancy danger signs were 2.64 (AOR=2.64; 95%CI; 1.27, 5.49) times more likely to complete the continuum than their counterparts (Table 4).