Predictors of maternal and newborn health service utilization across the continuum of care in Ethiopia: A multilevel analysis.

BACKGROUND: The continuum of care for maternal and newborn health is a systematic approach for delivery of an integrated effective package of life-saving interventions throughout pregnancy, childbirth, and postpartum as well as across levels of service delivery to women and newborns. Nonetheless, in low-income countries, coverage of these interventions across the life cycle continuum is low. This study examined the predictors of utilization of maternal and newborn health care services along the continuum of care in Ethiopia.
METHODS: This was a cross-sectional population-based study. We measured maternal and newborn health care utilization practices among women who had live births in the last 12 months preceding the survey in Amhara, Oromia, SNNP, and Tigray regions of Ethiopia. We fitted multilevel random-effects logistic regression models to examine the predictors of the continuum of care accounting for the survey design, and individual, and contextual characteristics of the respondents.
RESULTS: Our analysis revealed that only one-fifth of women utilized maternal and newborn health services across the antepartum, intrapartum, and postpartum continuum; most women discontinued at the postpartum stage. Continued use of services varied significantly across wealth, model family, prenatal stay at maternity waiting homes, antenatal care in the first trimester, complete antenatal care service, and the administrative region at all antepartum, intrapartum, and postpartum stages. Moreover, family conversation during pregnancy [AOR: 2.12; 95% CI: 1.56-2.88], delivery by cesarean [AOR: 2.70; 95% CI: 1.82-4.02] and birth notified to health extension workers [AOR: 1.95; 95% CI: 1.56-2.43] were found to be predictors of the continuum of care at the postpartum stage.
CONCLUSION: In Ethiopia, despite good access to antepartum care, compliance with continuity of care across the pathway decreased with significant inequitable distributions, the poorest segment of the population being at most disadvantage. The main modifiable program factors connected to the continued uptake of maternal health services include family conversation, pregnant women conference, complete antenatal care, antenatal care in the first trimester, and birth notification.

Background

The continuum of maternal and newborn care is highlighted as a framework for delivering high-impact interventions across the pregnancy, childbirth, and postpartum life-course and the level of service delivery from community to the hospital to prevent maternal and newborn mortalities [1]. A recent systematic review also shows that the continuity of care from antepartum to postpartum periods may reduce the risk of combined neonatal, perinatal, and maternal mortality by 15% [2] and reduce neonatal and perinatal mortality risk by 21% and 16%, respectively [3].

In low-income countries, coverage of care is lowest during childbirth and postnatal period, and services are often fragmented further limiting the continuity of care [4-6]. The quality of antepartum and intrapartum care also influence women’s healthcare-seeking decisions [7, 8]. Uptake of services can drastically decline from the antenatal to the postnatal period, along with the continuum of care (CoC), due to quality concerns [9, 10]. A major drop-out occurs early in the CoC; women who did not complete antenatal care (ANC) may not receive skilled delivery or postnatal care (PNC) services [11].

Uptake of maternal and newborn health services across the CoC is affected by factors such as those related to the women themselves and their households [12]. Individual-and household-level factors include marital status, maternal education, partner’s education, wealth, women’s autonomy in decision making, awareness regarding PNC, and neonate being ill [11, 13–18]. Additionally, availability of transportation, living in an area far from the health facility, content of antepartum care like urine tests, and delivery at a health facility were identified as predictors significantly associated with CoC [13].

Understanding factors that contribute to creating gaps across service use during antenatal, delivery, and postnatal care is imperative for the successful continuity of maternal and newborn health services and eventually for improvements in maternal and newborn health outcomes [19]. However, evidence on factors associated with adherence to the continuum of maternal and newborn health care is not widely available. Nationally, there is a critical knowledge gap regarding the characteristics of a continuum of maternal and newborn health services [20]. Why the utilization of maternal and newborn health services has not improved in Ethiopia despite the huge national efforts and investments made in the last decade is not well known. The main purpose of this study was to identify factors associated with the use of maternal and newborn health care along the pathway of CoC in Ethiopia.

Methods

Context

Administratively, Ethiopia is divided into 12 geographic regions where regions are divided into zones, which are internally divided into woredas (i.e., districts) and each woreda into the smallest administrative unit called kebeles.

In Ethiopia packages of maternal and newborn health interventions are delivered through home-based, community-based, and facility-based service delivery modalities. The country’s health system has primary level care (encompassing a primary hospital, health centers, and health posts), secondary level care, and tertiary level care [21]. The primary health care provides preventive and promotive community and outreach services through the expansion of the Health Extension Program (HEP), the national flagship community-based health care delivery system, and the engagement of community volunteers [21, 22].

The country developed different strategies and programs to improve maternal survival which include strengthening and promoting skilled delivery through community mobilization [23]. The government has also provided ambulances to districts to mitigate transportation barriers, trained and deploy midwives and mid-level professionals, to improve access to, and utilization of maternal and newborn health services [22, 23].

In line with the national health agenda, the Last Ten Kilometers (L10K) project of JSI Research and Training Institute Inc. has supported the HEP to engage local communities to improve high-impact reproductive, maternal, newborn, and child health care behavior and practices in four of the most populous regions of the country (i.e., Amhara, Oromia, Southern Nations, Nationalities and Peoples [SNNP], and Tigray) since 2008. Between 2013 and 2017, the project scaled its platform activities in 115 woredas to engage local communities to identify and address barriers to access maternal and newborn health services particularly to identify pregnant women and ensure they received antenatal, intrapartum, and postpartum care. Besides, since 2014, L10K has implemented family conversation and birth notification strategies to promote birth preparedness, essential newborn care, and early postnatal care [24].

Data source

The data used for this study were obtained from a cross-sectional population-based study representing the 115 rural woredas which were carried out by L10K Project from October‐November 2017. The survey population included women of the reproductive age group (15–49 years) who had a live birth in the 12 months before the survey. The survey employed a two-stage stratified cluster sampling method stratified by the administrative region where kebeles were selected first as primary sampling units with the probability of selection being proportionate to its population size. This was undertaken to enumerate a representative sample of 2,724 women aged 15 to 49 years who had a live birth in the 12 months preceding the survey. The details of the design are described elsewhere [25].

The data were gathered through face-to-face interviews with mothers. During the interview, information about household and socio-demographic characteristics of mothers, awareness and access to health services, and experiences related to the women’s use of maternal health services, was collected from women with children in their first year of life. The questionnaire (S1 Appendix) was translated into local languages (Amharic, Oromifa, and Tigrigna). Details of data collection processes are described elsewhere [26].

The adequacy of the sample size to address the study objective was assured considering 95% confidence level (Zα/2 = 1.96), design effect (D = 2), and power of 80% for double population formula for comparative cross-sectional study design. Based on Anderson’s health-seeking behavior model [27], a health service utilization model that provides a framework to systematically describe factors that influence individual decisions to use (or not use) health care services, researchers considered different exposure variables [5, 28] including lack of women’s autonomy, 1–2 parity, no media exposure, no difficulty of distance to access medical care, no difficulty of transport arrangement to access medical care, no maternal education, and poorest wealth quintile as exposure and highest parity/5+, having autonomy, exposed to media, having difficulty of distance and transport arrangement to access medical care, higher education, and richest wealth quintile as non-exposure. Adding a 10% non-response rate, the maximum sample size obtained by women’s autonomy to healthcare decision-making was 2,501 for completed CoC at pregnancy [28].

Measurement

The outcome variables of interest of the study were the uptake of the CoC at antepartum, intrapartum, and postpartum stages: 1) continuum of care at the antepartum stage is women who received four or more antenatal care (ANC4+) visits, 2) continuum of care at the intrapartum stage is those women who continued use of skilled birth attendance after receiving ANC4+ visits, and 3) continuum of care at the postpartum stage or complete continuum is those women who received PNC for the mothers and their newborns, within six weeks of their delivery (either in a facility or at home) after receiving both ANC4+ visits and delivered by skilled assistance. Description and measurement of variables are presented in Table 1 below.

Table 1

Description and measurement of variables, Ethiopia, 2017.

Variables	Descriptions	Measurements
Utilization of ANC services (ANC 4+ visits)	Having health facility visits at least 4 times for pregnancy check-ups by skilled attendants during pregnancy. It is defined as continuum of care at ANC4+ (ANC4+ attainment).	Categorized into at least 4 visits and less than 4 visits or none
Skilled birth attendance (SBA)	It is defined as women who were assisted by a health professional (doctor, nurse, or midwife) during their last childbirth.	It is measured interview of women who were the primary person that assisted them with the delivery of their recent birth.
Postnatal care (PNC)	It is defined as women and their newborns who received postpartum care at the health facility or their home within six weeks of delivery.	It is measured interview of women whether pre-discharge care provided for them and their newborns after 24-hours of stay for whom delivered at the health facility and any postnatal check-up for mother’s and newborn’s health by a health care provider within 6 weeks of giving birth. The check-ups for the newborn and mother were separately inquired.
Continuum of care	In this study, the continuum of care is defined as the proportion of women who received maternal services at the pregnancy, delivery, and post-delivery stages. Accordingly, women who received all the following components were considered to have completed the continuum of care;	Continuum of care at delivery was obtained from ANC4+ and SBA variables; while a complete continuum of care was obtained from ANC4+, SBA, PNC variables
	At least four ANC visits (ANC4+) by health service providers at a health facility or home, Delivery assisted by a health professional (i.e., doctor, nurse, midwife, or health officer/), and At least one PNC check-ups for mother within six weeks after delivery by health service providers at a health facility or home At least one PNC check-ups for mother within six weeks after delivery by health service providers at a health facility or home	The outcome for Model II is 1 for receiving antenatal care and skilled birth attendance, and 0 for receiving ANC but not SBA. After delivery, some women received PNC and some did not. Thus we fit Model III among women who received ANC and SBA to identify factors associated with completion of the CoC. The two categories of the outcome for Model III are 1 for receiving ANC, SBA, and PNC, and 0 for receiving ANC & SBA but not PNC.
Complete ANC	Complete ANC service was measured by the four content or items of care women received during ANC visits (i.e., blood pressure measured, weight taken, and blood and urine tested during last pregnancy).	Information on these items of ANC content was derived from the interviewing women whether they received these services as part of their ANC consultation during their last pregnancy. Based on the information, we created a complete ANC service if women received all four contents of care or otherwise no complete ANC service if women did not receive all four contents of care.
Distance from health facility	Approximate reporting walking distance from respondents home to the nearest health post or health center in minutes	The average distance was computed for each respondent and dichotomized as < 30 minutes and > = 30 minutes
Maternity waiting homes (MWHs)	These are residential facilities located near a hospital or a health center to accommodate women in their final weeks of pregnancy to bridge the geographical gap in obstetric care for women with poor access to facilities [29].	Measured by asking whether respondents stayed at the MWHs during their last weeks of pregnancy or not
Family conversation	Family conversation is a forum conducted at the house of a pregnant woman with her family members and relatives who are expected to support her during pregnancy, labor, delivery, and the postpartum period. It creates an opportunity to discuss issues such as birth preparedness and essential newborn practices with all these family members together.	It is measured by asking women regarding their attendance of at least one family conversation at home during their last pregnancy.
Birth notification	It is a strategy introduced to promote early postnatal care.	It is measured through interviews of women whether they took measures to inform the Health Extension Workers (HEWs) about their childbirth immediately after delivery or not.
Pregnant women’s conference	Pregnant women’s conference is a pregnant women’s group meeting for peer learning to seek maternal and newborn health care which is facilitated by health care providers. It creates an opportunity to discuss issues such as birth preparedness and essential newborn practices with all these family members together.	It is measured by asking women about their attendance at least one conference during their last pregnancy.
Model family	Model families are defined as those households who received training from HEWs, acquire the necessary knowledge, skills, and behavior in health practices and demonstrate practical changes in the use of health service programs and serve as models in their community [21].	Measured by asking whether the respondent’s family is currently recognized as a model family or not
Household wealth index	A wealth index score was constructed for each household with the principal component analysis of the household’s possessions (electricity, watch, radio, television, mobile phone, telephone, refrigerator, table, chair, bed, electric stove, and kerosene lamp), and household characteristics (type of latrine, water source, floor, and wall material). Subsequently, households were ranked according to wealth score and then divided into five quintiles using the Principal Component Analysis method [30].	Household assets ownership was assessed and the wealth index was computed by using principal component analysis. The wealth status was categorized into five groups and ranked from the poorest to the wealthiest quintile.

For selecting predictor variables at individual and community levels, we adopted Anderson’s behavioral model for healthcare use [27]. The individual-level variables include wealth status, maternal education, distance to the health facility, being a model family, participation in pregnancy conference, having family conversation, early ANC booking, complete ANC service, and use of MWHs for prenatal stay. Besides, infant’s birth weight, mode of delivery and birth notification are included as predictor variables for CoC at the postpartum stage. The community variables considered in the study include region and area of residence (clusters/kebeles).

Data analysis

Data were analyzed using Stata version 15. The characteristics of the sample respondents were described by a set of background characteristics. The difference in the characteristics of the respondents was examined using Pearson’s chi-square statistics adjusted for cluster survey design effects.

Bivariate and multivariable mixed-effects logistic regression analyses were used to examine the predictors of the CoC accounting for cluster survey design, and the individual, and contextual characteristics of the respondents. We fitted three sequential random-effects logit regression models to examine the patterns of care-seeking and factors predicting the continuation of care. We fitted Model I among women receiving ANC4+ as the outcome (i.e., coded 1 for receiving ANC4+, otherwise 0); Model II among women who received ANC4+ to determine the factors associated with the continuity having skilled birth attendance (i.e., coded 1 for receiving ANC4+ and SBA, otherwise 0); and Model III fitted for women who received ANC4+ and SBA to identify factors associated with women returning for PNC visits or completion of the CoC (i.e., coded 1 for receiving ANC, SBA, and PNC, otherwise 0). The random-effects model accounts for the fact that people who live in the same area share similar characteristics and examine the proportion of variance explained by community-level factors (unobserved). We present the adjusted odds ratios and confidence intervals at the 95% level wherever applicable.

The global Wald’s statistics, the likelihood ratio test of the cluster-level random effects, and sensitivity of the quadrature approximation were used to assess the goodness-of-fit of the models. Regression diagnostic Akaike Information Criterion (AIC) was used to determine the suitability of the model.

Ethics approval and consent to participate

For this study, we obtained permission to use the data from the JSI, and ethics approval was obtained from the Research and Ethics Committee of the Department of Health Studies of the University of Gondar (reference number V/P/RCS/05/2505/2019; dated 25 August 2019). The original study was ethically approved by the Ethical Review Boards of Amhara, Oromia, SNNP, and Tigray Regional Health Bureaus, and JSI. Verbal consent from respondents was sought and documented by interviewers before interviewing. Voluntary participation was ensured during interviews [25].

Results

Characteristics of the study participants

A total of 2,724 women with live births in the 12 months preceding the survey were included in this analysis. The socio-demographic characteristics of respondents are described elsewhere [26].

Utilization of maternal and newborn health services

Most women (n = 2,481; 91.2% [95% CI: 89.1–93.0]) received at least one ANC during their recent pregnancy. However, only about a quarter of them (n = 698; 26.3% [95% CI: 23.9–28.9]) started ANC in their first trimester of pregnancy. A little more than half n = 1,387; 56.3% [95% CI: 54.5–58.2]) of them received complete ANC services (i.e., weight taken, blood pressure measured, blood test, and urine test conducted). About two-thirds (n = 1,816; 67.4% [95% CI: 62.8–71.6]) of them were attended by a skilled health provider (doctor, nurse, or midwife) during delivery at their last birth and about 5.6% of these were delivered through cesarean section. However, about 36.8% (n = 1,002; [95% CI: 33.3–40.4]) of women received PNC service within six weeks after delivery following their last childbirth; 29.4% of them received PNC within 48 hours; and 34.1% within 7 days.

Continuum of maternal and newborn health care

As presented in Fig 1 of the error bar, a little higher than half (52.3%) of the women continued receiving ANC4+, 42.4% continued for skilled birth attendance, and about one-fifth (21.5%) of women received all three packages of services during pregnancy, childbirth, and postpartum periods. However, one-fifth of them did not receive any maternal and newborn health services. Also, around 8.4% of women receive at least one ANC from skilled providers but did not receive the other two services. About 12.4% of them received skilled birth assistance during childbirth without receiving pregnancy care. Similarly, few women received PNC (1.6%) without having antepartum and intrapartum care About 10% and 21% of women discontinued the CoC at the delivery and postnatal stage, respectively (Table 2).

Fig 1

Utilization of skilled maternal and newborn care across the continuum.

Table 2

Achievement of the continuum of maternal and newborn health care in Ethiopia, 2017.

Antepartum (ANC4+)	Intrapartum (ANC4+ & SBA)	Postpartum (ANC4+, SBA, & PNC)	Number (n)	Percent
Yes	Yes	Yes	564	21.47
Yes	Yes	No	550	20.95
Yes	No	Yes	40	1.45
Yes	No	No	221	8.43
No	Yes	Yes	327	12.46
No	Yes	No	325	12.38
No	No	Yes	42	1.60
No	No	No	556	21.18
Total			2,626	100.00

The characteristics of women who completed different stages of care are shown in Tables 3 and 5. Of those women who resided within 30 minutes walking distance from the facility, 713 (55.3%) had four and above ANC visits, 584 (45.7%) continued to skilled delivery, 312 (24.4%) completed CoC at the postnatal stage. Besides, the distribution of the utilization of the package of maternal and newborn health services across the continuum significantly varied by region, educational status of the mother, and wealth quintile. Tigray and Oromia regions appeared to be better performing than Amhara region. Those mothers with better education residing near health facilities and those with better wealth appeared to better use the services across the continuum (Table 3).

Table 3

Distribution of maternal health services across the CoC in Ethiopia, by sociodemographic characteristics, 2017.

Characteristics	Continued at antenatal care (received ANC 4)		Continued at delivery stage		Complete continuum (continued at postpartum)
	n (%)	95% CI	n (%)	95% CI	n (%)	95% CI
Maternal age
<20	83 (52.6)	42.2–62.7	73 (46.6)	36.6–56.8	32 (20.8)	13.8–30.2
20–34	1,077 (52.6)	48.9–56.3	859 (42.4)	38.3–46.7	447 (22.1)	19.0–25.4
35–49	227 (50.6)	43.7–57.4	183 (41.0)	34.2–48.2	85 (19.1)	14.5–24.6
Education
No education	721 (47.5)	43.4–51.7	538 (35.8)	31.5–40.3	249 (16.6)	13.7–19.9
Primary	310 (53.5)	48.0–58.9	252 (43.8)	37.7–50.0	141 (24.5)	19.7–30.2
Higher	355 (64.1)*	58.5–69.3	324 (59.2)*	53.0–65.0	173 (31.7)*	26.7–37.1
Religion
Orthodox	791 (50.2)	45.7–54.8	464 (41.2)	36.3–46.3	302 (19.3)	16.0–23.2
Protestant	284 (56.7)	50.2–62.9	223 (45.1)	37.7–52.7	148 (29.8)	23.1–37.7
Muslims	301 (54.1)	45.5–62.5	242 (44.0)	34.6–53.9	111 (20.2)	15.3–26.1
Traditional/others	11 (53.9)	36.2–70.7	5 (24.8)	11.2–46.4	3 (16.7)**	7.2–34.0
Wealth quintile
Poorest	224 (41.7)	35.0–48.8	166 (31.0)	25.0–37.6	68 (12.7)	9.0–17.7
Poorer	265 (50.0)	44.1–55.9	214 (40.7)	34.4–47.5	105 (19.9)	15.4–25.3
Middle	265 (50.1)	44.1–56.1	210 (40.0)	34.2–46.1	104 (19.8)	15.4–25.0
Richer	312 (58.0)	53.0–62.8	244 (46.4)	40.5–52.5	138 (26.2)	21.4–31.7
Richest	321 (61.8)*	55.3–67.9	280 (54.5)*	47.9–60.9	150 (29.1)*	24.2–34.6
Distance to the nearest health facility
<30 min	713 (55.3)	50.6–59.8	584 (45.7)	40.6–50.8	312 (24.4)	20.9–28.4
>=30 min	674 (49.5)**	45.2–53.7	530 (39.3)**	34.8–44.0	252 (18.7)*	15.4–22.4
Region
Amhara	379 (42.3)	36.5–48.7	285 (32.1)	26.1–38.7	114 (12.9)	9.0–18.0
Oromia	314 (45.2)	39.0–51.6	229 (33.5)	27.0–40.8	144 (21.0)	15.8–27.5
SNNP	433 (64.5)	58.8–69.9	368 (55.5)	48.1–62.7	179 (27.0)	21.5–33.3
Tigray	261 (66.1)*	58.2–73.1	232 (59.2)*	50.7–67.2	127 (32.4)*	25.9–39.5
Total	2,653 (52.3)	48.7–55.9	2,626 (42.4)	38.5–46.5	2,626 (21.5)	18.6–24.6

*P-value <0.001;

** P-value <0.05.

As presented in Table 4, the distribution of maternal health services across the CoC also varied by health-service-related and obstetric characteristics. The CoC at antepartum, intrapartum, and postpartum stages significantly differed among women who stayed at MWHs, who were recognized as model family, who attended at least 3 pregnancy conferences during last pregnancy, those who attended at least one family conversation during pregnancy, received ANC in the first trimester, and received complete ANC, as compared to those who did not. While those women who delivered through cesarean section and their birth was notified to HEWs received complete CoC significantly better than their counterparts (Table 4).

Table 4

Distribution of maternal health services across the CoC in Ethiopia, by health-service related and obstetric characteristics, 2017.

Characteristics	Continued at antenatal care (received ANC 4)		Continued at delivery stage		Complete continuum (continued at postpartum)
	n (%)	95% CI	n (%)	95% CI	n (%)	95% CI
Prenatal stay at MWHs
No	1,075 (49.6)	45.8–53.4	803 (37.5)	33.3–41.9	394 (18.4)	15.5–21.6
Yes	312 (64.2)*	57.7–70.3	311 (64.2)*	57.7–70.2	170 (35.2)*	29.6–41.2
Family currently recognized as a model family
No	848 (47.1)	43.2–51.0	648 (36.5)	32.3–40.8	276 (15.5)	13.1–18.3
Yes	539 (63.3)*	58.3–68.0	466 (54.9)*	49.4–60.4	288 (34.0)*	29.2–39.1
Attended at least 3 pregnancy conference during last pregnancy
No	1,167 (50.1)	46.4–53.8	929 (40.4)	36.4–44.4	460 (20.0)	17.3–23.0
Yes	220 (67.6)*	60.0–74.4	185 (56.9)*	47.4–66.0	104 (32.0)*	24.2–40.9
At least one family conversation that took place at home during pregnancy
No	1,228 (50.9)	47.2–54.5	970 (40.6)	36.6–44.7	452 (18.9)	16.3–21.8
Yes	159 (66.5)*	57.7–74.3	144 (61.2)*	52.4–69.4	112 (47.4)*	38.1–57.0
ANC in the first trimester
No	912 (46.7)	42.8–50.5	724 (37.5)	33.4–41.7	364 (18.8)	15.9–22.2
Yes	475 (68.0)*	62.3–73.3	390 (56.2)*	50.2–62.1	200 (28.8)*	24.4–33.6
Complete ANC
No	494 (42.6)	38.0–47.3	348 (30.2)	25.9–34.9	171 (14.9)	12.0–18.3
Yes	893 (59.9)*	55.8–63.8	766 (51.9)*	47.5–56.3	393 (26.6)*	22.9–30.6
Mode of delivery
Vaginal delivery	1,282 (51.0)	47.3–54.7	1,009 (40.6)	36.5–44.8	503 (20.2)	17.4–23.4
Cesarean section delivery	105 (75.5)*	66.0–83.0	104 (75.5)*	66.0–83.0	61 (44.2)*	35.6–53.3
Birth notification
No	712 (45.9)	41.4–50.4	537 (35.0)	30.4–39.8	213 (13.9)	11.3–16.9
Yes	675 (61.3)*	57.0–65.4	577 (52.9)*	48.2–57.5	351 (32.2)*	27.8–36.8
Total	2,653 (52.3)	48.7–55.9	2,626 (42.4)	38.5–46.5	2,626 (21.5)	18.6–24.6

*<0.001;

**<0.05.

Predictors of the continuum of care

Except for maternal education, other covariates remain statistically significant in the multivariable analysis. As presented in Model I, Table 5, the use of ANC4+ visits at the stage of pregnancy is significantly associated with the administrative region, wealth quintile, staying at MWHs, being model family, attendance at pregnancy conferences, and booked for ANC in the first trimester. Those women in the richest wealth quintile are more likely to complete the pathways to maternal and newborn health care. The odds of continued use of antepartum care among women who started ANC in the first trimester of pregnancy are about two times higher [adjusted odds ratio (AOR): 2.76; 95% confidence interval (CI): 2.24–3.41] than those who booked late.

Table 5

Factors associated with the CoC at antenatal, delivery, and postnatal care in Ethiopia, 2017.

Characteristics	Model 1 (ANC stage)			Model II (ANC4 & skilled delivery)			Model III (CoC)
	# of respondents	AOR (95% CI)	p-value	# of respondents	AOR (95% CI)	p-value	# of respondents	AOR (95% CI)	p-value
Maternal education
No education	1,518	1.00		1,502	1.00		1,502	1.00
Primary	581	0.98 (0.79–1.23)	0.893	576	1.00 (0.79–1.26)	0.990	576	1.05 (0.81–1.36)	0.708
Higher	555	1.07 (0.84–1.36)	0.594	548	1.32 (1.03–1.69)	0.030	548	1.20 (0.92–1.56)	0.184
Wealth quintile
Poorest	537	1.00			1.00			1.00
Poorer	529	1.70 (1.28–2.25)	<0.001	535	1.80 (1.33–2.42)	<0.001	535	1.66 (1.17–2.34)	0.001
Middle	529	1.59 (1.19–2.12)	0.002	526	1.55 (1.14–2.10)	0.005	526	1.42 (1.00–2.03)	0.051
Richer	538	2.57 (1.89–3.51)	<0.001	525	2.41 (1.74–3.34)	<0.001	525	2.36 (1.64–3.39)	<0.001
Richest	520	2.61 (1.90–3.60)	<0.001	525	2.99 (2.14–4.18)	<0.001	525	2.03 (1.40–2.94)	<0.001
Distance to the health facility				515
<30 min	1,291	1.11 (0.93–1.34)	0.254	1,278	1.23 (1.02–1.50)	0.034	1,348	1.33 (1.07–1.64)	0.009
> = 30 min	1,362	1.00		1348	1.00		1,278	1.00
Family recognized as a model family
Yes	852	1.57 (1.29–1.90)	<0.001	848	1.44 (1.18–1.78)	<0.001	848	1.59 (1.28–1.97)	<0.001
No	1,801	1.00		1,778	1.00		1,778	1.00
Prenatal stay at MWHs
Yes	486	1.38 (1.08–1.76)	0.009	484	2.34 (1.82–3.00)	<0.001	484	1.97 (1.53–2.54)	<0.001
No	2,167	1.00		2,142	1.00		2,142	1.00
Attended at least 3 pregnancy conferences during last pregnancy
Yes	326	1.96 (1.46–2.62)	<0.001	325	1.77 (1.33–2.37)	<0.001	325	1.28 (0.95–1.71)	0.105
No	2,327	1.00		2,301	1.00		2,301	1.00
Attended at least 1 family conversation during pregnancy
Yes	238	1.33 (0.97–1.82)	0.077	236	1.43 (1.04–1.96)	0.027	236	2.12 (1.56–2.88)	<0.001
No	2,415	1.00		2,390	1.00		2,390	1.00
Antenatal care utilization in the first trimester
Yes	698	2.76 (2.24–3.41)	<0.001	694	2.03 (1.65–2.49)	<0.001	694	1.46 (1.17–1.81)	0.001
No	1,955	1.00		1,932	1.00		1,932	1.00
Complete ANC
Yes	1,491			1,478	2.13 (1.74–2.61)	<0.001	1,478	1.80 (1.43–2.26)	<0.001
No	1,162			1,150	1.00		1,150	1.00
Mode of delivery
Vaginal delivery							2,488	1.00
Cesarean section delivery							138	2.70 (1.82–4.02)	<0.001
Birth notification
Yes								1.95 (1.56–2.43)	<0.001
No								1.00
Community characteristics
Region
Amhara	893	1.00		888			888	1.00
Oromia	694	1.32 (0.94–1.86)	0.107	682	1.49 (1.02–2.16)	0.037	682	2.78 (1.86–4.16)	<0.001
SNNP	671	2.94 (2.09–4.13)	<0.001	663	3.30 (2.29–4.74)	<0.001	663	2.73 (1.87–3.99)	<0.001
Tigray	395	2.82 (1.97–4.04)	<0.001	393	3.36 (2.30–4.92)	<0.001	392	2.85 (1.91–4.24)	<0.001
Community-level intercepts (SE)		0.22 (0.04)			0.07 (0.01)			0.02 (0.004)
Random effects
Community-level variance (SE)		0.39 (0.09)			0.46 (0.10)			0.34 (0.10)
Log-likelihood ratio test		48.23*			52.87*			24.32*
Model fit statistics
ICC (SE)		0.11 (0.02)			0.12 (0.02)			0.09 (0.02)
AIC		3212.72			3031.01			2539.54

Model II presents the predictors of continuation of care from pregnancy to delivery among women who received ANC 4+. All variables remain significant. Besides, higher maternal education is significantly associated with the CoC at the delivery stage [AOR: 1.32; 95% CI: 1.03–1.69]; those women who attended at least one family conversation during pregnancy had 43% higher odds of complete antepartum and intrapartum skilled care than those counterparts [AOR: 1.43; 95% CI: 1.04–1.96]; and those women whose homes were within less than 30 minutes walking distance to the health facility, had 30% more odds of continually utilizing package of services at delivery stage [AOR: 1.23; 95% CI: 1.02–1.50].

In Model III, administrative region, wealth quintile, being model family, having booked ANC in the first trimester, and having received complete ANC remained significantly associated with complete CoC at the postnatal stage among women who received ANC4+ visits and SBA. Additionally, attendance at least 1 family conversation during pregnancy, mode of delivery, and birth notified, significantly associated with complete CoC at the postnatal stage. Family conversation [AOR: 2.12; 95% CI: 1.56–2.88], cesarean delivery [AOR: 2.70; 95% CI: 1.82–4.02], and birth notified to HEWs [AOR: 1.95; 95% CI: 1.56–2.43] were found to be predictors of complete CoC (Table 5).

As presented above, wealth quintile, being model family, prenatal stay at MWHs, ANC booking in the first trimester, complete ANC, and administrative region were predictors for the CoC at all stages: at antepartum, intrapartum, and postpartum care. Besides, attendance at least 3 pregnancy conferences during the last pregnancy was a predictor for continued antepartum and intrapartum care. Family conversation predicted continued use of skilled delivery and postnatal care.

Discussion

Our analysis revealed that only one-fifth of women fully utilized maternal and newborn health service packages across the continuum of care. The compliance with continuity of care across the pathway showed a significant inequality that left the poorest and those living far from health facilities behind. Moreover, uptake of complete CoC at the postpartum stage varied by mode of delivery and birth notification.

The coverage of CoC is slightly higher as compared with the findings of similar studies undertaken in Ethiopia [31, 32] and other sub-Saharan African countries [11, 14, 33]. However, it is lower than in South Asian countries [11, 19, 28]. The uptake of services declined from antepartum to postpartum stages, along with the CoC, similar to other studies [15, 17]. The CoC framework has been highlighted to link service pathways along the CoC and each contact with the health system is an opportunity to amplify the effect of subsequent contact. However, in low-income countries, service delivery is often fragmented and weakly implemented, especially during the postnatal period [4].

The importance of more frequent ANC visits for a positive pregnancy experience and its linkage with subsequent use of skilled assistance during delivery and postnatal services is well recognized [34]. Nevertheless, many women drop out from the pathway of the continuum in sub-Saharan African countries [11] including Ethiopia [31, 32]. Likewise, another significant cohort of women also dropped from skilled assistance at delivery which is one of the main strategies to ensure safe motherhood and combat maternal mortality [35, 36] indicating the country is still a long way from universal access to skilled childbirth care. The postnatal period is critical for women as they may develop life-threatening complications that can, however, be promptly treated if postnatal care is accessed [37]. There is a need to redefine the delivery strategies to strengthen the facility-based PNC, staying for at least 24-hours after delivery and providing care as recommended by WHO and the national guide [20, 38]. With the current practice of early discharge [5] and difficulties to revisit facility due to lack of transport, costs, and cultural constraints [39] women and babies would not receive appropriate CoC at the postnatal stage.

Inequity due to wealth, where the poorer women were more disadvantaged is similar to other studies conducted in low-and-middle-income countries [11, 13, 14, 17, 19, 28, 33]. That is contrary to the pro-poor strategies adopted in many countries for increasing access to services for all women to end preventable maternal and neonatal mortality [40]. Evidence also shows that when interventions are offered as a package without adequate equity considerations, the poorest segment of the population would be very disadvantaged [41]. The regional differences in the delivery of the package of maternal and newborn health services across the continuum underlined the need to monitor CoC at the subnational level as documented in previous studies [14, 28, 33, 42].

Distance to facility influenced the odds of receiving CoC at the delivery and postpartum stages, but not for the antepartum stage. Mixed findings were reported by other studies. A study by Sakuma et al. [18] reported distance from the health facility was negatively associated with the CoC. However, a study by Mohan et al. [17] reported that distance to the nearest facility was not associated with an increased likelihood of receiving CoC. In Ethiopia, though the government expanded access to facilities and free ambulance transport, the country has poor road access, difficult terrain, and roads that are often unreachable during the rainy season. A recent study in Ethiopia affirms that women living in remote areas with no access to transportation did not have access to maternal health services [43]. This might discourage women to visit the facility for skilled birth, particularly during the night time and rainy season. Moreover, despite new health centers and primary hospitals being built in Ethiopia to achieve universal access to primary health care [21], there is still the need to further expand the reach to women for skilled delivery.

Early booking for antenatal care is the entry point for continued use of maternal and newborn health services. This is linked with the use of other maternal and newborn health services across the continuum pathway, as documented in previous studies [11, 16, 18, 28]. Early consultation and complete ANC service were also found to be associated with continued use of delivery and PNC care [10]. Uptake of complete antenatal care is connected to subsequent utilization of skilled birth attendance and postnatal care which is in line with previous studies [13]. Previous studies also documented urine samples taken were identified as predictors significantly associated with the continuation of care from having skilled birth attendants to receiving postnatal care [13, 15]. As such, program managers and policy-makers need to promote strategies and interventions to make ANC care more accessible, both in terms of early booking and content of ANC service. The level of integration and content of care at each visit should be given due attention and closely monitored.

The use of MWHs during pregnancy is found to be an independent factor regarding the continued use of skilled maternal and newborn health care in this study. Previous studies also revealed that the use of MWHs was linked to the use of maternal and newborn health services including facility birth and use of essential and emergency obstetric care [44]. This indicates a need to expand MWHs to bridge geographic barriers to access maternal health services as well as strengthen the readiness of the existing MWHs to host women postpartum and provide care in the critical period [44-46].

Being recognized as a model family was associated with the uptake of CoC. This shows the contribution of the HEP in mobilizing the community and HEWs efforts of educating families on uptake of maternal and newborn health services across the CoC pathway. Previous studies in Ethiopia also demonstrate that the model family was more likely to use maternal and newborn health services [47].

Household-level interactions with husband, neighbors, mothers-in-law, and pregnant women, as well as pregnant women’s group discussion facilitated by health care workers, have positive effects on the utilization of the continuum of maternal and newborn health care. Mothers, relatives, or women development groups who informed HEWs about the birth immediately after delivery and mothers who attended at least one family conversation during pregnancy and attended at least three pregnancy conferences during the last pregnancy, were more likely to complete CoC than their counterparts. This is in agreement with a previous observational study made on collaborative community-based quality improvement intervention in Ethiopia, which involved family meetings and labor and birth notification contributing to the increased receipt of PNC within 48 hours by skilled providers or HEWs [48]. This notification system might motivate HEWs to do home visits for the provision of PNC to the mother and the newborn.

Maternal education was significantly associated with the uptake of maternal health services at the delivery stage. However, in contrast to most studies [11, 31–33], maternal education and uptake of maternal health service across the continuum at antepartum and postpartum stages did not show significant association. While, in this study, maternal education has a positive association with the uptake of SBA [26] but with the utilization of PNC [26]. Possible explanations would be; first, educated women might book ANC late due to false self-efficacy and confidence as well as a perception that multiple visits were not necessary that might have led them to have fewer ANC visits [26]. Similarly, a study in Tanzania showed that women with primary or more education presented later in pregnancy than women with no schooling might be the reason for negative associations between education and four or more ANC visits [17]. Second, the lack of association with PNC might have related to the PNC service delivery strategy where PNC services in Ethiopia are provided mainly home-based. In line with this, some studies in Ethiopia also showed that maternal education did not have a significant association with the utilization of PNC [49, 50]. Third, we speculate that due to low-quality education, women with primary or more education in rural settings might not bring the necessary behavioral changes for better access to health or financial services than their counterparts that affects their autonomy to seek health services. Lastly, this might have also been related to the national pro-poor health policies and strategies where maternal and newborn health services in Ethiopia are provided free of charge.

The findings of this study may have been affected by social desirability and recall biases despite our efforts to minimize biases using memory aids, pretested the survey tools, trained data collectors, and allocated adequate days for data collection. Some aspects of the content of care and frequency of ANC visits might be over-reported which would positively overestimate the association between the outcome variable. In defining PNC, we included both postpartum pre-discharge care at the health facility as well as postpartum care provided after discharge to see the full spectrum of uptake of postpartum care. Accordingly, we examined predictors for pre-discharge and after-discharge care together which theoretically be different for pre-discharge care at the facility and care after discharge or at home. Another limitation of this study would be the presence of unmeasured confounders that would be correlated both with the outcome and predictor variables included in the model, particularly level-2 endogeneity arising from correlations between included individual characteristics and omitted community-level variables and/ or reverse causality of variables that would cause the model to be endogenous [51]. To minimize the potential endogeneity problems, we excluded predictor variables with the unclear direction of the effect with the outcome variables, for instance, “Complete ANC” was excluded from Model I of predicting ANC4+ uptake.

Conclusion

The main modifiable program factors connected to the continued uptake of maternal health services include family conversation, pregnant women conference, complete ANC, ANC in the first trimester, and birth notification. As such, program managers should give focus on these practical program strategies to ensure that community or household level interactions including family conversation during pregnancy, early identification of pregnancy and link to care, and complete ANC services, notification of birth for early PNC are available to all women to improve the continuity of maternal health service use.

Survey questionnaire.

Survey questionnaire we used to collect information from study participants. The first sheet contains variable definitions (data dictionary) in English and other local languages (Amharic, Oromiffa, and Tigregna), and the second sheet contains variable answer choices.

(XLSX)

Click here for additional data file.

Survey dataset.

This is survey data with variables and their values we used for the analysis.

(XLS)

Click here for additional data file.

4 Sep 2021

PONE-D-20-37879

Predictors of maternal and newborn health service utilization across the continuum of care in Ethiopia: a multilevel analysis

PLOS ONE

Dear Dr. Tiruneh,

Thank you for submitting your manuscript to PLOS ONE. After careful consideration, we feel that it has merit but does not fully meet PLOS ONE’s publication criteria as it currently stands. Therefore, we invite you to submit a revised version of the manuscript that addresses the points raised during the review process.

Please pay particular attention to addressing the concerns and points of methodological and interpretative clarification raised by Reviewer 1.

Please submit your revised manuscript by Oct 12 2021 11:59PM. If you will need more time than this to complete your revisions, please reply to this message or contact the journal office at plosone@plos.org. When you're ready to submit your revision, log on to https://www.editorialmanager.com/pone/ and select the 'Submissions Needing Revision' folder to locate your manuscript file.

Please include the following items when submitting your revised manuscript:

A rebuttal letter that responds to each point raised by the academic editor and reviewer(s). You should upload this letter as a separate file labeled 'Response to Reviewers'.

A marked-up copy of your manuscript that highlights changes made to the original version. You should upload this as a separate file labeled 'Revised Manuscript with Track Changes'.

An unmarked version of your revised paper without tracked changes. You should upload this as a separate file labeled 'Manuscript'.

If you would like to make changes to your financial disclosure, please include your updated statement in your cover letter. Guidelines for resubmitting your figure files are available below the reviewer comments at the end of this letter.

If applicable, we recommend that you deposit your laboratory protocols in protocols.io to enhance the reproducibility of your results. Protocols.io assigns your protocol its own identifier (DOI) so that it can be cited independently in the future. For instructions see:  http://journals.plos.org/plosone/s/submission-guidelines#loc-laboratory-protocols . Additionally, PLOS ONE offers an option for publishing peer-reviewed Lab Protocol articles, which describe protocols hosted on protocols.io. Read more information on sharing protocols at  https://plos.org/protocols?utm_medium=editorial-email&utm_source=authorletters&utm_campaign=protocols .

We look forward to receiving your revised manuscript.

Kind regards,

Jamie Males

Staff Editor

PLOS ONE

Journal Requirements:

When submitting your revision, we need you to address these additional requirements.

1. Please ensure that your manuscript meets PLOS ONE's style requirements, including those for file naming. The PLOS ONE style templates can be found at

https://journals.plos.org/plosone/s/file?id=wjVg/PLOSOne_formatting_sample_main_body.pdf and https://journals.plos.org/plosone/s/file?id=ba62/PLOSOne_formatting_sample_title_authors_affiliations.pdf

2. We note that your study is closely related to the following publication, on which you are an author:

https://bmcpregnancychildbirth.biomedcentral.com/articles/10.1186/s12884-020-03254-7

Although you have briefly cited the above study in the discussion section of your article, we feel that the scientific rationale of the current study and the contribution that it makes to the field should be better justified.  Therefore, please cite and discuss the above study in the introduction and discussion sections of your manuscript, clarifying how the present work is related to the previously published paper.

Please note that our second publication criterion states that "If a submitted study replicates or is very similar to previous work, authors must provide a sound scientific rationale for the submitted work and clearly reference and discuss the existing literature. Submissions that replicate or are derivative of existing work will likely be rejected if authors do not provide adequate justification.  http://www.plosone.org/static/publication.action#results.

Thank you for your attention to this request.

3. Your ethics statement should only appear in the Methods section of your manuscript. If your ethics statement is written in any section besides the Methods, please move it to the Methods section and delete it from any other section. Please ensure that your ethics statement is included in your manuscript, as the ethics statement entered into the online submission form will not be published alongside your manuscript.

4. We noticed you have some minor occurrence of overlapping text with the following previous publication(s), which needs to be addressed:

- https://bmcpregnancychildbirth.biomedcentral.com/articles/10.1186/s12884-020-03254-7

The text that needs to be addressed involves the results section.

In your revision ensure you cite all your sources (including your own works), and quote or rephrase any duplicated text outside the methods section. Further consideration is dependent on these concerns being addressed.

[Note: HTML markup is below. Please do not edit.]

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. Is the manuscript technically sound, and do the data support the conclusions?

The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented.

Reviewer #1: Yes

Reviewer #2: Yes

**********

2. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #1: Yes

Reviewer #2: Yes

**********

3. Have the authors made all data underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #1: Yes

Reviewer #2: Yes

**********

4. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

Reviewer #1: No

Reviewer #2: Yes

**********

5. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #1: I have attached a separate word document with the same comments (for correct formatting).

Review comments for PONE-D-20-37879

Overall:

This is an interesting paper and I believe the findings are important in the context of Ethiopia. However, I do see a lot of areas that need major/significant revision. I also see a lot of grammatical errors throughout the manuscript. Hence, I would suggest that the authors first address the following issues before the manuscript can be considered for publication in PLOS ONE.

Introduction:

• The following sentence needs to be revised in order to clarify what you mean

o “The continuum of maternal and newborn care provides a framework for delivering high-impact interventions organized in health service packages, ensuring appropriate linkages between family and community care, outreach and outpatient services, and the facility level across the tier to prevent maternal mortality (2).”

• CoC is not defined at first mention in the main body of the text. Please define all acronyms at first mention within the body of the text.

• I do not think that calling individual, interpersonal and household level determinants of service use “demand-side” factors is an appropriate use of the term. Instead of classifying factors as “demand-side” or “supply-side”. I would suggest just calling them what they are (i.e. individual factors, interpersonal factors, household factors, etc.)

• The author mentions that the “evidence on factors associated with adherence to the continuum of maternal and newborn health care is not widely available”. While this is largely true, there are studies done in low- and middle-income countries that looked at determinants of continuum of care. Please cite some of these studies and briefly talk about their findings to inform your study in the introduction. I see that you cite other relevant studies in the discussion section. Also, briefly mention them in the introduction section. Below is a list of references to get started:

o Determinants of continuum of care for maternal, newborn and child health services in rural Khammouane, Lao PDR by Saki Sakuma, Junko Yasuoko, Khampheng Phongluxa and Masamine Jimba (PLOS ONE; https://doi.org/10.1371/journal.pone.0215635)

o Assessing the Continuum of Care Pathway for Maternal Health in South Asia and Sub-Saharan Africa by Kavita Singh, William T. Story and Allisyn C. Moran (Maternal Child Health Journal; doi: 10.1007/s10995-015-1827-6)

o Factors associated with the continuum of care for maternal, newborn and child health in The Gambia: a cross-sectional study using Demographic and Health Survey 2013 by Jiyoung Oh, Juyoung Moon, Jae Wook Choi and Kyunghee Kim (BMJ Open; http://dx.doi.org/10.1136/bmjopen-2019-036516)

o Analysis of dropout across the continuum of maternal health care in Tanzania: findings from a cross-sectional household survey by Diwakar Mohan, Amnesty E LeFevre, Asha George, Rose Mpembeni, Eva Bazant, Neema Rusibamayila, Japhet Killewo, Peter J Winch and Abdullah H Baqui (Health Policy and Planning; https://doi.org/10.1093/heapol/czx005)

o Levels and determinants of continuum of care for maternal and newborn health in Cambodia-evidence from a population-based survey by Wenjuan Wang and Rathavuth Hong (BMC Pregnancy and Childbirth; doi: 10.1186/s12884-015-0497-0)

• The following sentence is incomplete and grammatically incorrect. Please check throughout the manuscript to correct errors like this.

o “Why the utilization of maternal and newborn health services has not improved in Ethiopia despite the huge national efforts and investments made in the last decade.”

Methods:

• For readers who may not be familiar with “Anderson’s health-seeking behavior model”, please briefly describe the model in the text.

• The author states that the outcome variable of interest was the uptake of the CoC. However, this seems like an oversimplification of what the authors actually did. I see at least three outcome variables: ANC 4+ as the outcome, continuum of care up to delivery and continuum of care through postnatal care. Please describe each outcome variable separately and in detail in the opening paragraph of the “Measurement” sub-section.

• In Table 1, the descriptions for SBA and PNC are ambiguous. For example, defining them as “proportion of women” seems incorrect as you are looking at how an individual woman responded to the corresponding questions.

• In Table 1, the author mentions that PNC included pre-discharge care at the health facility as well as care provided after discharge. Theoretically, determinants for pre-discharge care at the facility and determinants for care after discharge or at home should be treated differently. However, if you decide to keep them together, I would mention how the nature of the determinants could be different for pre-discharge care and after-discharge/home care in the discussion/limitations section.

• I would advise reconsidering the variable “Complete ANC” in the analysis models. This is because the direction of the effect between Complete ANC and ANC4+ is unclear. The authors are trying to predict ANC4+ but ANC4+ may in turn cause Complete ANC. Reverse causality causes the model to be endogenous and estimates unreliable. I would suggest omitting “Complete ANC” entirely when predicting ANC4+.

• Also, have the authors checked for multicollinearity between model family, pregnant women’s conference, family conversation and maternity waiting homes? I wonder if there is high collinearity between these variables and if there are, it could throw off standard errors. If there is high collinearity, I suggest dropping variables that are very highly collinear.

• The three sequential models are fine to use but you could also just use the entire sample for all three models:

o Model 1:

�  Outcome = 1 if ANC4+

�  Outcome = 0 if not

o Model 2:

�  Outcome = 1 if ANC4+ & SBA

�  Outcome = 0 if not

o Model 3:

�  Outcome = 1 if ANC4+ & SBA & PNC

�  Outcome = 0 if not

o If you decide to keep the three sequential models, make sure to be clear that the findings from Model 2 and Model 3 apply to a sub-sample. This means that based on your findings (from Model 2 and Model 3), program decision-makers can gain insight on a sub-population of women who have followed through to ANC (Model 2) or delivery (Model 3) but not about the general population of pregnant women. However, if you decide to use the alternative models described in the bullet points above, you can have the advantage of generalizing your findings to the general population of pregnant women.

Results:

• There are a few places where the author explains what the findings “mean”. I would suggest reserving interpretations of the findings to the discussion section. The results section should just report the findings and not try to interpret them.

• I don’t think it’s appropriate to say “CoC at the antepartum stage”. Just using antenatal care is not a “continuum of care”.

• Be clear that Model 2 and Model 3 need to be interpreted in the context of the sub-sample. For example, say “In Model 3, administrative region, wealth quintile, being model family, having booked ANC in the first trimester and having received complete ANC remained significantly associated with complete COC at the postnatal stage among women who received four or more ANC visits and skilled birth attendance.”

Discussion:

• Overall, the discussion section is well-organized, rich and clearly written.

• If you decide to keep the three sequential models, I would ask the authors make it very clear in the discussion section that the findings (particularly from Model 2 and Model 3) are not generalizable to the larger population. Rather, those findings (from Model 2 and Model 3) specifically apply to a sub-sample of women who have followed through to ANC (Model 2) or delivery (Model 3). Hence, programmatic implications are only applicable to those women who have followed through to ANC or delivery and not to the general population of pregnant women.

• I would also focus on fixing grammatical errors and awkward sentences.

• Lastly, I would suggest expanding the discussion of methodological limitations a little bit (i.e. potential issues of endogeneity).

Reviewer #2: the authors investigated the predictors of maternal and newborn health service utilization across the continuum of care in Ethiopia. the authors did a great work and have added to the body of knowledge on maternal health.

the authors should edit their writeup and make sure where ever they use an abbreviation for the first time, they first indicate it in full and the abbreviation in bracket.

**********

6. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #1: No

Reviewer #2: No

[NOTE: If reviewer comments were submitted as an attachment file, they will be attached to this email and accessible via the submission site. Please log into your account, locate the manuscript record, and check for the action link "View Attachments". If this link does not appear, there are no attachment files.]

While revising your submission, please upload your figure files to the Preflight Analysis and Conversion Engine (PACE) digital diagnostic tool, https://pacev2.apexcovantage.com/. PACE helps ensure that figures meet PLOS requirements. To use PACE, you must first register as a user. Registration is free. Then, login and navigate to the UPLOAD tab, where you will find detailed instructions on how to use the tool. If you encounter any issues or have any questions when using PACE, please email PLOS at figures@plos.org. Please note that Supporting Information files do not need this step.

Submitted filename: Review comments for PONE-D-20-37879.docx

Click here for additional data file.

8 Nov 2021

A point-by-point response to reviewers' comments

Journal: PLOS ONE

PONE-D-20-37879

Predictors of maternal and newborn health service utilization across the continuum of care in Ethiopia: a multilevel analysis

PLOS ONE

The authors would like to appreciate and thank the reviewers and editors for the constructive comments. Our point-by-point responses are below each of the comments in italics.

Reviewer #1:

Overall:

This is an interesting paper and I believe the findings are important in the context of Ethiopia. However, I do see a lot of areas that need major/significant revision. I also see a lot of grammatical errors throughout the manuscript. Hence, I would suggest that the authors first address the following issues before the manuscript can be considered for publication in PLOS ONE.

Introduction:

• The following sentence needs to be revised in order to clarify what you mean

o “The continuum of maternal and newborn care provides a framework for delivering high-impact interventions organized in health service packages, ensuring appropriate linkages between family and community care, outreach and outpatient services, and the facility level across the tier to prevent maternal mortality (2).”

Thanks for the comment. To clarify for the readers, it is revised as follows, “The continuum of maternal and newborn care is highlighted as a framework for delivering high-impact interventions across the pregnancy, childbirth, and postpartum life-course and the level of service delivery to prevent maternal and newborn mortalities (2).”

• CoC is not defined at first mention in the main body of the text. Please define all acronyms at first mention within the body of the text.

Comment well taken and addressed in this version.

• I do not think that calling individual, interpersonal and household level determinants of service use “demand-side” factors is an appropriate use of the term. Instead of classifying factors as “demand-side” or “supply-side”. I would suggest just calling them what they are (i.e. individual factors, interpersonal factors, household factors, etc.)

Many thanks for the comment. Comment well taken and addressed in this version.

• The author mentions that the “evidence on factors associated with adherence to the continuum of maternal and newborn health care is not widely available”. While this is largely true, there are studies done in low- and middle-income countries that looked at determinants of continuum of care. Please cite some of these studies and briefly talk about their findings to inform your study in the introduction. I see that you cite other relevant studies in the discussion section. Also, briefly mention them in the introduction section. Below is a list of references to get started:

o Determinants of continuum of care for maternal, newborn and child health services in rural Khammouane, Lao PDR by Saki Sakuma, Junko Yasuoko, Khampheng Phongluxa and Masamine Jimba (PLOS ONE; https://doi.org/10.1371/journal.pone.0215635)

o Assessing the Continuum of Care Pathway for Maternal Health in South Asia and Sub-Saharan Africa by Kavita Singh, William T. Story and Allisyn C. Moran (Maternal Child Health Journal; doi: 10.1007/s10995-015-1827-6)

o Factors associated with the continuum of care for maternal, newborn and child health in The Gambia: a cross-sectional study using Demographic and Health Survey 2013 by Jiyoung Oh, Juyoung Moon, Jae Wook Choi and Kyunghee Kim (BMJ Open; http://dx.doi.org/10.1136/bmjopen-2019-036516)

o Analysis of dropout across the continuum of maternal health care in Tanzania: findings from a cross-sectional household survey by Diwakar Mohan, Amnesty E LeFevre, Asha George, Rose Mpembeni, Eva Bazant, Neema Rusibamayila, Japhet Killewo, Peter J Winch and Abdullah H Baqui (Health Policy and Planning; https://doi.org/10.1093/heapol/czx005)

o Levels and determinants of continuum of care for maternal and newborn health in Cambodia-evidence from a population-based survey by Wenjuan Wang and Rathavuth Hong (BMC Pregnancy and Childbirth; doi: 10.1186/s12884-015-0497-0)

Many thanks for the valid comments. All these articles are cited in the Background and Discussion sections in the revised version.

• The following sentence is incomplete and grammatically incorrect. Please check throughout the manuscript to correct errors like this.

o “Why the utilization of maternal and newborn health services has not improved in Ethiopia despite the huge national efforts and investments made in the last decade.”

Comment well taken and addressed.

Methods:

• For readers who may not be familiar with “Anderson’s health-seeking behavior model”, please briefly describe the model in the text.

Comment is well taken and a brief description of the model included.

• The author states that the outcome variable of interest was the uptake of the CoC. However, this seems like an oversimplification of what the authors actually did. I see at least three outcome variables: ANC 4+ as the outcome, continuum of care up to delivery and continuum of care through postnatal care. Please describe each outcome variable separately and in detail in the opening paragraph of the “Measurement” sub-section.

Comment well taken and addressed as follows, “The outcome variables of interest of the study were the uptake of the CoC at antepartum, intrapartum, and postpartum stages: 1) continuum of care at the antepartum stage is women who received ANC 4+ visits, 2) continuum of care at the intrapartum stage is those women who continued use of skilled birth attendance after receiving ANC4+ visits, and 3) continuum of care at the postpartum stage or complete continuum is those women who received PNC for the mothers and their newborns, within six weeks of their delivery (either in a facility or at home) after receiving both ANC4+ visits and delivered by a skilled assistance.”

• In Table 1, the descriptions for SBA and PNC are ambiguous. For example, defining them as “proportion of women” seems incorrect as you are looking at how an individual woman responded to the corresponding questions.

Comment is well taken and addressed in this version.

• In Table 1, the author mentions that PNC included pre-discharge care at the health facility as well as care provided after discharge. Theoretically, determinants for pre-discharge care at the facility and determinants for care after discharge or at home should be treated differently. However, if you decide to keep them together, I would mention how the nature of the determinants could be different for pre-discharge care and after-discharge/home care in the discussion/limitations section.

Thanks for the valuable comment. In the discussion section, we included the following sentence as a limitation of this paper. “In defining PNC, we included both postpartum pre-discharge care at the health facility as well as postpartum care provided after discharge to see the full spectrum of uptake of postpartum care. Accordingly, we examined predictors for pre-discharge and after-discharge care together which theoretically be different for pre-discharge care at the facility and care after discharge or at home.”

• I would advise reconsidering the variable “Complete ANC” in the analysis models. This is because the direction of the effect between Complete ANC and ANC4+ is unclear. The authors are trying to predict ANC4+ but ANC4+ may in turn cause Complete ANC. Reverse causality causes the model to be endogenous and estimates unreliable. I would suggest omitting “Complete ANC” entirely when predicting ANC4+.

Thanks for the valid and educational comment. It is well taken and Complete ANC is omitted from Model I. The results including the result Table are edited accordingly.

• Also, have the authors checked for multicollinearity between model family, pregnant women’s conference, family conversation and maternity waiting homes? I wonder if there is high collinearity between these variables and if there are, it could throw off standard errors. If there is high collinearity, I suggest dropping variables that are very highly collinear.

Thanks a lot for the valid comments. We have checked for multi-collinearity to see the assumption of independence of variables in the multiple regression model and we found out a mean-variance inflation factor (vif) of 1.05 that there is no multi-collinearity problem.

• The three sequential models are fine to use but you could also just use the entire sample for all three models:

o Model 1:

�  Outcome = 1 if ANC4+

�  Outcome = 0 if not

o Model 2:

�  Outcome = 1 if ANC4+ & SBA

�  Outcome = 0 if not

o Model 3:

�  Outcome = 1 if ANC4+ & SBA & PNC

�  Outcome = 0 if not

o If you decide to keep the three sequential models, make sure to be clear that the findings from Model 2 and Model 3 apply to a sub-sample. This means that based on your findings (from Model 2 and Model 3), program decision-makers can gain insight on a sub-population of women who have followed through to ANC (Model 2) or delivery (Model 3) but not about the general population of pregnant women. However, if you decide to use the alternative models described in the bullet points above, you can have the advantage of generalizing your findings to the general population of pregnant women.

Thanks for the detailed explanation. We defined the three models exactly how you described them above. Few differences in the sample size for the three models are due to missing values where we recoded “Don’t know” values as missing. In the analysis section, we clearly defined how the outcome variables are coded 1 or 0 for further clarification.

Results:

• There are a few places where the author explains what the findings “mean”. I would suggest reserving interpretations of the findings to the discussion section. The results section should just report the findings and not try to interpret them.

Comment is well taken and revisited in this version.

• I don’t think it’s appropriate to say “CoC at the antepartum stage”. Just using antenatal care is not a “continuum of care”.

Comment is well taken and revisited in this version.

• Be clear that Model 2 and Model 3 need to be interpreted in the context of the sub-sample. For example, say “In Model 3, administrative region, wealth quintile, being model family, having booked ANC in the first trimester and having received complete ANC remained significantly associated with complete COC at the postnatal stage among women who received four or more ANC visits and skilled birth attendance.”

Comment well taken and revisited in this version.

Discussion:

• Overall, the discussion section is well-organized, rich and clearly written.

Thanks much.

• If you decide to keep the three sequential models, I would ask the authors make it very clear in the discussion section that the findings (particularly from Model 2 and Model 3) are not generalizable to the larger population. Rather, those findings (from Model 2 and Model 3) specifically apply to a sub-sample of women who have followed through to ANC (Model 2) or delivery (Model 3). Hence, programmatic implications are only applicable to those women who have followed through to ANC or delivery and not to the general population of pregnant women.

Thanks for the comment. As we explained earlier we defined the three models exactly how you described them above. In this version, we clearly defined how the outcome variables are coded 1 or 0.

• I would also focus on fixing grammatical errors and awkward sentences.

Comment well taken and addressed.

• Lastly, I would suggest expanding the discussion of methodological limitations a little bit (i.e. potential issues of endogeneity).

Thanks for the valid comment. The following sentences are included in the limitation section of the revised manuscript. “Another limitation of this study would be the presence of unmeasured confounders that would be correlated both with the outcome and predictor variables included in the model, particularly level-2 endogeneity arising from correlations between included individual characteristics and omitted community-level variables and/ or reverse causality of variables that would cause the model to be endogenous [52]. To minimize the potential endogeneity problems, we excluded predictor variables with the unclear direction of the effect with the outcome variables, for instance, “Complete ANC” was excluded from Model I of predicting ANC4+ uptake.” Line 401-8; page 27

Reviewer #2: the authors investigated the predictors of maternal and newborn health service utilization across the continuum of care in Ethiopia. the authors did a great work and have added to the body of knowledge on maternal health.

Thanks!!

the authors should edit their writeup and make sure where ever they use an abbreviation for the first time, they first indicate it in full and the abbreviation in bracket.

Comment is well taken and addressed in this version.

Submitted filename: Point-by-point response letter.docx

Click here for additional data file.

6 Dec 2021

24 Jan 2022

A point-by-point response to reviewer/editorial

Journal: PLOS ONE

PONE-D-20-37879R1

Predictors of maternal and newborn health service utilization across the continuum of care in Ethiopia: a multilevel analysis

The authors would like to appreciate and thank the reviewers and editors for the constructive comments. Our point-by-point responses are below each of the comments in italics.

Journal Requirements:

Please review your reference list to ensure that it is complete and correct. If you have cited papers that have been retracted, please include the rationale for doing so in the manuscript text, or remove these references and replace them with relevant current references. Any changes to the reference list should be mentioned in the rebuttal letter that accompanies your revised manuscript. If you need to cite a retracted article, indicate the article’s retracted status in the References list and also include a citation and full reference for the retraction notice.

Thank you for the valid comments. References cited are now edited. Articles under review

Additional Editor Comments (if provided):

This is an interesting work. They choose 3 indicators and analyse them as a cascade of a continuum of care (CoC) antepartum (4 antenatal visits), intrapartum (skilled birth attendance conditional the woman had the 4 antenatal visits), and postpartum (postnatal care among those who completed the intrapartum step as defined here). Then they present prevalence of these indicators per characteristics and elements chosen from Anderson’s health-seeking behaviour model.

Thank you so much.

However, few issues:

1. Please add some description of the administrative division of Ethiopia. At least the reader will understand what is a woreda.

Comment well taken. A paragraph is added under the Methods section to describe the administrative division of the country. (Line 79-81, page 5)

2. Please provide 95% confidence intervals (CI) for each step. That could be accomplished with a bar plot for the antepartum, intrapartum and postpartum. Please, make sure that those CI are adjusted for the clustering.

Comment well taken. The continuum of care coverage across antepartum, intrapartum and postpartum with error bars is presented in Figure 1, separately attached.

a. Tables 3 and 4 could have those 95% confidence intervals rather than the p-values. Again, make sure those CI are adjusted for the cluster sampling.

Thank you for the valid comments. The 95% CI adjusted for the cluster sampling is included in both Tables 3 and 4.

3. Table 3 - for education, I believe complete continuum figures are a repetition of the “continued at antenatal care”.

Thanks a lot for comment. You are right. It was an error. It is now fixed.

4. Table 5 - Please add a row with the number of observations included in each model

Thanks. Number of observations in each model is included in Table 5.

Comments to the Author

Reviewer #3: As stated by a previous reviewer, the entire manuscript needs copy editing. Specific issues include:

Comment well taken and copy edited.

Line 47-48: This first sentence is confusing and frankly, not necessary.

Comment well taken and edited accordingly.

Line 50-51: I believe a modifier may be missing here, such as “level of service delivery needed to prevent…”

Comment well taken and edited accordingly.

Line 63: What are “supply-side” factors? Number of physicians/nurses/facilities?

Comment well taken edited accordingly.

Line 68: Needs a reference

Comment well taken.

Line 124-128: The direction of “no difficulty of distance and transportation arrangement to access medical care” as an “exposure” variable does not make sense given the direction of the other variables; please explain. Additionally, II am not sure that “exposure” is the correct term. Predictor is more appropriate. You were not “exposing” women to these factors with an intervention.

Thanks for the comment. We checked the adequacy of our sample size using a double population, using case and control or exposure and non-exposed groups to determine the predictor variables of the study.

Line 134-136: At line 135, “four or more antental care (ANC4+)” appears in item 2 but should appear in item 1, as that is its first use.

Comment well taken and addressed.

Line 137: What is PNC? (Spell it out before abbreviating it.)

Comment well taken and spelt out.

Line 149: I am surprised you did not include the infant’s birth weight, estimated gestational age at birth, health at birth (did the infant need to stay in a neonatal special care unit, did the infant have a congenital anomaly), or survival. These factors may influence receipt of care at the delivery and postnatal stages. Were these items asked?

Thanks for the valid comments. Variable infant’s birth weight was collected and included in the analysis; but not significantly associated with COC at postpartum stage. And this is indicated in the revised version of the manuscript. However, we did not collect variables like estimated gestational age at birth, health at birth and/or presence of a congenital anomaly.

Table 3, 4: What are these proportions? Neither columns nor rows sum to 100% in any group. Also, what is the denominator for the N’s?

Thank you for the comment. The denominator for Tables 3 and 4 are presented as “Total’’ column at the end of each table. The proportions presented in these tables are actual percentages of each category; their complements are not presented which we believe readers can get by subtracting from 100%.

Submitted filename: Response to Reviewers.docx

Click here for additional data file.

15 Feb 2022

Predictors of maternal and newborn health service utilization across the continuum of care in Ethiopia: a multilevel analysis

PONE-D-20-37879R2

Dear Dr. Tiruneh,

We’re pleased to inform you that your manuscript has been judged scientifically suitable for publication and will be formally accepted for publication once it meets all outstanding technical requirements.

Within one week, you’ll receive an e-mail detailing the required amendments. When these have been addressed, you’ll receive a formal acceptance letter and your manuscript will be scheduled for publication.

An invoice for payment will follow shortly after the formal acceptance. To ensure an efficient process, please log into Editorial Manager at http://www.editorialmanager.com/pone/, click the 'Update My Information' link at the top of the page, and double check that your user information is up-to-date. If you have any billing related questions, please contact our Author Billing department directly at authorbilling@plos.org.

If your institution or institutions have a press office, please notify them about your upcoming paper to help maximize its impact. If they’ll be preparing press materials, please inform our press team as soon as possible -- no later than 48 hours after receiving the formal acceptance. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information, please contact onepress@plos.org.

Kind regards,

Orvalho Augusto, MD, MPH

Academic Editor

PLOS ONE

Additional Editor Comments (optional):

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. If the authors have adequately addressed your comments raised in a previous round of review and you feel that this manuscript is now acceptable for publication, you may indicate that here to bypass the “Comments to the Author” section, enter your conflict of interest statement in the “Confidential to Editor” section, and submit your "Accept" recommendation.

Reviewer #3: (No Response)

**********

2. Is the manuscript technically sound, and do the data support the conclusions?

The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented.

Reviewer #3: Yes

**********

3. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #3: Yes

**********

4. Have the authors made all data underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #3: Yes

**********

5. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

Reviewer #3: Yes

**********

6. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #3: I was not clear about my previous concern about lines 126-127.

The "exposure" variables that you list indicate problems - lack of autonomy, no media exposure, no maternal education, poorest wealth quintile. You also include "no difficulty of distance to access medical care" and "no difficulty of transport arrangement to access medical care." I believe that, given the direction of the other variables, that you mean "difficulty of distance and transport arrangement to access medical care".

If I am correct, at lines 128 and 129, you will want to remove "having difficulty of distance and transport arrangement to access medical care" and replace it with "no difficulty of distance or transport arrangement to access medical care."

Otherwise, thank you for the excellent revision.

**********

7. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #3: No

17 Feb 2022

PONE-D-20-37879R2

Predictors of maternal and newborn health service utilization across the continuum of care in Ethiopia: a multilevel analysis

Dear Dr. Tiruneh:

I'm pleased to inform you that your manuscript has been deemed suitable for publication in PLOS ONE. Congratulations! Your manuscript is now with our production department.

If your institution or institutions have a press office, please let them know about your upcoming paper now to help maximize its impact. If they'll be preparing press materials, please inform our press team within the next 48 hours. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information please contact onepress@plos.org.

If we can help with anything else, please email us at plosone@plos.org.

Thank you for submitting your work to PLOS ONE and supporting open access.

Kind regards,

PLOS ONE Editorial Office Staff

on behalf of

Dr. Orvalho Augusto

Academic Editor

PLOS ONE