Using Andersen's behavioral model of health care utilization in a decentralized program to examine the use of antenatal care in rural western Ethiopia.

BACKGROUND: In Ethiopia, most women do not make the minimum number of antenatal care (ANC) visits recommended by WHO. This study modeled predisposing, enabling, need, and external environmental factors in the utilization of decentralized health facilities for ANC services in rural western Ethiopian communities.
METHODS: A community-based, cross-sectional study was conducted in Gida Ayana Woreda (District) among 454 women. Data were collected through structured questionnaires. Multinomial logistic regression was used to model the association between the explanatory variables and the use of recommended and fewer than recommended visits for ANC with reference to the base model, no ANC visits.
RESULTS: Only 15.2% of women made the recommended minimum number of ANC visits. Women with fewer than 2 children (AOR 10.7; 95% CI 3.0-8.4) were 10.7 times more likely received ANC service as recommended. Women with a delivery of 2 or more (AOR 9.7; 95% CI 3.7-5.2) home visits by health extension workers (HEWS) were 9.7 times more likely receiving minimum ANC services. Involvement in gainful activities had 4 times higher log odds of seeking recommended ANC (AOR 4.0; 95% CI 1.4-11.7). Women who experienced high fever were more likely to obtain the recommended ANC services (AOR 7.1; 95% CI 2.9-7.5). Residents of Ayana Kebele decentralization entity were 60% more likely to make the recommended number of visits to ANC (AOR 24.6; 95% CI 4.8-15.2).
CONCLUSIONS: Number of children, home visits, gainful activities, monthly income, high fever, and decentralized administrative kebele were strongly linked with recommended ANC schedule. The need for a program intervention aimed at meeting WHO recommendations for ANC visits include economizing birth size and spacing; improving home attendance by HEWs, knowledge of pregnancy complications and benefits of minimum ANC visits, local socio-economic development measures targeting poor women/households; further decentralization of health system improving proximity to ANC in rural western Ethiopia.

Introduction

About 303,000 maternal deaths were reported worldwide in 2015. Of these, 99% were in the developing world, making the maternal mortality rate (MMR) in that region 239 per 100,000 live births (LBs), which was 20 times higher than in industrialized countries. Sub-Saharan African women accounted for roughly 66% (201 000) of the global maternal deaths and had the highest MMR, 546 deaths per 100,000 LBs in 2015 [1].

In Ethiopia, 13 017 maternal deaths were reported in 2015 [2]. The global burden of disease studies of 2013 and 2015 revealed MMR of 497 and 410 per 100,000 LBs, respectively, showing no significant change between the two studies [2,3]. The prevention of maternal mortality is a priority for the World Health Organization; the UN Sustainable Development Goal (SDG) and the Ethiopian government [4,5]. In the SDG period, the target is to reduce the global MMR to less than 70 per 100,000 LBs by 2030 with no country having MMR more than 140 per 100,000 LBs [5]. In 2015, the Health Sector Transformation Plan (HSTP) of Ethiopia targets for improving maternal health is to reduce MMR from 420 per 1000,000 LBs in 2015 to 199 per 100,000 in 2020 [4]. Achiveing this target by the year 2020 will also enable the country to reach her SDG3’s promise of less than 140 MMR per 100,000 LBs in 2030. However, the countrywide 2016 Demographic Health Survey (EDHS) documented a MMR of 412 [6], far short of the HSTP and SDG3 targets [1,4]. Most maternal deaths (90%) are avoidable with timely interventions [7,8].

Studies carried out elsewhere have found that simple ANC interventions such as monitoring blood pressure and body weight, giving vaccinations, and providing counseling on pregnancy and danger signs are highly effective preventive measures [7,9-12]. Non-utilization of local ANC programs may help explain the persisting high rates of pregnancy complications in Ethiopia [9,13,14]. Despite the Ethiopian government’s efforts to improve maternal health and bring facilities closer to mothers through decentralization programs implemented in the early 1990s [15,16,17], a recommended minimum number of ANC utilization remains low [18-20]. The 2016 EDHS, which covered all the regions of Ethiopia, found that 49.3% of women in the most populous administrative region, Oromia, did not receive ANC, and results were similar for Somali Region [6]. This statistics suggests a gap in understanding of the impact of multifaceted factors on the utilization of the minimum number of ANC visits, particularly in Oromia Region, where antenatal outcomes are poor compared to national figures [4]. Preventable maternal health risks may be managed with early detection [7,10,13,21,22]. This study examined antenatal service utilization behavior of women in a remote rural area in western Ethiopia.

Various studies have reported that ANC utilization is driven by factors such as awareness among service users and the wider communities, knowledge of maternal pregnancy and risks, community customs, previous facility use, parity and pregnancy complications, individual attitudes and health-care-seeking behaviors, household income, occupation, decision power, home visit and availability and accessibility of health facilities [13,23-26]. However, the determinants of utilization of ANC vary across different geographical locations and contexts, different cultures and beliefs, and socio-economic and demographic settings [10,12,19,25,27,28]. There is little information about women’s utilization of decentralized health facilities for ANC and underlying factors in Oromo culture which has its own peculiar geographic, socio-economic, and cultural characteristics that may affect the utilization of the minimum number of ANC services. We adapted the behavioural model framework of Andersen [26] for use of health services to identify the factors potentially facilitate or impede minimum number of antenatal health services seeking behavior at individuals and community levels [13,22,29]. The model groups and predicts that a series of factors predisposing, enabling, and need and external factors influence the utilization of health services. Predisposing factors are socio-demographics characteristics; enabling factors facilitates individuals to use services such as availability of resources such as income, access to free services, availability and access to the service; need factors are physical conditions illness or disease conditions that motivate service use [26]. ANC outcomes need to be modeled as functions of these factors [13,22,29,30]. Studies employing the model in Ethiopia have used secondary sources; thus there is a need for studies using primary data [20,23,29]. Hence, the objective of this study was to investigate the above domains of determinants influencing the use of decentralized health facilities for the recommended number of ANC services with the aim of informing policy makers and practitioners responsible for planning, administering, and delivering maternal health service programs.

Methodology

Study setting

The study was conducted in Gida Ayana Woreda [15], Oromia Region, rural western Ethiopia, about 450 km from Addis Ababa, and 112 km from Nekemte, the capital of Eastern Wollega Zone. The area of the woreda is about 1,502 square km and organized into 7 urban and 21 rural kebeles (the smallest administrative units in Ethiopia). According to the 2013 population projection release and Oromia Regional State, the woreda had a total population of 140,484, including 47,040 child-bearing women and 10, 577 women of reproductive age, 15–49 years [31,32]. The woreda has 1 primary hospital, 5 health centers, and 28 health posts [32]. All promotive and preventive health servies and basic essential obstetric care are provided in the health center and health posts where as comprehensive essential obstetric care is provided in the primary hospital [33]. There were also 8 private drug shops, 3 private drug venders and 1 clinic under NGO ownership [32].

Study design and period

A community-based, cross-sectional design was used in this study, from November 2016 to January 2017.

Sample and recruitment

The sample size was determined using a single population proportion formula. Following a previous study [34], a proportion of 32.7%, a 95% confidence interval (CI), a margin of error of 5%, and with 2 design effect (since the selection was conducted in two stages: at kebele and household level) of 1.5 were used. Thus, a minimum adequate sample size was determined using the statistical estimation method [35]. Since the source population was estimated less than 10,000; sample size correction was performed. Then, 5% non-response rate was added to obtain the adequate sample size of 459.

Two-stage sampling using a simple random sampling technique was employed to select an appropriate representative study population. In the first stage, four representative kebeles were randomly selected using the lottery technique. In the four selected kebeles, women who reported to have had their last birth during the 5 years prior to the study were identified with the help of female HEWs and women team leaders. In the second stage, eligible women were sampled using Microsoft Office Excel-generated random numbers proportional to the estimated number of women who had given birth in the respective kebeles during those 5 years. When two women were living in the same household, recent births were considered in determining whom to interview. Where study mothers were not available during the survey, they were visited again the next day and, if not available then, they were considered to be non-responders. Inclusion criteria of our study were women aged 15–49 years who gave birth to children during the 5 years prior to the survey. Women who reside in the respective kebele for a minimum of 8 months prior to data collection were not considered in this survey.

Outcome

ANC is defined here as at least one visit to a doctor, nurse, midwife, or trained traditional birth attendant during pregnancy [36]. The nominal dependent variable of the study was the number of visits to ANC service clinics. ANC users were categorized into three groups according to the WHO recommendations for ANC visits [37], irrespective of when in the course of the pregnancy the visits occurred; the groups were as follows: the logit or the log-odds of having y = 2 [those who made the recommended 4 or more ANC visits], y = 1 [those who made 1 to 3 ANC visits, fewer than the recommended number]; and y = 0 [no ANC visits]. No ANC visit was a base model for the first two modeled categories.

Explanatory variables

In the present study, based on Andersen’s behavioral model (S1 Fig) of health care utilization theory [25,26,38], age as a three-categorical variable (19 or less, 20–34, 35–48 years), mother’s and father’s education (college or higher, secondary, primary, no education), marital status (married or other), mother’s religion (Christian or Muslim), and decision making regarding use of household resources (husband only, husband/wife, wife only) were examined as predisposing determinant factors of adherence to the recommended number of visits to ANC clinics.

Enabling factors were mother’s employment (skilled employee, small business/service, farmer, housewife), husband’s employment (skilled employee, merchant, farmer, other), home visits HEWs (more than one, one, none), distance to ANC (under 30 minutes, 30 minutes or more), household income (less than 50 $US, 50 $US or more), and possession of a radio/TV (yes or no).

Need factors were severe headache, vaginal bleeding/gush, swelling of hands/face, high fever, severe pain in the abdomen, high blood pressure, and blurred vision, each classified as a binary variable (yes or no). As external environmental factors, we examined the administrative kebeles where the women lived and in which health facilities were decentralized as well as whether the women lived in urban or rural locations. Using these external factors enabled us to explore differences in ANC utilization across space and to consider the available decentralized health facility type (health post, health center, hospital/clinic) as an enabling factor (Fig 1).

Fig 1

Conceptual framework for health-service utilization behavior modified from Andersen’s behavioral model.

Data collection and quality control

Data were collected using a structured questionnaire. The questionnaire was designed in English (S1 Text) and translated to Afan Oromo by staff of the English and Local Language departments of Wollega University. Data collection focused on institutional delivery; place of delivery; status of ANC visit and predisposing, enabling, perceived needs and external environmental factors affecting ANC use. Data were collected by 8 HEWs with BSc degrees or diplomas in health science who had previous experience in data collection in Gida Ayana Woreda. Data collection was supervised by two supervisors recruited from a local health center and coordinated by the corresponding author. The data collectors and supervisors were trained for two days in data collection techniques and ethics. A pilot study of 10% of the study population was carried out to test the survey instrument in an adjacent woreda (or Guto Gida) to ensure reliability, to check for clarity and comprehension. The supervisors and the principal investigator supervised and monitored data collection activities and checked all the complete questionnaires for consistency and missing data daily. Incomplete questionnaires deemed to have problems were returned to the interviewers for completion. The questionnaire was pretested for construct validity with a 10% sample in a nearby woreda and modified. Each completed questionnaire was checked daily to ascertain that all the questions were correctly answered to address data validity and reliability.

Data processing and analysis

We used statistical software EpiData Version 3.1 for preliminary data preparation and statistical software SPSS Version 24.0 for data analysis. Descriptive statistics were used to calculate the frequency distribution and proportions for categorical variables. For the normally distributed continuous variables, mean with SD was also used. The Variance Inflation Factor (VIF) > 10 indicates redundancy among explanatory variables [39,40]. Our ANC utilization model satisfied this criterion with VIF < 2.0. Associations between the number of ANC service visits in the three groups (those receiving the recommended number of ANC visits, those receiving fewer than recommended, and those receiving no ANC) and explanatory variables were calculated by the use of the binomial and multinomial logit (MNL) model.

The associated factors were examined using chi-square test and multivariable logistic regression analysis. All the significant variables in the bivariate analysis (p < 0.05) were included in the the final multinomial logistic (MNL) regression model because bivariate association between two variables does not necessarily imply a significant causal relationship between them. Therefore, a multivariate approach was applied to determine which factors best explain and predict health care service use outcome. The adequacy of the developed model was verified through the standard statistical mean of likelihood ratio test of goodness of fit [40]. Multicollinearity in the MNL model was detected by examining the standard error for the coefficients [11]. Adjusted odds ratios (AOR) with corresponding 95% CI estimates were used to describe the strength of associations of factors with recommended number of ANC visits and fewer than the recommended number of ANC visits versus no visits. The association of variables was found to be statistically significant at p ˂ 0.05.

Ethical consideration

The research protocol was reviewed and approved by Wollega University Research Ethics Approval Committee [Ref/No.WU–99529/H1-3/24/11/2016]. Permission was received from Gida Ayana Woreda Health Office. The purpose of the study was explained to all participants and a consent form approved by the Review Board was given to participants. Parents or legal guardians provided consent for all participants under age 18. We emphasized that participation was completely voluntary and that they had the right to withdraw any time during the interview without giving any reason. Confidentiality and anonymity were explained to all participants. We ensured that all participants understood the information given by asking them. The consent form was read aloud for women who could not read or write. Literate women were asked to read the consent form. A written consent in the form of a signature or a thumbprint was obtained from all of the participants.

Results

Socio-demographic characteristics and nature of home visit

A total of 459 women who had their last birth during the 5 years preceding the survey were enrolled in the study. The response rate was 98.9%. The mean age of mothers was 26.1 (±7.1) years. The mean number of children women gave birth to was 3.11 (±2.0). Nearly half of the mothers reportedly had no formal education. Over one-third of the husbands had completed secondary education and 33.9% completed college/higher education. Forty-nine percent of the respondents were of the Oromo ethnic group, 83.9% were married, and 55.9% were urban. Over half of mothers were Christian. One hundred ninety-eight (43.6%) were housewives and 124 (27.5%) were paid employees. Most husbands (46.7%) were subsistence farmers. Mean walking time it takes pregnant women to reach the nearest health facility was 51.1(±31) minutes, and the mean monthly household income was 47 (±15.1) $US. The majority of participants (56.2%) reported that they did not obtain home visit and support by HEWs (Table 1).

Table 1

Socio-demographic backgrounds of the study participants and home visits, in Gida Ayana Woreda, rural western Ethiopia, during Nov. 2016 to Jan 2017 (N = 454).

Variable	Variable categories	Number(n)	Percentage (%)
Age (in years)	19 or less	127	28.0
	20–34	255	56.1
	35 or more	72	15.9
Education	No formal education	224	49.3
	Primary	81	17.8
	Secondary	73	16.1
	College/higher	76	16.7
Husband’s education	No formal education	90	19.8
	Primary	49	10.8
	Secondary	161	35.5
	College/higher	154	33.9
Occupation	Paid employee	124	27.5
	Small business/service	98	21.6
	Farmer	34	7.3
	Housewife	198	43.6
Husband’s occupation	Skilled employee	86	18.9
	Merchant	114	25.1
	aInformal activity	42	9.3
	Farmer	212	46.7
Marital status	Married	381	83.9
	bOther	73	16.1
Ethnicity	Oromo	222	48.9
	Amhara	144	31.7
	Tigre	88	19.4
Religion	Christian	260	57.3
	Muslim	194	42.7
Urban-rural location of the women	Urban	254	55.9
	Rural	200	44.1
Mean number of births (SD)	3.11(±1.9)
Mean walking time to ANC clinic in min (SD)	51.1(±30.9)
Mean monthly family income in $US (SD)	c47.0(±15.1)
Home visit by HEWs	2 times or more	132	29.0
	One time	67	14.8
	No	255	56.2

aInformal activity: day laborer/weaving/students.

bOther: single/widowed/divorced. SD: Standard Deviation.

cAverage exchange rate of 1$US was 21.43 Ethiopian Birr between November 2016-January 2017.

ANC service use and decentralized facility attended

The majority (55.1%) of the women obtained ANC from a decentralized health center facility. The women who made the recommended number of visits to ANC constituted 15.2% of the participants while 49.6% made fewer than the recommended number and 35.2% did not seek services. We found that 89.4% and 66.2% of respondents of Ayana and Ejere Kebeles, respectively, visited the ANC clinic, and lower proportions did so in Angar and Lalistu Kebeles. Of the women who utilized ANC services, 41.5% made their first visit in the second trimester of their pregnancy. Among those who made no ANC visit, 46.2%, 25.0%, and 11.2% mentioned lack of awareness about the importance of pregnancy care, transportation problems, and long waiting times, respectively, as reasons for not using ANC services (Table 2).

Table 2

Utilization of ANC and local health facilities attended, in Gida Ayana Woreda, rural western Ethiopia, Nov. 2016-Jan. 2017 (N = 454).

Variable	Variable categories	Number(n)	Percentage(%)
ANC visits	Recommended # ANC visits	69	15.2
	Fewer than recommended visits	225	49.6
	No visits	160	35.2
Decentralized facility visited	Hospital/clinic	25	8.5
	Health center	162	55.1
	Health post	107	36.4
Decentralized administrative kebele by ANC visits	Ayana	84	89.4
	Ejere	51	66.2
	Angar	92	57.5
	Lalistu	67	54.5
Timing of 1st ANC visit	1st trimester	96	32.6
	2nd trimester	122	41.5
	3rd or 4th trimesters	76	25.9
cReason for no ANC visit	Lack of awareness	74	46.25
	Transportation problem	40	25.0
	Waiting time	18	11.25
	Illness was not severe	14	8.75
	Heavy workload	10	6.25
	Others	4	2.5
	Total	160	100

cmultiple responses were possible.

Factors influencing visits to ANC in a decentralized facility

The results of bivariate analyses of number of ANC visits and the independent variables showed that walking distance to neaby facility, sever headache, vaginal bleeding and rural urban residence appeared to be positively associated with minimum number of recommended ANC received (p < 0.05). Maternal age at last pregnancy, mother’s and husband’s education, marital status, religion, number of births, decision on family resources, type of decentralized facility, monthly income, occupation, home visit by HEWs, availability of radio/orTV, high fevere, swelling of hands/face, abdominal pain and Kebele in which decentralized health facilities were located were also significantly associated (p < 0.01) with the recommended schedule of ANC visits. No statistically significant associations were found for high blood pressure and blurred vision assessed (p > 0.05).

In the following sections, we present the influence of each determinant factor on recommended number of ANC visits and fewer than the recommended ANC visits versus no ANC visits as established through the multinomial regression analysis and shown in Table 3.

Table 3

Multinomial regression analysis for factors influencing number of women’s ANC visits, in Gida Ayana Woreda, rural western Ethiopia, Nov. 2016-Jan. 2017 (N = 454).

			Status of ANC visits. Base model: No-visits category
			Recommended visits		Fewer than recommended visits
Factor	Variable	Variable categories	AOR (95% CI)	p-value	AOR (95% CI)	p-value
Predisposing factors	Religion	Christian	1.6(0.6–3.7)	0.279	3.3(1.7–6.5)**	0.001
		Muslims	1		1
	Marital status	Married	0.5(0.1–1.5)	0.253	2.3 (0.9–5.8)	0.06
		others	1		1
	No.of children	Fewer than 2	10.7(3.0–8.4)**	0.001	9.2(3.6–23.0)**	0.001
		2 to 3	5.5(1.5–2.4)*	0.01	4.6(1.8–11.5)**	0.001
		4 or more	1		1
	Age	19 or less	1.7(0.4–6.5)	0.402	1.0(0.4–2.7)	0.906
		20–34	0.6(0.2–2.3)	0.557	0.9(0.3–2.1)	0.814
		35 or more	1		1
	Education	College/higher	1.1(0.2–5.1)	0.902	0.7(0.2–2.7)	0.719
		Secondary	0.8(0.2–2.7)	0.743	0.6(0.2–1.7)	0.395
		Primary	1.7(0.5–6.0)	0.373	2.1(0.8–5.1)	0.092
		No education	1		1
	Husband’s education	College/higher	2.4(0.4–12.7)	0.299	0.8(0.2–3.1)	0.771
		Secondary	1.7(0.3–7.7)	0.468	1.3(0.4–4.1)	0.64
		Primary	1.2(0.4–3.9)	0.662	0.9(0.4–2.0)	0.878
		No education	1		1
	Decision on family resource	Husband	1.5(0.6–3.9)	0.334	1.9(0.9–4.0)	0.075
		Wife/husband	1.6(0.6–4.6)	0.316	3.9(1.8–8.4)**	0.001
		Wife	1		1
Enabling factors	Employment	Employee	2.1(0.7–5.8)	0.133	1.4(0.6–2.9)	0.365
		Small business/service	4.0(1.4–11.7)**	0.009	2.2(1.0–4.8)*	0.05
		Farmer	2.3(0.4–13.4)	0.349	1.8(0.3–9.0)	0.424
		Housewife	1
	Husband’s employment	Employee	1.5(0.2–8.1)	0.623	2.0(0.5–7.7)	0.291
		Merchant	1.0(0.3–3.1)	0.972	1.5 (0.6–3.5)	0.341
		Other	0.8 (0.1–4.0)	0.885	1.3(0.4–3.9)	0.603
		Farmer	1		1
	Family income	50 $US or more	2.8(1.2–6.2)*	0.011	2.1(1.1–3.8)*	0.013
		Less than 50 $US	1		1
	Home visit by HEWs	More than one time	9.7(3.7–5.2)**	0.001	4.2(1.9–8.9)**	0.0001
		One time	9.5(2.9–3.7)**	0.001	4.8(1.8–13.0)**	0.001
		No visit	1		1
	Walking time to ANC clinic	Less than 30 minutes	1.7(0.7–4.4)	0.226	2.4(1.2–5.0)*	0.013
		30 minutes or more	1		1
	Possession of radio/or TV	Yes	1.7(0.7–4.1)	0.224	1.7(0.8–3.3)	0.108
		No	1		1
	Decentralized facilities	Hospital/clinic	1.1(0.2–4.6)	0.889	0.4(0.1–1.6)	0.239
		Health center	2.3(0.9–5.8)	0.08	2.2(1.1–4.5)*	0.019
		Health post	1		1
Need factors	Severe headache	Yes	2.9(1.1–7.5)*	0.026	3.7(1.8–7.6)**	0.001
		No	1		1
	Vaginal bleeding/gush	Yes	1.7(0.6–4.6)	0.239	2.2(1.1–4.8)*	0.036
		No	1		1
	Swelling of hands/face	Yes	1.2(0.5–2.9)	0.601	0.9(0.5–1.8)	0.99
		No	1		1
	High fever	Yes	7.1(2.9–7.5)**	0.001	4.1(1.9–8.5)**	0.001
		No	1		1
	Severe pain in abdomen	Yes	0.8(0.3–2.4)	0.794	1.1(0.5–2.5)	0.676
		No	1		1
External Environmental factors	Kebeles in which decentralized health facilities were located	Ayana	24.6(4.8–15.2)**	0.001	8.2(2.1–3.5)**	0.002
		Angar	2.0(0.3–11.3)	0.398	0.9(0.2–3.3)	0.951
		Lalistu	1.2(0.3–5.1)	0.742	0.8(0.3–2.5)	0.806
		Ejere	1		1
	Urban-rural residence	Urban	2.1(1.1–3.7)**	0.013	1.4(0.9–2.1)	0.8
		Rural	1		1

1 = Reference category.

** significant at p < 0.01

*significant at p < 0.05

Predisposing factors

Holding other variables constant, Christian women were 3.3 times more likely than Muslims to make 3 or fewer ANC visits than no visits. Nevertheless, identification as Christian was not significantly associated with utilization of the recommended number of ANC visits although the corresponding log odds figure was higher (AOR = 1.6; 95% CI 0.6–3.7, p > 0.05).

Women who had fewer than 2 children were 10.7 times and 9.2 times more likely to make the recommended number of ANC visits (AOR = 10.7; 95% CI 3.0–8.4, p ˂ 0.01) and fewer than the recommended number (AOR = 9.2; 95% CI 3.6–23.0, p ˂ 0.01), respectively, than no visits compared to women who reported 4 children or more. Likewise, women with 2 to 3 children had log odds of 5.5 times and 4.6 times higher of making the recommended number of ANC visits (AOR = 5.5; 95% CI 1.5–2.4, p ˂ 0.05) and fewer than the recommended ANC visits (AOR = 4.6; 95% CI 1.8–11.5, p ˂ 0.01), respectively, than no visits compared to women who had 4 or more children.

Gendered decision making about resource use by husband (AOR = 1.5; 95% CI 0.6–3.9, p > 0.05) and wife/husband (AOR = 1.6; 95% CI 0.6–4.6, p > 0.05) influenced women’s maternal health service seeking behavior in choosing the recommended number of ANC visits rather than no visits, but the association was not statistically significant. However, when all other factors were held constant, shared decision making (wife/husband) about household resource use was 3.9 times more likely to result in fewer than the recommended ANC clinic visits than no visits compared to wife-only decision making (AOR = 3.9; 95% CI 1.8–8.4, p ˂ 0.01).

Enabling factors

When all other determining factors were held constant, women who operated small businesses had 4.0 times (AOR = 4.0; 95% CI 1.4–11.7, p ˂ 0.01). and 2.2 times (AOR = 2.2; 95% CI 1.0–4.8, p ˂ 0.01) higher log odds of choosing recommended and fewer than the recommended number of ANC visits, respectively, than no visits compared to women who identified as housewives.

Women in households with monthly household income of 50 $US or more were nearly 3 times and 2 times more likely to make the recommended (AOR = 2.8; 95% CI 1.2–6.2, p < 0 05) and fewer than recommended number of ANC visits (AOR = 2.1; 95% CI 1.1–3.8, p < 0.05), respectively, than no visits compared to those reporting less than 50 $US in household income.

Women who were visited 2 times or more than 2 times by HEWs were 9.7 times and 4.2 times, respectively, more likely to make the recommended (AOR = 9.7; 95% CI 3.7–5.2, p < 0.01) and fewer than recommended number of ANC visits (AOR = 4.2; 95% CI 1.9–8.9, p < 0.01) than no visits compared to those reporting no visits by HEWs. Likewise, women who were visited one time had higher log odds of making the recommended (AOR = 9.5; 95% CI 2.9–3.7, p < 0.01) and fewer than recommended (AOR = 4.8; 95% CI 1.8–13.0, p < 0.01) number of ANC visits by 50% and 80%, respectively, than no visits compared to those who were not visited by HEWs.

Walking time strongly influenced the utilization of health facilities, specifically the number of ANC visits, but the relationship was not statistically significant (AOR = 1.7; 95% CI 0.7–4.4, p > 0.05). Furthermore, the log odds ratio of receiving fewer than 4 ANC visits versus no visits was 40% higher for women living closer than 30 minutes from the nearest ANC clinics (AOR = 2.4; 95% CI 1.2–5.0, p < 0.05).

When comparing decentralized health facility types of hospital/clinic (AOR = 1.1; 95% CI 0.2–4.6, p > 0.05). and health center (AOR = 2.3; 95% CI 0.9–5.8, p > 0.05), no significant association was found with the utilization of the recommended number of ANC visits versus no visits. For women who received services at decentralized front-line health posts, the log odds of making fewer than the recommended number of ANC visits versus no visits to health centers was 20.0% higher (AOR = 2.2; 95% CI 1.1–4.5, p < 0.05).

Need factors

When all other determining factors were held constant, women who reportedly felt severe headaches were found to be 90% (AOR = 2.9; 95% CI 1.1–7.5, p < 0.05) and 70% (AOR = 3.7; 95% CI 1.8–7.6, p < 0.01) more likely to make the recommended and fewer than recommended number of ANC visits, respectively, than those not reporting headaches. The log odds ratio for making fewer than the recommended number of ANC visits was 20% higher (AOR = 2.2; 95% CI 1.1–4.8, p < 0.05) for women who experienced vaginal bleeding/gush compared to those who did not. Likewise, the odds of women who had high fevers were 7.1 and 4.1 times higher for having the recommended (AOR = 7.1; 95% CI 2.9–7.5, p < 0.01) and fewer than recommended (AOR = 4.1; 95% CI 1.9–8.5, p < 0.01) number of visits to ANC facilities, respectively, than those who felt no fever.

External environmental factors

When comparing the utilization of ANC services in the four study kebeles in which government decentralized health facilities were located, women of Ayana had higher odds of making the recommended number of visits (AOR = 24.6; 95% CI 4.8–15.2, p < 0.01) and fewer than the recommended number (AOR = 8.2; 95% CI 2.1–3.5, p < 0.01) than making no visits compared to women of Ejere Kebele. Furthermore, when all other factors were held constant, residents of urban settlements were 2.1 times more likely to make the recommended number of ANC service visits (AOR = 2.1; 95% CI 1.1–3.7, p < 0.01) compared to rural residents. However, the urban-rural difference did not seem to significantly influence the choice of fewer than the recommended number of ANC service visits in spite of higher log odds(OR = 1.4; 95% CI 0.9–2.1, p > 0.05).

Discussion

Although increasing the number of ANC visits has contributed to a drastic reduction in the maternal death rate in low-income countries during the past 30 years, the majority of women in sub-Saharan Africa, including Ethiopia, still do not make the WHO-recommended 4 ANC visits or more during the pregnancy period [21]. This study found that only 15.2% of the 454 participants received the recommended number of ANC visits; 49.6% made fewer than the recommended ANC visits and 35.2% reported no visits for ANC.

Our findings identified a number of predisposing, enabling, need, and environmental factors influencing the choices regarding ANC visits: religion, number of children, woman’s occupation, home visit by HEWs, walking time to health facility, monthly income, severe headache, vaginal bleeding/guish, high fever, availability of decentralized ANC facilities, decentralized administrative kebele, and urban-vs-rural residence of the women.

After adjusting for all variables, Christian women were found to be 30% more likely to make fewer than the recommended minimum number of ANC visits than no visits compared to Muslim women although religion was not significantly associated with making the recommended number of ANC visits. This finding is consistent with a study in northeastern Ethiopia, where Christians were 2.2 times more likely to make fewer than the recommended number of ANC visits compared to Muslim women [41]. In Nigeria, Christian women were more likely to make the recommended number of ANC visits than fewer than the recommended number than Muslims [42]. The higher level of ANC use among women of certain religions could be attributed to theological differences and differences in lifestyle across various beliefs [12]. In Nepal, Christians and Hindus were 50% and 30%, respectively, more likely to make the recommended and fewer than recommended number of ANC visits versus no visits compared to women of other religions [11].

Women experiencing their first pregnancy and those who had 2 or 3 children were 10.7 and 5.0 times more likely to make the recommended minimum number of ANC visits, respectively, than no visits compared to baseline. Recent systematic reviews and meta-analyses reveal that women with first pregnancies are more likely than multiparous women to make the recommended number of ANC visits due to fear of complications with the first birth [21,22,29]. Multiparous women tend to believe there is less risk to current pregnancy due to their previous birthing experiences and their negative perceptions of the environment in local health institutions regarding cleanliness, equipment quality, and behavior of providers [13,21,27,43].

Decision making status on family resourece was associated with increased log odds of utilizing minimum antenatal care services among mother. Our results also showed that women who were able to decide with their partners on family resource use had higher log odds of choosing fewer than the recommended number of ANC visits by 90.0%, which implies that housewives with some autonomy in this area were able to make at least some visits to clinics. This suggests that women who were not constrained by a patriarchal structure were better able to utilize ANC services. This finding corroborates a study in northern Ethiopia [18] in which decision making by wives and husbands separately was associated with 45% and 65% lower numbers making the recommended and fewer than recommended number of ANC visits, respectively, compared to couples who made the decision jointly.

Amongst the enabling factors, the odds for utilizing minimum number of ANC clinics increased among mothers who were engaged in non-housewife types of occupations. Our data showed that the log odds ratios associated with gainful small business activities were 4.0 times higher for making the recommended minimoum number of ANC visits. The odds ratios corresponding to the other categories of women’s occupation were also higher. However, their p-values did not demonstrate statistically significant association with ANC service visits. Engaging in skilled employement and small businesses as income sources among mother was associated with increased odds of utilizing recommended minimum number of ANC services which is consistent with those of previous studies in Ethiopia [21,24]. A study in Nigeria [44] reported that women who operated small businesses were 6 times more likely to make the recommended number of ANC clinics visits. Similar studies in Nepal and China also corroborate our findings [11,45].

Furthermore, we found that women in households reporting monthly income of $US 50.00 or more made the recommended minimum number of ANC visits and fewer than recommended visits at a rate 80% and 10% higher than no visits, respectively, compared to those with household incomes of $US 49 or less. Women from high household incomes were more likely to be able to afford health services, and their associated costs, including transportation costs [46,47]. Low household income denoted a major deterrent to mothers to seek prompt care. This variation might be an area of concern for policy makers. A study in Sodo Woreda [21] reported that 13% of rural women with higher cash incomes made the ANC visits recommended by WHO. Similar findings were reported in Nigeria, where women in wealthier households were 4.0 times more likely to make the recommended minimum number of ANC visits [48]. The differentials were 2.71 times in studies in Afghanistan [49], 8.8 times in Nepal [11], and 3.3 times in China [45].

Women who were visited 2 or more times by HEWS were 70% and 20% more likely to make the recommended and fewer than recommended number of visits, respectively, than women who were not visited. Women who were visited only once made the recommended minimum ANC visits and fewer than recommended number of visits at a 50% and 80%, respectively, higher rate. A qualitative study in Kafa Zone, southwestern Ethiopia [50] reported that women preferred to be seen by HEWs who they knew rather than health workers they did not know. Similar oversea findings were reported by several other studies [12,49,51,52]. This finding may be important to develop intervention strategies that more strengethened a systematic and regular home visits by health workers which helps women to improve their utilization of a minimum ANC services or more at home or health post with low opportunity costs.

Women who visited decentralized health centers were 20% more likely to make 3 or fewer ANC visits than those with utilized nearby bottom-line health posts. This finding corroborates those of other studies, which predicted higher log odds for making at least 3 recommended visits to ANC clinics [51,53]. Women were likely to visit health centers that were better equipped and more user friendly than others [54]. But this result was not consistent with the studies in Kaffa Zone, southwestern Ethiopia where majority of study participants reported that they used a minimum number of ANC at the health post rather than the health centre because of the physical distance, the cost of health services at health centres or the hospital and because women preferred to be seen by HEWs who they knew rather than health workers they did not know [50], and in Rwanda, where decentralized health posts were over utilized compared to health centers because health posts were better supplied with maternal resources and attracted most of the local women [55].

After adjusting for all variables, women were found to be 40% more likely to make fewer than recommended minumum number of ANC visits to health facilities located within walking distance of half an hour than to facilities located at greater distances. Similar extensive studies elsewhere [18,24,56] pointed out that walking distance to the available health facilities and time needed to reach these health facilities influences health-seeking behavior and was associated with the utilization of a minimum number of health services. An extensive study in Ethiopia reported that utilization of health facilities declined with distance from maternal service users homes [57].

Severity of pregnancy complication or illness also increased seeking care in health facilities and associated with the utilization of the minimum number of ANC visits [47]. Amongst the need factors, our data show that recognizing the severity of illness by danger signs of severe headaches increases the likelihood of utilizing minimum ANC visits and women with severe headaches had 9.0 times higher log odds of making the recommended number of visits for ANC; headaches appeared to motivate mothers to attend the minimum number of ANC clinics, similar to other studies [18,21,30,43,58]. Furthermore, an increased odds of utilizing minimum ANC services was observed in mothers of high severity of illness by danger signs of high fevere. The study found that women reporting high fevers were 7.1 times times more likely to make the recommended minimum number of ANC visits. This finding is in line with studies in Hadiya Zone in southern Ethiopia [59], west Bengal in India [60], and rural Bangladesh [43]. The reason women with fevers appear motivated to utilize ANC could be that mothers with a history of complications have personal experience that helps them understand the life-threatening condition and makes them inclined to seek preventive maternal care [21].

Log odds of attending recommended minimum number of visits to ANC were 60% among women of Ayana kebele, woreda’s capital, than outer administrative decentralisation units in the rural areas. There were geographical variations in the use of ANC among women across kebeles of different socio-cultural groups [13] with government-decentralized health facilities. Another study found wide interregional disparities in ANC use in Ethiopia, with Oromia Region having the lowest use of all regions except Somali [6]. In India, except for the southern region, as well as in Pakistan, Nigeria, and South Sudan, ANC utilization rates are low [30,48,49]. Low ANC utilization in these areas may be due to historical, socioeconomic, and cultural conditions across these physical settings and community groups.

The log odds of urban residence of women was significantly associated with adherence to the utilization of minimum number of ANC visits than rural women by 10% higher rate. This finding is consistent with meta-analysis in Ethiopia [61]. In Nepal, urban women were 7 and 2 times more likely to make the recommended and fewer than recommended number of ANC visits, respectively [11]. Researchers suggest the reasons for the higher use might be the better quality of care and greater accessibility to professionals in urban areas. Non-significant association between urban residence and fewer than recommended number of ANC visits has been reported elsewhere [27,48].

Our study had a number of limitations. All data were self-reported by the women participants and were not triangulated with other sources, which may have resulted in bias. The study also did not capture institution-based factors such as health providers’ behavior and accessibility and quality of services, all of which influence the health-seeking behavior of women. Moreover, the long recall period may have introduced information bias.

Conclusion

Despite the fact that ANC attendance impacts safe motherhood and reduces maternal deaths, this study found unacceptably low adherence to the recommended number of ANC visits in rural western Ethiopia. Only 15.2% of women studied made the recommended number of visits to ANC clinics, 49.6% made fewer than the recommended number, and 35.2% did not obtain any ANC. The results of this study confirm the importance of several factors in women’s making the WHO-recommended number of visits to ANC clinics: number of births, occupation, visits by HEWs, household income, headache, fever, decentralized administrative kebele, and urban-rural location. Religion, gendered decisions about resource use, walking time to ANC services, decentralized facility type, and vaginal bleeding were significant predictors of fewer than the recommended number of ANC visits.

These results indicate that the study of broadly-based interventions considering the socioeconomic, cultural, demographic, and environmental context of communities may be useful in identifying barriers to ANC utilization and promoting adherence to the recommended number of maternal visits in rural western Ethiopia. Health campaigns conducted through kebele and woreda health services as well as mass media may promote health-seeking behavior of pregnant women and increase the awareness of communities, religious leaders, and other stakeholders about the recommended number of ANC visits so as to reduce maternal and neonatal mortality. However, policies and programs must extend beyond community awareness of the need for adequate and appropriate maternal care, use of family planning to control birth size and spacing, and address also long-term multi-sectoral development issues. Broadly based interventions need to focus on motorized rural roads, public transport, livelihoods and income generation, and gender equity. Interventions must increase the number and the coverage of home visits by HEWs and upgrade and equip front-line health posts or further extend the decentralization of health centers in rural areas. These actions will ensure that the predisposing, enabling, need, and external environmental factors that promote health-seeking behavior are in place to achieve increased ANC utilization and reduction in maternal deaths.

Origonal Andersen’s behavioural health care seeking framework.

(DOCX)

Click here for additional data file.

English language survey questionnaire developed to study the utilization of decentralized health facilities for minimum number ANC visits, rural western Ethiopia.

(DOCX)

Click here for additional data file.

11 Dec 2019

PONE-D-19-25468

Using Andersen’s behavioral model of health care utilization in a decentralized program to examine the use of ANC in rural western Ethiopia

PLOS ONE

Dear Mr. Tolera,

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.

We would appreciate receiving your revised manuscript by Jan 25 2020 11:59PM. When you are 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.

If you would like to make changes to your financial disclosure, please include your updated statement in your cover letter.

To enhance the reproducibility of your results, we recommend that if applicable you deposit your laboratory protocols in protocols.io, where a protocol can be assigned its own identifier (DOI) such that it can be cited independently in the future. For instructions see: http://journals.plos.org/plosone/s/submission-guidelines#loc-laboratory-protocols

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). This letter should be uploaded as separate file and labeled 'Response to Reviewers'.

A marked-up copy of your manuscript that highlights changes made to the original version. This file should be uploaded as separate file and labeled 'Revised Manuscript with Track Changes'.

An unmarked version of your revised paper without tracked changes. This file should be uploaded as separate file and labeled 'Manuscript'.

Please note while forming your response, if your article is accepted, you may have the opportunity to make the peer review history publicly available. The record will include editor decision letters (with reviews) and your responses to reviewer comments. If eligible, we will contact you to opt in or out.

We look forward to receiving your revised manuscript.

Kind regards,

Kannan Navaneetham

Academic Editor

PLOS ONE

Journal Requirements:

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

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 http://www.plosone.org/attachments/PLOSOne_formatting_sample_main_body.pdf and http://www.plosone.org/attachments/PLOSOne_formatting_sample_title_authors_affiliations.pdf

Please ensure that you refer to Figure 2 in your text as, if accepted, production will need this reference to link the reader to the figure.

Please include additional information regarding the survey or questionnaire used in the study and ensure that you have provided sufficient details that others could replicate the analyses. For instance, if you developed a questionnaire as part of this study and it is not under a copyright more restrictive than CC-BY, please include a copy, in both the original language and English, as Supporting Information.  If the original language is written in non-Latin characters, for example Amharic, Chinese, or Korean, please use a file format that ensures these characters are visible.

Please state whether you validated the questionnaire prior to testing on study participants. Please provide details regarding the validation group within the methods section.

Please change your reference to "p=0.000" to "p<0.001" or as similarly appropriate, as p values cannot equal zero.

[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: Yes

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: Title:

• Please avoid the use of abbreviations like (ANC) in the title.

Abstract:

• In the result part, it is better to describe the magnitude (15.2%) and then the factors. “Adherence to the recommended number of ANC service visits was a function of predisposing, enabling, need, and external environmental factors. Women who made the recommended number of ANC visits constituted 15.2% of all subjects.”

• Does this statement “Women’s involvement in gainful activities had higher log odds of seeking the recommended ANC services compared to housewives.” based on the logistic regression model result? If so, you should write the Odds ratio form the regression.

Introduction

• The statistics reported in the first paragraph, “About 300,000 maternal deaths were reported worldwide in 2013”, is rather reflected in the 2015 WHO maternal mortality report. Would you please double-check the figure?

• Paragraph two: “The Ethiopian Demographic 45 Health Surveys of 2005 and 2011 revealed maternal death ratios of 673 and 676 per 100,000 live 46 births, respectively, showing no change between the two studies.” The maternal mortality definition used by the DHS is different from the formal WHO definition used as it considers all pregnancy-related deaths. Hence using the figure from the DHS might not be correct in this context.

• Please summarise the last three paragraphs of the introduction in one succinct sentence that addresses the justification and aim of the study.

Method

• Study setting: Add some brief information regarding the health infrastructure of the study area.

• You don’t need to mention the name of the kebeles “(Ayana, 115 Angar, Ejere, and Lalistu)

• What is the sampling unit for this study? (Household or woman)?

• Inclusion criteria: Are women who reside in the area for less than six months included?

• Explanatory variables: The description regarding the Andersen Newman’s model (conceptual model) could be summarised and moved to the introduction, and focus on the list of predictor variables for this particular study.

• Please move or remove this “None of the participants refused to be interviewed. Five women wanted to end the interviews early due to personal appointments they had to attend to; they were reported as non-responders” to the appropriate section

Result

• Marital status (n=383) Vs Husband’s education and occupation (n=454)??

• Moslem or Muslim?

• Description at kebele, “We found that 89.4% and 66.2% of respondents of Ayana and Ejere Kebeles, respectively, visited the ANC clinic, and lower proportions did so in Angar and Lalistu Kebeles.” may not be required.

• Factors associated with ANC: The bivariate analysis could be merged with the multivariate model in one table to show how the variables are transferred from the first model to the next. Or else, authors can summarise the bivariate analysis in one paragraph without a table and then directly move to describe the analysis results of the final model.

• What is the (n=??) for the regression model?

• I’m not sure why authors consider the kebeles in the regression, as these might have insignificant differences in terms of socio-economic and demographic factors.

Discussion

• The discussion was nicely presented except the issues raised below

• You don’t need to refer to Table 4 in the discussion. Move Table 4 from this section

• The magnitude of use of the ‘minimum number of ANC’ was not discussed.

• Again, the factors do not need to be categorized here in the discussion. First, discuss the magnitude, and then focus on the modifiable factors associated with the utilization of the minimum number of ANC visits.

Reviewer #2: Dear Prof Kannan Navaneetham

Thank you for inviting me to review this interesting paper. I have a very minor collection of comments and hoping my comments are very easy for the author to make correction and I recommend you to accept the paper without modification. The author should also check my comments within the PDf version which I attached it with the word version of my comments. I look forward your invitation again when you have an MS that that be should be reviewed very carefully.

Thanking you.

Here are my minor comments which will improve the paper

Title

Q1. Title: Using abbreviation in title ANC alone is confusing, please write both the expansion and the abbreviations?

Q2. The short title is not actually short: I recommend this, Health care utilization and antenatal care Services in rural western Ethiopia.

Abstract

Q3. Abstract: Re-write the methods and result section separately for clarity to readers, please see my comments within the PDF

Q4. From the abstract result section, for all AOR, please include the confidence interval to see how the association is strong or weak? This is very key information for the scientific community.

Q5. From conclusion of the abstract, utilization of recommended ANC services was strongly linked with predisposing, enabling, need, and external factors? This seems very general better to mention those factors which are actually associated with among predisposing, enabling, need, and external factors due to all factors of predisposing, enabling, need, and external factors are not actually a factor for the use of ANC?

Q6. From line 34, which intervention you recommended as a researcher?

Q7. Key word is not applicable for PLoS ONE publication, thus, delete it.

Introduction

Q8. Referencing style should be in line with the PLoS OEN guideline? This comment for all throughout your MS.

Q9. Line 46-50 should be replaced by the current HSTP (Health Sector Development Plan) and SDG goals, why you sued the expired MDG and HSDP information?

Q10. Better to delete lines 81-83 since it is repetition with line 93-94. Keeping line 93-94 idea is good.

Methodology

Q11. Better to put the map of the study area, if the author put the study area in another published of his paper, he can cite that paper without putting the map within this paper. This gives a clear view for the readers of this paper.

Q12. Line 100 and line 102 refrence should be corrected. Use CSA reference for line 102 AND LINE 100 for Eastern Wollega Zone, Wone Finance and Economic Development Office. Physical and

561 Socio Economic Profile of Gidda Ayana Woreda. Nekemte, Ethiopia; 2015. Making both at the same time seems both data available in both sources, which could not be in real situation?

Q13. Line 104, better to write a ‘community based’ than population based. Again, the langue is not clear, Please see my comment within the PDF for line 104.

Q14. Line 110 to 112 had confusing ideas, if the source population less than 10,000, sample size correction is done and adding 5% non-response rate does not the justification. Adding a non-response rate either less than or greater than 10,000 is a must. Please re-write it/

Q15. Line 121 of using recent births, why not you select randomly one of the mother?

Q16. Line 125-126 seems it will incur bias? Btter to delete it.

Q17. Line 131. Cite the WHO recommendation source for ANC categorization?

Q18. Line 136 to 162 should not be one paragraph? Make it at-least 3

Q19. Line 165-66, re-write, What does it mean literature? Did you mean published paper

Q20, the full name and abbreviation for health extension workers written wt line 117, then at line 170 use only the abbreviations, such comment works for all. Once you used both the full name and the abbreviation, then use the abbreviation, check also about EDHS, ANC and etc.

Q21. Line 181 and 812, re-write for clarity.

Q22 Line 178-193, make it at least two paragraph?

Q23. Line 194. consent is part of ethics and no need to write as a topics , make it the title ‘Ethical consideration”

Result

Q24. Line 219, The exchange rate should be during your data collection period? And please write the time from your result display or in the table?

Q 25. Line 220, Table 1 should have information about mean age, and mean number of children women gave birth? Putting by text from line 211 is not enough?

Q26. Table 1 topic needs modification. There is no cultural related data in the table so that saying Socio-cultural and demographic backgrounds not correct? But the Table also have HEWs information then, you can find out better title as reflecting what exist in the Table?

Q27. For all Tables write the study period and the correct location of the study area by including rural western Ethiopia.

Q28. Liens 236-238 should be deleted since the bivariate association has no value? But I understand that Table 3 about multivariable analysis result and if so, the Table 3 title saying bivariate is total wrong and confusing? Please check

Q29. If Table 3 about Bivaraite analysis, your note on 235 to 307 should be deleted since bivariate data is noting for decision making and the association have no value.

Discussion

Q30. Please bring Table 4 above the discussion at result section?

Q31. Discussion better to be re-written? Repeating the result section at the discussion is not recommended, either you use result and discussion within the same topic or separately, I thought you have a separate topic for discussion, so in your discussion section focus on discussions part? No need to write each result again? At the result section, put all your result and focus on the main findings, you wasted your time at the result by bivariate result at Table 4, but the most important finding is Table 4.

Q31. Line 451, write the reason why no grant number

Q 32, 455, in your acknowledgement, please acknowledge also data collectors supervisors, Addis Abbaa University, Department f Geography at AAU and etc. This tells how you are very careful

**********

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: Yes: Metadel Adane (PhD)

Assistant Professor of Water and Public Health

Department of Environmental Health

Wollo University

Dessie,. Ethiopia

[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 to be viewed.]

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 us at figures@plos.org. Please note that Supporting Information files do not need this step.

Submitted filename: PONE-D-19-25468-Metadel comments.pdf

Click here for additional data file.

Submitted filename: PONE-D-19-25468-comment.doc

Click here for additional data file.

31 Dec 2019

Resubmission date: December 24, 2019

Manuscript ID: PONE-D-19-25468

Title: “Using Andersen’s behavioural model of health care utilization in a decentralized program to examine the use of antenatal care in rural western Ethiopia”

Authors: Habtamu Tolera, Tegegne Gebre-Egziabher, Helmut Kloos

Dear Dr. Navaneetham,

Thank you for your letter dated December 11, 2019. We were pleased to know that our manuscript was considered potentially acceptable for publication in PLOS ONE, subject to adequate revision as requested by the reviewers. Based on the instructions provided in your letter, we uploaded the file of the rebuttal letter; the marked-up copy of the revised manuscript highlighting the changes made in the original submitted version and the clean copy of the revised manuscript.

We have revised the manuscript by modifying the abstract, introduction, methods, results, discussion and other sections, based on the comments made by the reviewers and using the journal guidelines. Accordingly, we have marked in red color all the changes made during the revision process. Appended to this letter is our point-by-point response to the comments made by the editor and the two reviewers.

We agreed with almost all the comments/questions raised by the editor and the reviewers and provided justification for disagreeing with some of them. We would like to take this opportunity to express our thanks to the editor and the reviewers for their valuable comments and to thank you for allowing us to resubmit a revision of the manuscript.

I hope that the revised manuscript is accepted for publication in PLOS ONE.

Sincerely yours

Habtamu Tolera

PhD Candidate in Socio-Economic Development in Addis Ababa University, Ethiopia

Lecturer and Researcher in Wollega University, Ethiopia

Phone: +251 (0) 912015545

E-mail address: habtol@yahoo.com

Response to the academic Editor’s and the reviewers’s comments

Response to the Editor-in-Chief’s comments

General comment:

DearMr.Tolera,

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. We would appreciate receiving your revised manuscript by Jan 25 2020 11:59PM. When you are 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.

If you would like to make changes to your financial disclosure, please include your updated statement in your cover letter.

To enhance the reproducibility of your results, we recommend that if applicable you deposit your laboratory protocols in protocols.io, where a protocol can be assigned its own identifier (DOI) such that it can be cited independently in the future. For instructions see: http://journals.plos.org/plosone/s/submission-guidelines#loc-laboratory-protocols

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). This letter should be uploaded as separate file and labeled 'Response to Reviewers'.

• A marked-up copy of your manuscript that highlights changes made to the original version. This file should be uploaded as separate file and labeled 'Revised Manuscript with Track Changes'.

• An unmarked version of your revised paper without tracked changes. This file should be uploaded as separate file and labeled 'Manuscript'.

• Please note while forming your response, if your article is accepted, you may have the opportunity to make the peer review history publicly available. The record will include editor decision letters (with reviews) and your responses to reviewer comments. If eligible, we will contact you to opt in or out.

• We look forward to receiving your revised manuscript.

Kind regards,

Kannan Navaneetham

Academic Editor

PLOS ONE

Dear Editor, thank you for your positive decision and for your letter allowing us resubmitting our revised manuscript for publication. We all agreed, reconsidered and revised all concerns and commented raised in a point-by-point manner as you and your two reviewers rightly suggested. We addressed all points alternatively in the following sections. Thank once again!!

1. Journal Requirements:

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

1.1. 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 http://www.plosone.org/attachments/PLOSOne_formatting_sample_main_body.pdf and http://www.plosone.org/attachments/PLOSOne_formatting_sample_title_authors_affiliations.pd.

Thank you for your remark. We really appreciate your comments. We have read journal’s author guidelines and tried to meet PLOS ONE's citation style and formatting requirements, including those for file naming as academic editor suggested.

1.2. Please ensure that you refer to Figure 2 in your text as, if accepted, production will need this reference to link the reader to the figure.

Thank you for your comment. We refer to Figure 2 in the original submission as supportive information, S1_Fig.Doxc file, in the revised submission. You can check a list on page 34 lines 686 under Supportive information” statement.

1.3. Please include additional information regarding the survey or questionnaire used in the study and ensure that you have provided sufficient details that others could replicate the analyses. For instance, if you developed a questionnaire as part of this study and it is not under a copyright more restrictive than CC-BY, please include a copy, in both the original language and English, as Supporting Information. If the original language is written in non-Latin characters, for example Amharic, Chinese, or Korean, please use a file format that ensures these characters are visible.

Thank so much for your observation. We have uploaded the minimum anonymized data set as a supporting information files as suggested which include English language survey questionnaire (S1_text.docx), see the revised manuscript of “Supportive information" statement on page 34 lines 687-689.and Original Andersen Behavioural model for healthcare seeking (S1_Fig.Doxc).

1.4. Please state whether you validated the questionnaire prior to testing on study participants. Please provide details regarding the validation group within the methods section.

Thank so much for your observation. As mentioned above, we reconsidered your concern and discussed in details under sub-section” Data collection and quality control” in this revised report. See page 7-8, lines 170-180.

1.5. Please change your reference to "p=0.000" to "p<0.001" or as similarly appropriate, as p values cannot equal zero

Thank so much for your observation and we revised as you fairly commented us. See Table 3 on pages 17-18, lines 336-337.

Responses to Reviewer comments # 1

1. Title:

1.1. Please avoid the use of abbreviations like (ANC) in the title.

Thank you for your remark. We have agreed with your concern and taken out this acronym, “ANC”, from the study’s title as commented. See page 1, line 3.

2. Abstract

2.1. In the result part, it is better to describe the magnitude (15.2%) and then the factors. “Adherence to the recommended number of ANC service visits was a function of predisposing, enabling, need, and external environmental factors. Women who made the recommended number of ANC visits constituted 15.2% of all subjects”

First of all, we appreciate your constructive comment. We observed your concern much because it reshaped and further strengthen the value of the manuscript, and well-kept the flows of ideas. We have reworked accordingly. See page 1, lines 19-27 of the ‘Results’ section under Abstract.

2.2. Does this statement “Women’s involvement in gainful activities had higher log odds of seeking the recommended ANC services compared to housewives.” based on the logistic regression model result? If so, you should write the Odds ratio form the regression

Thank you so much for your comment. Hence, we incorporated the remark as suggested above. See page 1, lines 23-24 of the ‘Results’ section under Abstract.

3. Introduction

3.1. The statistics reported in the first paragraph, “About 300,000 maternal deaths were reported worldwide in 2013”, is rather reflected in the 2015 WHO maternal mortality report. Would you please double-check the figure?

Thank you for your remark. As per of your concern, different global figures in different publications or reports are available there online. We double checked and used recent report as you commented us and incorporated global figures reflected in the 2015 WHO maternal mortality report: Trends maternal mortality. 1990 to 2015: Estimates by WHO, UNICEF, UNFPA, World Bank Group and the United Nations Population Division. Geneva: World Health Organization.; 2015. See pages 1-2, lines 36-40 under the ‘Background’ section.

3.2. Paragraph two: “The Ethiopian Demographic Health Surveys of 2005 and 2011 revealed maternal death ratios of 673 and 676 per 100,000 live births, respectively, showing no change between the two studies.” The maternal mortality definition used by the DHS is different from the formal WHO definition used as it considers all pregnancy-related deaths. Hence using the figure from the DHS might not be correct in this context.

We appreciate your suggestion. As commented above, we have taken out local DHS figures we cited in the original submission and used WHO global figures on country’s report in place as per of your suggestion. See page 2, lines 41-43 under the ‘Background’ section.

3.3. Please summarize the last three paragraphs of the introduction in one succinct sentence that addresses the justification and aim of the study.

Thank so much for your observation. Yes, we have understood your concern and merged the last three paragraphs of the original submission and modified to a concise paragraph as you indicated above in this revised submission. See pages 3-4, lines 69-93 under the ‘Background’ section.

4. Method

4.1. Study setting: Add some brief information regarding the health infrastructure of the study area.

We incorporated this concern as rightly indicated above. See page 4, lines 102-106 under the ‘Study setting’ section.

4.2. You don’t need to mention the name of the kebeles “(Ayana, 115 Angar, Ejere, and Lalistu).

Alright, as a reviewer indicated in his report, we have taken out the name of the kebeles from ‘Sample and recruitment’ section under the ‘Methodology. See page 5, lines 119.

4.3. What is the sampling unit for this study? (Household or woman)?

We appreciate your observation. We used ‘Household’ or ‘woman’ interchangeably unintentionally. However, the actual case, sampling unit is the pregnant women not the household. Thus, modification has been made accordingly where it appears across the text in this revised submission. See page 5, lines 120-122, under this revised submission.

4.4. Inclusion criteria: Are women who reside in the area for less than six months included?

Thank you for reminding us to mention. Women who came from somewhere else, outside the study woreda and did not reside in the study kebeles 6 months or fewer preceding data collection were not included in the study. This is indicated in the revised submission. See page 6, lines 129-130, under the sub-section ‘Sample and recruitment’.

4.5. Explanatory variables: The description regarding the Andersen Newman’s model (conceptual model) could be summarized and moved to the introduction, and focus on the list of predictor variables for this particular study.

As marked above, we have summarized description about Andersen Model. See in the “Background” section on page 3-4, lines 78-88. See also the revised list of explanatory variables on pages 6-7, from lines 140-160 in the sub-section ‘Explanatory variables’ under the ‘Methodology’

4.6. Please move or remove this “None of the participants refused to be interviewed. Five women wanted to end the interviews early due to personal appointments they had to attend to; they were reported as non-responders to the appropriate section.

Thank you so much. We have taken out the above texts as per of your comment.

5. Result

5.1. Marital status (n=383) Vs Husband’s education and occupation (n=454)??

We checked everything about the concerns. we have checked but not observed no any arithmetical errors with regard to, n or N, in Table 1 or may we not understand a reviewer’s report?

5.2. Moslem or Muslim?

Thank so much for your comment on our inconsistency in not using one of the above across the manuscript. We have used “Muslim” to keep consistency throughout the whole revised submission. See, e.g., page 14, line 263 and elsewhere.

5.3. Description at kebele, “We found that 89.4% and 66.2% of respondents of Ayana and Ejere Kebeles, respectively, visited the ANC clinic, and lower proportions did so in Angar and Lalistu Kebeles.” may not be required.

Thank you so much for your observation. We all authors needed to see some variations observed in proportions of ANC service utilization among pregnant women across kebeles in which decentralized public primary health facilities were located. So kebeles are administrative decentralization entities in which health facilities or maternal programs including health workers were transferred one step down or are decentralization proxy variable. But mainly and primarily, we practically observed that there are some variations in cultural practice among study kebeles because Ayana and Lalistu kebeles are almost Oromo dominated host communities while Angar and Ejere are Amhara and Tigre dominated, originally settlers, from other regions outside Oromia region. This’s is our primary interest to analyse the frequency distribution of ANC utilization using kebeles as a categorical variable and the reason we also included kebeles in our regression model. See Table 2 on pages 12-13, and pages 12, lines 238-240 sub-section “ANC service use and decentralized facility attended”; see also the ‘External environmental factors’ part under ‘Result’ on page 17, from lines 325-329.

5.4. Factors associated with ANC: The bivariate analysis could be merged with the multivariate model in one table to show how the variables are transferred from the first model to the next. Or else, authors can summarise the bivariate analysis in one paragraph without a table and then directly move to describe the analysis results of the final model

Yes, as indicated above, bivariate analyses outputs are crude and are not as such important for decisions. Hence, we have summarised the data in the bivariate Table 3, in old submission in one paragraph in the new submission, see page 13, lines 249-258 and finally we deleted a Bivariate Table 3 of the original submission.

5.5. What is the (n=??) for the regression model?

We appreciate your remark. As indicated above, n=?? is deleted in the new submission because it is our, an excused technical error. See Table 3, on page 17, line 336.

5.6. I’m not sure why authors consider the kebeles in the regression, as these might have insignificant differences in terms of socio-economic and demographic factors.

Thank, we appreciate your observation. This question is a part of questions raised in subsection 5.3 and it is well answered there. See above, subsection 5.3.

6. Discussion: the discussion was nicely presented except the issues raised below

6.1. You don’t need to refer to Table 4 in the discussion. Move Table 4 from this section

We appreciate your suggestion. and we have incorporated both concerns. See the moved Table 3 on pages 17-18, lines 336-337 under the Result section in this revised submission, we also deleted unnecessary refer to Table 4 in the old submission under “Discussion” part.

6.2. The magnitude of use of the ‘minimum number of ANC’ was not discussed.

Thank you so much for constructive comment. As suggested, we have tried to satisfy a reviewer’s concern. We have given emphasis in the revised “Discussion” part of the new submission for the magnitude of use of the ‘minimum number of ANC visits as much as possible. See pages 18-24, lines 337-460.

6.3. Again, the factors do not need to be categorized here in the discussion. First, discuss the magnitude, and then focus on the modifiable factors associated with the utilization of the minimum number of ANC visits.

Thank you very much also. Authors have also taken out the categories used in the

“Discussion” section of the original submission. See the revised version on pages

19-24, lines 350-463.

Responses to Reviewer comments # 2 , Dr. Metadel Adane

General Comments:

Thank you for inviting me to review this interesting paper. I have a very minor collection of comments and hoping my comments are very easy for the author to make correction and I recommend you to accept the paper without modification. The author should also check my comments within the PDf version which I attached it with the word version of my comments. I look forward your invitation again when you have an MS that that should be reviewed very carefully.

Thank you for your kind recognition of our efforts. We also appreciate the positive and invaluable assessment and feedbacks you offer us to add the value on this work. Thank you once also. In the following sections, we have tried to respond on reviewer #2’s comments, concerns.

Title

Q1. Title: Using abbreviation in title ANC alone is confusing, please write both the expansion and the abbreviations?

Thank you so much for your observation. Authors have agreed with your concern and taken out this acronym, “ANC” in the title as you fairly suggested so as to avoid confusion among future readers of the article, see page 1, line 3.

Q2. The short title is not actually short: I recommend this, Health care utilization and antenatal care services in rural western Ethiopia.

We appreciate your concern. We have incorporated the comment indicated above during our system-based online uploading to the journal in the revised submission.

Abstract

Q3. Abstract: Re-write the methods and result section separately for clarity to readers, please see my comments within the PDF.

We have satisfied your concern. We have put both the methods and result sections under Abstract separately in the revised submission. See a front page, lines 14-27.But there is a word limit, 300 words, to describe the details

Q4. From the abstract result section, for all AOR, please include the confidence interval to see how the association is strong or weak? This is very key information for the scientific community.

We appreciate your invaluable comment and incorporated both AOR and 95%CI in both the “Result” section under the “Abstract” and throughout the manuscript in the revised submission. E.g., see page 1, lines 20-27; pages 14-18, lines 266-337.

Q5. From conclusion of the abstract, utilization of recommended ANC services was strongly linked with predisposing, enabling, need, and external factors? This seems very general better to mention those factors which are actually associated with among predisposing, enabling, need, and external factors due to all factors of predisposing, enabling, need, and external factors are not actually a factor for the use of ANC?

Yes, positive remark. This is modified as per of the above suggestion. See page 1 lines 28-29 in the revised version of the conclusion under the Abstract.

Q6. From line 34, which intervention you recommended as a researcher?

Thank you so much for positive comment and it has been rewritten to satisfy reviewer’s concern as suggested above. See page 1 lines 30-34 in the sub-section ‘Conclusion’ under the ‘Abstract’ in this revised submission.

Q7. Key word is not applicable for PLoS ONE publication, thus, delete it.

We appreciate the comment, excuse our failure of not keeping PLoS ONE’s author guide or template. As noticed, we have removed this section. See the revised submission in the front page.

Introduction

Q8. Referencing style should be in line with the PLoS OEN guideline? This comment for all throughout your MS.

Thank you, yes, we have to satisfy the requirements of the journal. We have corrected in-text citation style problem indicated above as per of PLoS OEN guideline, number in-text citations in rectangular bracket throughout the entire revised version. See the first in-text citation, as example on page 2 line 40. Thank you once again.

Q9. Line 46-50 should be replaced by the current HSTP (Health Sector Development Plan) and SDG goals, why you used the expired MDG and HSDP information?

It is a positive comment and we have modified accordingly. See page 2 from lines 41-43 in the “Introduction” section in the new submission.

Q10. Better to delete lines 81-83 since it is repetition with line 93-94. Keeping line 93-94 idea is good.

We appreciate your comment. As indicated, we have removed the sentences from lines 81-83 in original submission and kept lines 93-94 . Of course, we almost modified the “Introduction” of the study to satisfy some concerns raised by reviewer # 1. See pages 2-4, lines 35-93 of the new submission.

Methodology

Q11. Better to put the map of the study area, if the author put the study area in another published of his paper, he can cite that paper without putting the map within this paper. This gives a clear view for the readers of this paper.

Thank you. As mentioned above, we have cited another work. See page 4 line 96 under a sub-section “Study setting”.

Q12. Line 100 and line 102 reference should be corrected. Use CSA reference for line 102 AND LINE 100 for Eastern Wollega Zone, Wone Finance and Economic Development Office. Physical and 561 Socio Economic Profile of Gidda Ayana Woreda Nekemte, Ethiopia; 2015. Making both at the same time seems both data available in both sources, which could not be in real situation?

We accept the comment. We used both sources purposively because the 2007 Census release reported the total population as Gida-Kiremu Woreda. Hoever, after census the woreda was divided in to two woredas, Gida Ayana and Kiremu Woredas, and even same kebeles cross the border spilt into two, one side to Gida Ayana and the other side to Kiremu Woreda in 2008. This a challenge for us and we used both counts by the woredas and counts by the census. That’s why we cite both. So, this is the case and better if we cite both sources as they are originally indicated. See page 4 lines 99-102.

Q13. Line 104, better to write a ‘community based’ than population based. Again, the langue is not clear, please see my comment within the PDF for line 104.

We have incorporated the concern as per of the suggestion. It is also rewritten. See the revised version on page from 5 lines 108-109 in the sub-section “Study design and period” under Methodology part.

Q14. Line 110 to 112 had confusing ideas, if the source population less than 10,000, sample size correction is done and adding 5% non-response rate does not the justification. Adding a non-response rate either less than or greater than 10,000 is a must. Please re-write it.

We have agreed with your concern and revised the issue as rightly suggested See page 5, lines 113-117 in the revised submission

Q15. Line 121 of using recent births, why not you select randomly one of the mothers?

We preferred to interview a mother with a recent birth if more than one housewife in a given household because it is more logical that this late mother recalls more of her reproductive history than that of her counterpart who gave births earlier than her in same household. Look at page 5 lines 120-122.

Q16. Line 125-126 seems it will incur bias? Better to delete it.

We have removed the entire sentence as a reviewer indicated above.

Q17. Line 131. Cite the WHO recommendation source for ANC categorization?

We have incorporated to satisfy the above concern. See page 6, from lines 134-35 in the sub-section “Outcome variable” under Methodology.

Q18. Line 136 to 162 should not be one paragraph? Make it at-least 3

Aright, we have revised accordingly. Look at pages 6-7 from lines 141-160, in the new submission, a sub-section called “Explanatory variables”

Q19. Line 165-66, re-write, what does it mean literature? Did you mean published paper?

Thank you for the concern. As indicated, our interest is to mean it a published paper. Hence, we have rewritten as suggested. See page 7, line 164 in the revised version under a sub-section “Data collection and quality control”

Q20, the full name and abbreviation for health extension workers written at line 117, then at line 170 use only the abbreviations, such comment works for all. Once you used both the full name and the abbreviation, then use the abbreviation, check also about EDHS, ANC and etc.

We have appreciated you for the invaluable comment. We have incorporated the above specific comments and related suggestions across the manuscript. See e.g., page 7 line 168 in this revised version.

Q21. Line 181 and 812, re-write for clarity.

Ok, revision is made as mentioned. See page 8 lines 185-187.

Q22 Line 178-193, make it at least two paragraphs?

Changes have been made as per of your suggestion. See pages 8-9 lines 182-203 of this revised version.

Q23. Line 194. consent is part of ethics and no need to write as a topic, make it the title ‘Ethical consideration”

We appreciate you for your observation. This is revised to satisfy the comment. See page 9 line 204.

Result

Q24. Line 219, The exchange rate should be during your data collection period? And please write the time from your result display or in the table?

Truly, it is revised accordingly. See the foot note of Table 1 on page 12, line 233 in this revised submission.

Q 25. Line 220, Table 1 should have information about mean age, and mean number of children women gave birth? Putting by text from line 211 is not enough?

This is revised as suggested above. Look at page 10 lines 226-227.

Q26. Table 1 topic needs modification. There is no cultural related data in the table so that saying Socio-cultural and demographic backgrounds not correct? But the Table also have HEWs information then, you can find out better title as reflecting what exist in the Table?

We have revised incorporating all comments raised. See e.g., title of Table 1 page 10, lines 230-231; page 9 line 217, in the new submission.

Q27. For all Tables write the study period and the correct location of the study area by including rural western Ethiopia.

This is revised accordingly. See e.g., pages 10 lines 230-231.

Q28. Liens 236-238 should be deleted since the bivariate association has no value? But I understand that Table 3 about multivariable analysis result and if so, the Table 3 title saying bivariate is total wrong and confusing? Please check

We appreciate your concern. As per of your comment and also suggestions from the second reviewer we have deleted Table 3 or Bivariate Table. Rather, as reviewer #1 marked in his report as well we have preferred to summarizing Table 3 bivariate data into one concise paragraph immediately before proceeding discussion of multivariable analyses in the revised submission. See page13 lines

249-258. Finally, Table 3 was deleted.

Q29. If Table 3 about Bivariate analysis, your note on 235 to 307 should be deleted since bivariate data is noting for decision making and the association have no value

.

Thank you for your worry. Table 3 in the original version is as you said Bivariate Table. But the analysis preceding Table 3 or Bivariate Table is a Multivariable regression analysis. Thus, our note on lines 235 to 307 on the original submission should not be deleted since it is a multivariable regression analysis. We have moved Multivariable Table from the “Discussion” section to coincide the analysis section presented under “Results”. See pages 13-18 lines 259-337 in the revised submission.

Discussion

Q30. Please bring Table 4 above the discussion at result section?

Alright, we have done it as commented and we modified the Table into Table 3. See pages 17-18 lines 336-337.

Q31. Discussion better to be re-written? Repeating the result section at the discussion is not recommended, either you use result and discussion within the same topic or separately, I thought you have a separate topic for discussion, so in your discussion section focus on discussions part? No need to write each result again? At the result section, put all your result and focus on the main findings, you wasted your time at the result by bivariate result at Table 4, but the most important finding is Table 4.

Thank you so much for your positive comments. We have revised the “Discussion” part as suggested. See pages 19-24 lines 350-463 in the revised submission.

Q31. Line 451, write the reason why no grant number.

Thank, the reason is that the amount of money offered is very small and is for direct data collection. It is difficult to label it as a full Ph.D. program fund and give a grant number. This is not the customary of the two institutions.

Q 32, 455, in your acknowledgement, please acknowledge also data collectors and supervisors, Addis Ababa University, Department of Geography at AAU and etc. This tells how you are very careful.

Thank, as you mentioned we acknowledged data collectors and the supervisors. See page 26 lines 506-507 in the new submission under ‘Acknowledgement” section. However, PLoS One author guideline restricts us not to include funding institution or competing interest’s information in Acknowledgments section.

We would like to thank the reviewers and editors for evaluating our manuscript. We have tried to address all the concerns in a proper way and believe that our paper has been considerably improved. We would be happy to make further corrections if necessary and look forward to hearing from you all soon.

Submitted filename: Response to Reviwers.docx

Click here for additional data file.

13 Jan 2020

Using Andersen’s behavioral model of health care utilization in a decentralized program to examine the use of antenatal care in rural western Ethiopia

PONE-D-19-25468R1

Dear Dr. Tolera,

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

Within one week, you will receive an e-mail containing information on the amendments required prior to publication. When all required modifications have been addressed, you will receive a formal acceptance letter and your manuscript will proceed to our production department and be scheduled for publication.

Shortly after the formal acceptance letter is sent, an invoice for payment will follow. To ensure an efficient production and billing process, please log into Editorial Manager at https://www.editorialmanager.com/pone/, click the "Update My Information" link at the top of the page, and update your user information. 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 enable them to help maximize its impact. If they will be preparing press materials for this manuscript, you must inform our press team as soon as possible and 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.

With kind regards,

Kannan Navaneetham

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 #1: All comments have been addressed

Reviewer #2: All comments have been addressed

**********

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 #1: Yes

Reviewer #2: Yes

**********

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

Reviewer #1: Yes

Reviewer #2: 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 #1: Yes

Reviewer #2: 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 #1: Yes

Reviewer #2: 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 #1: (No Response)

Reviewer #2: The author is a smart researcher. He addressed every concern in scientific manner and I really appreciate his commitment.

**********

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 #1: No

Reviewer #2: Yes: Dr Metadel Adane (PhD in Water and Public Health).

Department of Environmental Health

Wollo University

Dessie, Ethiopia

17 Jan 2020

PONE-D-19-25468R1

Using Andersen’s behavioral model of health care utilization in a decentralized program to examine the use of antenatal care in rural western Ethiopia

Dear Dr. Tolera:

I am 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 notify them about your upcoming paper at this point, to enable them to help maximize its impact. If they will be preparing press materials for this manuscript, 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.

For any other questions or concerns, please email plosone@plos.org.

Thank you for submitting your work to PLOS ONE.

With kind regards,

PLOS ONE Editorial Office Staff

on behalf of

Professor Kannan Navaneetham

Academic Editor

PLOS ONE