Factors Associated With the Timing of Antenatal Care Seeking in Bangladesh

Abstract

This study investigates the factors affecting the timing of antenatal care (ANC) seeking among Bangladeshi women using the 2007 Bangladesh Demographic and Health Survey data. Overall, 52% of the most recently pregnant women sought skilled ANC. The most pronounced reason for not receiving the services was that it was “not needed.” Of the women who sought ANC, 57% entered for ANC at the fourth month or later. The multivariate regression analysis revealed that maternal age, women’s education, residence, wealth index, pregnancy intention status, child’s birth order, and wanting more children were the important determinants of ANC services utilization. Wanting more children and contraception use were no longer independent contributors to late ANC entry. These barriers should also be addressed as part of health interventions through information education and communication programs to early entry to ANC. The need to reduce financial barriers to care is a major implication of this study.

Keywords

antenatal care Bangladesh pregnancy women’s education logistic regression

Introduction

Routine antenatal care (ANC) includes medical interventions and advice that a woman receives during pregnancy and is a key entry point for pregnant women to receive a broad range of health promotion and preventive health services,¹ including knowledge about healthy practices during pregnancy, nutritional support, prevention and treatment of anemia, diagnosis, treatment of other diseases, and tetanus toxoid immunization.² Late ANC or inadequate attendance at antenatal clinics has been associated with poor pregnancy outcomes, such as low birth weight, prematurity, and increased delivery intervention.³ However, there is not much literature on the risks and protective factors associated with the timing of ANC for mothers.⁴

An estimated 358 000 maternal deaths occurred worldwide in 2008, a 34% decline from the levels in 1990. Despite this decline, developing countries continued to account for 99% of the deaths.⁵ Also, a significant proportion of morbidities occur in developing countries.⁶ Although ANC might not have the potential to predict and avert obstetric emergencies during pregnancy and childbirth, it exposes women to health education on risk factors and encourages them to deliver with a skilled attendant or in a health facility.⁷ Several studies have shown that women who started ANC attendance early and attended frequently were more likely to be assisted during delivery by a skilled attendant compared with those who initiated ANC late and attended only a few visits.^8,9 Recent studies have suggested that women who knew about risk factors were more likely to utilize health facilities for delivery than those without this knowledge.⁸

Early entry to ANC is important for early detection and treatment of adverse pregnancy-related outcomes.¹⁰ The World Health Organization (WHO) recommends that pregnant women without uncomplicated pregnancies in developing countries should receive at least 4 ANC visits, with the first visit starting before the fourth month of gestation, whereas in developed countries such as the United Kingdom and the United States, ANC is recommended within the first 3 months of pregnancy.^11-14 ANC is generally considered to have a positive effect on the health of both the mother and baby. Guidelines in ANC are diverse and not entirely evidence based.¹⁵ Although there is no consensus about the optimal number of antenatal visits, it is proved that inadequate ANC is related to a worse pregnancy outcome.^16,17 Thus, it is important for health policy makers to understand what factors have an influence on the timing of ANC visits.

Bangladesh is committed to the fifth Millennium Development Goal (MDG-5) target of reducing its maternal mortality ratio by three-quarters between 1990 and 2015. Since the early1990s, Bangladesh adopted a strategy of improving access to facilities equipped and staffed to provide emergency obstetric care (EmOC). Utilization of ANC increased substantially from 24% in 1991 to 60% in 2004, with a relatively greater increase in rural than urban areas. Professional attendance (medical, midwifery, trained health workers, etc) at delivery increased by 50% (from 9% to 14%, more rapidly in rural than urban areas), and caesarean sections trebled (from 2% to 6%), but these indicators remained low even by developing country standards.¹⁸ ANC and professional attendance at birth were more common among urban, rich and educated women. Disparities become more pronounced at the extremes of the socioeconomic spectrum. Wider provision of EmOC has clearly increased access to obstetric surgery, although much of the increase occurred in the private sector, and only the wealthier women benefit from these interventions.¹⁸ Despite commendable progress in improving uptake of ANC and in equipping health facilities to provide EmOC, the very low utilization of these facilities, especially by poor women, is a major impediment to meeting MDG-5 in Bangladesh.¹⁸

Bangladesh has achieved important health gains over the recent decades, particularly in reducing maternal mortality and infant and child mortality. Despite this, little is known regarding maternal health care services utilization and, in particular, the timing of ANC services use among women of Bangladesh—a developing and agricultural country where the vast majority live in rural areas. Moreover, there may be enormous gaps with regard to utilization of reproductive health care services among women from different socioeconomic strata. This study aims to investigate the factors affecting the receiving of ANC services and, in particular, the timing of ANC among women of Bangladesh using a nationally representative and cross-sectional data set.

The Andersen Health-Seeking Behavior Model

The analysis of health care services utilization is frequently based on the behavioral model developed by Andersen.¹⁹ This model assumes that health-seeking behavior is the result of interaction between characteristics of individuals, the population, and the surrounding environment.¹⁰ The model consists of several main components, such as predisposing components, enabling, needs, personal health behavior, outcomes, and environment. The predisposing component refers to the characteristics of the person and includes demographic, social position, and attitudinal variables. The enabling component consists of conditions that make health care use available to the person. The third component—illness level—is most directly related to health care use and comprises perceived illness and diagnosed illness. In the context of earlier studies, illness level can be translated to pregnancy history and the current course of the pregnancy.¹⁵

Methods

Data for this study have been taken from the 2007 Bangladesh Demographic and Health Survey (BDHS). The 2007 BDHS is a nationally representative survey, which gathered information from 10 996 ever-married women from rural and urban areas. The survey followed a multistage cluster sampling procedure. It obtained detailed information on fertility levels, marriage, fertility preferences, awareness and use of family planning methods, breastfeeding practices, nutritional status of women and young children, childhood mortality, maternal and child health, and knowledge and attitudes regarding HIV/AIDS and other sexually transmitted infections. A total of 6058 births were recorded in the survey from rural and urban areas that occurred during the past 5 years preceding the survey date. Information regarding ANC was collected from the most recent pregnancy of 4905 women and forms the basis of this study. The 2007 BDHS was conducted under the authority of the National Institute for Population Research and Training (NIPORT) of the Ministry of Health and Family Welfare. It was implemented by Mitra and Associates, a Bangladeshi research firm located in Dhaka. Technical assistance was provided by Macro International Inc, through the MEASURE DHS program. Financial support for the survey was provided by the US Agency for International Development (USAID/Bangladesh).The details of the survey are given elsewhere.²⁰

Outcome Variables

In this study, we considered the most recent births delivered by the women in the past 5 years preceding the survey date. The outcome variables of this study are 2: (1) ANC seeking; and (2) the timing of receiving ANC services. An ANC visit was defined as a contact with medically trained health personnel to follow up pregnancy. All women were asked, “When you were pregnant with (name of last child), did you see anyone for a medical checkup?” Following this, another question was asked: “How many months pregnant were you when you first received ANC for this pregnancy?” The duration of pregnancy was given in continuous form by months. The first outcome variable was dichotomous, and the responses were recorded as either “yes” or “no.” The second outcome variable was classified into: “early entry to ANC,”, “late entry to ANC,” and “did not receive any ANC service.” “Early entry to ANC” was defined as entering ANC before the fourth month of pregnancy, whereas “late entry” was defined as entering ANC at the fourth month or later. This cutoff point was considered following the direction provided by the World Health Organization.¹⁴

Exposure Variables

The ANC visits were analyzed in association with predisposing, enabling, environmental, and illness characteristics as related to the pregnancy. Maternal age and women’s education were the predisposing factors examined. The enabling factor included in this study is the wealth quintile of the households. The wealth index used in this study was developed and tested in a large number of countries to measure inequalities in household income, use of health services, and health outcome.²⁰ It is an indicator of the level of wealth that is consistent with expenditure and income measures. The wealth index is constructed from data on household assets, including ownership of durable goods (such as televisions and bicycles) and dwelling characteristics (such as source of drinking water, sanitation facilities, and construction materials). To create the wealth index, each asset was assigned a weight (factor score) generated through principal component analysis, and the resulting asset scores were standardized in relation to a normal distribution with a mean of 0 and standard deviation of 1. Each household was then assigned a score for each asset, and the scores were summed for each household; individuals were ranked according to the total score of the household in which they resided. The sample was then divided into quintiles from 1 (lowest) to 5 (highest). The details of the measurement of wealth quintiles have been given elsewhere.²⁰ The environmental factors included for analyses are region and place of residence, and pregnancy-related determinants are pregnancy intention status and parity.

Sample Size and Statistical Analyses

We used children’s files for this study from the 2007 BDHS raw data set. The births that occurred during the 5 years preceding the survey were recorded in this file. Information such as birth history, birth order, ANC, delivery care, delivery assistance, and postnatal care has been gathered from the respective mothers of the children. To analyze the factors affecting the utilization of ANC service, we included all respondents (4905 women) from whom information regarding ANC services for the most recent birth was collected. This would help improve our understanding of which factors influence the use and nonuse of ANC services. Besides, to capture the impact of the factors on the late receiving of ANC services, we excluded those women who did not receive any ANC service from multivariate analyses. With this background, the sample size for analyzing the late receiving of ANC service stood at 2536 women.

Simple cross-tabulation and bivariate and multivariate statistical analyses were performed in this study. Simple cross-tabulation was undertaken to describe percentage distribution of the sample women. Bivariate techniques—namely, χ² tests were used to examine the association between the outcome and exposure variables. Multivariate binary logistic regression analyses were conducted to assess the effect of socioeconomic factors on receiving any ANC service and late entry to ANC. Both the outcome variables were made dichotomous. For the first outcome variable, the women who did not receive any ANC service were coded as “0” and “1” otherwise. For the second outcome variable, the women who received ANC early were coded as “0,” and those who received ANC late were coded as “1.” For both outcome variables, the model-fitting process involved 4 stages of estimation. The first model, model I included only predisposing factors. Model II included predisposing and environmental factors. Model III was constructed using the predisposing, environmental, and enabling factors. Model IV, the final model, included all contextual factors considered in this study. The results of the logistic regression analysis are shown using odds ratios with 95% confidence intervals. All the statistical analyses were performed in SPSS v17 (SPSS Inc, Chicago, IL).

Ethical Considerations

The study used nationally representative data extracted from the 2007 BDHS. The BDHS was conducted as a part of the global Demographic and Health Surveys (DHS) program. It was designed to provide data to monitor the population and health situation in Bangladesh as a follow-up to the previous BDHS surveys. The survey was conducted under the authority of the National Institute for Population Research and Training (NIPORT) of the Ministry of Health and Family Welfare. The BDHS followed all protocols prescribed by the WHO. Moreover, the data are widely used for public health issues. Thus, it was not necessary to get the ethical approval of the review board of Islamic University, Kushtia, Bangladesh.

Results

Table 1 shows the percentage distribution of women by their background characteristics. Of the women, 27% gave birth during their teens, 32% were 20 to 24 years old, 22% were 25 to 29 years old, and 19% were 35 years or older. Slightly more than a quarter had no formal education, 31% had some primary education, 36% had a secondary level of education, and 7% had higher education. In terms of wealth index, 22% were poorest, 21% were poorer, 19% were of medium status, 20% were richer, and 18% were richest. The largest group of women was from the Dhaka region, followed by Rajshahi, Chittagong, Khulna, Sylhet, and Barisal regions, respectively. More than three-fifths were from rural areas. Of the births reported, 31% were unintended. Exactly two-fifths of the women reported that they wanted more children, and 58% were current contraceptive users. Almost 32% of the births were first-order, 27% were second-order, 17% were third-order, 11% were fourth-order, and 13% were fifth- or higher-order pregnancies. Just 52% had visited at least once for ANC services.

Table 1.

Percentage Distribution of Women by Their Background Characteristics, BDHS 2007.

Characteristics	n^a	Percentage
Maternal age (years)
<20	1330	27.1
20-24	1569	32.0
25-29	1065	21.7
30-34	624	12.7
35-49	317	6.5
Education
No education	1282	26.2
Primary	1517	30.9
Secondary	1778	36.3
Higher	327	6.7
Wealth index
Poorest	1068	21.8
Poorer	1045	21.3
Middle level	932	19.0
Richer	958	19.5
Richest	902	18.4
Region
Barisal	313	6.4
Chittagong	1030	21.0
Dhaka	1556	31.7
Khulna	503	10.3
Rajshahi	1118	22.8
Sylhet	384	7.8
Residence
Urban	1039	21.2
Rural	3866	78.8
Wanted last child
Intended	3394	69.2
Unintended	1509	30.8
Want more children
No	2798	58.4
Yes	1925	40.1
Undecided	71	1.5
Current FP method user
No	2037	41.5
Yes	2868	58.5
Birth order
First	1566	31.9
Second	1314	26.8
Third	828	16.9
Fourth	541	11.0
Fifth+	656	13.4
Received ANC service
No	2368	48.3
Yes	2537	51.7
Total	4905	100.0

Abbreviations: BDHS, Bangladesh Demographic and Health Survey; FP, family planning; ANC, antenatal care.

The total number of women is not always equal to 4905 because of missing values.

Entry to ANC

Figure 1 shows the percentage distribution of women by the time of first antenatal checkup for their most recent pregnancy. As shown in the figure, 48.3% of the women did not receive any ANC services. Only 2.8% of the women received ANC services in the first month of their pregnancy, 7.5% received ANC services in the second month, 11.8% received ANC services in the third moth, and the rest received ANC services in the fourth month or later.

Figure 1.

Percentage distribution of women by the time (in months) at first antenatal checkup.

Table 2 shows the early and late entry to ANC visits among women who visited at least once for ANC services, by their background characteristics. As shown in the table, 43% of the women entered early to ANC, and 57% delayed in obtaining ANC services. All the available predisposing, enabling, environmental, and pregnancy-related factors were significantly associated with entry to ANC visits. Late ANC seeking was more prevalent among the young, women with no formal education, the poor, rural residents, women from the Rajshahi region, those with an intended pregnancy, those who desired more children, those who were current contraceptive users, and for third- and higher-order pregnancies.

Table 2.

Prevalence of ANC Seeking Among Women of Bangladesh by Timing and Their Background Characteristics, BDHS 2007.

	Timing of ANC Visits
Characteristics	Early Visit (n = 1080)	Late Visit (n = 1456)	Not Visited (n = 2369)	χ²
Maternal age (years)				84.23^a
<20	21.4	34.4	44.2
20-24	22.2	32.8	45.0
25-29	24.2	25.7	50.1
30-34	21.5	25.5	53.0
35-49	17.0	16.4	66.7
Women’s education				789.62^a
No education	10.3	18.4	71.3
Primary	16.5	27.4	56.1
Secondary	28.7	38.8	32.4
Higher	57.0	34.6	8.4
Wealth index				853.70^a
Poorest	8.5	22.3	69.2
Poorer	12.4	23.9	63.7
Middle	17.6	30.3	52.1
Richer	28.5	37.1	34.5
Richest	46.9	36.6	16.5
Region				67.73^a
Barisal	19.4	24.2	56.3
Chittagong	23.5	28.9	47.6
Dhaka	22.5	25.7	51.8
Khulna	25.4	37.1	37.5
Rajshahi	19.6	35.4	45.0
Sylhet	20.9	26.0	53.1
Residence				252.43^a
Urban	37.3	33.9	28.8
Rural	17.9	28.5	53.6
Wanted last child				32.24^a
Intended	22.6	31.7	45.7
Unintended	20.7	25.2	54.1
Want more children				69.43^a
No	20.0	27.1	52.9
Yes	25.3	34.1	40.6
Undecided	17.9	31.1	51.0
Current FP method user				10.39^b
No	20.6	28.4	51.0
Yes	23.0	30.6	46.4
Birth order				347.28^a
First	29.5	37.6	32.9
Second	23.0	30.3	46.7
Third	17.4	29.6	53.0
Fourth	17.8	21.2	60.9
Fifth+	11.3	16.6	72.1
Total	22.0	29.7	48.3

Abbreviations: ANC, antenatal care; BDHS, Bangladesh Demographic and Health Survey; FP, family planning.

Level of significance, P < .001.

Level of significance, P < .01.

Reasons for Not Receiving ANC Services

Table 3 shows the reasons for not receiving ANC for the recent births reported by the women. The women who did not receive ANC services were asked, “Why did you not receive ANC services?” To record the response to this question in the interview schedule, there were 13 closed-ended and coded options, such as too far, inconvenient service hour, unpleasant staff, service too expensive, and so on. From these options, the prompt response provided by the respondents to this particular question was recorded. The most reported reason as mentioned by the women was that “service was not needed” (59.3%), followed by “service was too expensive” (20.4%), “did not know the need of the service” (5.7%), “did not have permission from husband to go for ANC” (4.1%), “the facility place is too far” (3.3%), and “religious reason” (2.6%).

Table 3.

Reasons for Not Receiving ANC Services.

Reasons for Not Receiving ANC Services	n	Percentage
Too far	77	3.3
Inconvenient service hour	10	0.4
Unpleasant staff	9	0.4
Lack of experienced staff	2	0.1
Lack of privacy	12	0.5
Inadequate drug supply	5	0.2
Service too expensive	482	20.4
Religious reason	62	2.6
Service not needed	1404	59.3
Did not know the need	135	5.7
Did not have permission to go	98	4.1
Did not know place	32	1.4
Other reasons	57	2.4

Abbreviation: ANC, antenatal care.

Results of Multivariate Logistic Regression

Table 4 demonstrates the results of multivariate logistic regression analyses for ANC seeking among women of Bangladesh. Model I shows that the younger women were significantly more likely to receive ANC services than women aged 35 to 49 years. Women’s education was positively associated with ANC seeking. Model II shows that the effect of age remained unchanged with slight attenuation for women’s education. The women of Chittagong, Khulna, Rajshahi, and Sylhet regions were more likely to receive ANC services than those of Barisal region. As expected, the likelihood of ANC seeking was significantly lower among rural women than their urban sisters.

Table 4.

Logistic Regression Modeling for ANC Seeking Showing Odds Ratios (and 95% Confidence Intervals) by Background Characteristics.

Characteristics	Model I	Model II	Model III	Model IV
Maternal age (years)
<20	1.35 (1.02-1.79)^a	1.38 (1.03-1.85)^a	1.52 (1.12-2.05)^b	1.23 (0.89-1.71)
20-24	1.45 (1.10-1.91)^b	1.45 (1.09-1.93)^b	1.55 (1.16-2.08)^b	1.39 (1.03-1.88)^a
25-29	1.27 (0.96-1.69)^c	1.28 (0.95-1.71)^c	1.34 (0.99-1.81)^a	1.25 (0.92-1.69)
30-34	1.44 (1.06-1.95)^b	1.46 (1.07-2.00)^b	1.41 (1.02-1.95)^a	1.36 (0.98-1.88)^a
35-49	1.00	1.00	1.00	1.00
Women’s education
No education	1.00	1.00	1.00	1.00
Primary	1.88 (1.60-2.21)^d	1.86 (1.58-2.20)^d	1.60 (1.35-1.90)^d	1.55 (1.30-1.84)^d
Secondary	4.98 (4.21-5.88)^d	4.89 (4.13-5.80)^d	3.16 (2.63-3.79)^d	3.01 (2.49-3.62)^d
Higher	27.75 (18.28-42.13)^d	22.57 (14.79-34.45)^d	10.28 (6.82-15.96)^d	9.38 (6.02-14.62)^d
Region
Barisal		1.00	1.00	1.00
Chittagong		1.33 (1.01-1.75)^a	1.02 (0.77-1.36)	1.01 (0.75-1.34)
Dhaka		1.13 (0.87-1.48)	0.89 (0.67-1.16)	0.87 (0.66-1.14)
Khulna		1.91 (1.40-2.61)^d	1.57 (1.15-2.16)^b	1.55 (1.13-2.14)^b
Rajshahi		1.62 (1.23-2.14)^d	1.49 (1.13-1.97)^b	1.50 (1.14-1.99)^b
Sylhet		1.66 (1.20-2.30)^d	1.16 (0.83-1.63)	1.13 (0.80-1.59)
Residence
Urban		1.00	1.00	1.00
Rural		0.37 (0.31-0.43)^d	0.64 (0.53-0.77)^d	0.62 (0.51-0.75)^d
Wealth index
Poorest			1.00	1.00
Poorer			1.08 (0.89-1.31)	1.09 (0.90-1.32)
Middle			1.55 (1.27-1.89)^d	1.56 (1.28-1.90)^d
Richer			2.56 (2.08-3.14)^d	2.57 (2.08-3.16)^d
Richest			5.14 (3.95-6.70)^d	5.38 (4.11-7.05)^d
Wanted last child
Intended				1.19 (1.03-1.38)^b
Unintended				1.00
Want more children
No				1.00
Yes				1.23 (1.05-1.44)^d
Undecided				0.88 (0.50-1.53)
Current FP method user
No				1.00
Yes				0.95 (0.83-1.08)
Birth order
First				2.46 (1.86-3.25)^d
Second				1.47 (1.15-1.88)^d
Third				1.38 (1.08-1.77)^b
Fourth				1.38 (1.06-1.80)^b
Fifth+				1.00

Abbreviations: ANC, antenatal care; FP, family planning.

Level of significance, P < .05.

Level of significance, P < .01.

Level of significance, P < .10.

Level of significance, P < .001.

Model III in Table 4 reveals further attenuation of the effect of women’s education and place of residence on ANC seeking; however, the age effect became sharpened. It also showed regional variation of ANC seeking. The higher the wealth index, the higher the likelihood of ANC seeking. Model IV shows that maternal age, women’s education, region, place of residence, wealth index, pregnancy intention status, wanting more children, and birth order are significant determinants that influence ANC seeking. Being an adolescent age no longer had a significant effect on ANC seeking. Women with intended pregnancies who desired additional children and those with first-order pregnancies were significantly more likely to receive ANC services.

Table 5 shows the results of multivariate logistic regression analyses for late ANC entry. Model I showed that adolescents and young women were more likely to delay ANC seeking than their older counterparts. For instance, adolescents and younger women were significantly more likely to seek ANC services late compared with women aged 35 years or more. Education was significantly related to entry to ANC seeking. The lower women’s education, the more likely women were to seek ANC late. The odds of delay were 3-fold for women with no formal education, more than 2-fold for women with primary education, and 2-fold for women with secondary education compared with women who had higher education.

Table 5.

Logistic Regression Modeling for Late ANC Entry Showing Odds Ratios (and 95% Confidence Intervals) by Background Characteristics.

Characteristics	Model I	Model II	Model III	Model IV
Maternal age (years)
<20	1.83 (1.20-2.80)^a	1.63 (1.06-2.51)^a	1.54 (0.99-2.37)^b	1.96 (1.18-3.28)^a
20-24	1.73 (1.14-2.62)^a	1.66 (1.09-2.52)^a	1.59 (1.04-2.43)^b	1.77 (1.10-2.84)^a
25-29	1.30 (0.85-1.99)	1.28 (0.83-1.97)	1.24 (0.80-1.91)	1.21 (0.77-1.92)
30-34	1.33 (0.85-2.09)	1.27 (0.80-2.00)	1.27 (0.80-2.01)	1.25 (0.79-2.00)
35-49	1.00	1.00	1.00	1.00
Women’s education
No education	3.04 (2.21-4.19)^c	2.81 (2.03-3.90)^c	1.93 (1.36-2.76)^c	1.71 (1.16-2.50)^a
Primary	2.55 (1.91-3.39)^c	2.42 (1.81-2.25)^c	1.83 (1.34-2.49)^c	1.65 (1.19-2.29)^c
Secondary	1.99 (1.52-2.61)^c	1.88 (1.43-2.47)^c	1.65 (1.25-2.18)^c	1.55 (1.16-2.06)^c
Higher	1.00	1.00	1.00	1.00
Region
Barisal		1.00	1.00	1.00
Chittagong		1.06 (0.72-1.55)	1.23 (0.83-1.82)	1.19 (0.80-1.78)
Dhaka		1.01 (0.69-1.47)	1.17 (0.80-1.72)	1.16 (0.79-1.72)
Khulna		1.22 (0.80-1.84)	1.35 (0.88-2.05)	1.32 (0.86-2.02)
Rajshahi		1.45 (0.99-2.13)^b	1.52 (1.03-2.23)^b	1.51 (1.02-2.24)^b
Sylhet		0.94 (0.59-1.48)	1.13 (0.71-1.80)	1.09 (0.68-1.75)
Residence
Urban		1.00	1.00	1.00
Rural		1.51 (1.26-1.81)^c	1.18 (0.96-1.45)^d	1.21 (0.98-1.50)^b
Wealth index
Poorest			2.40 (1.72-3.36)^c	2.51 (1.78-3.55)^c
Poorer			1.79 (1.32-2.43)^c	1.81 (1.32-2.47)^c
Middle			1.69 (1.27-2.20)^c	1.67 (1.26-2.21)^c
Richer			1.36 (1.07-1.72)^a	1.33 (1.05-1.70)^a
Richest			1.00	1.00
Wanted last child
Intended				1.00
Unintended				1.18 (0.97-1.43)^d
Want more children
No				1.00
Yes				0.99 (0.79-1.25)
Undecided				1.58 (0.75-3.31)
Current FP method user
No				1.02 (0.85-1.21)
Yes				1.00
Birth order
First				1.00
Second				1.19 (0.93-1.54)
Third				1.68 (1.20-2.36)^c
Fourth				1.10 (0.72-1.67)
Fifth+				1.41 (0.87-2.28)^d

Abbreviations: ANC, antenatal care; FP, family planning.

Level of significance, P < .01.

Level of significance, P < .05.

Level of significance, P < .001.

Level of significance, P < .10.

Model II showed that the inclusion of environmental factors substantially attenuated the effect of predisposing factors. As compared with women of Barisal division, the women of Rajshahi division delayed significantly more when seeking ANC services. However, there was no significant variation in the likelihood of ANC seeking between the women of Barisal division and those of other administrative regions of the country. Urban women were significantly (P < .001) faster as compared with their rural counterparts in seeking ANC. Model III demonstrates that the inclusion of enabling factors further attenuated the effect of predisposing factors and slightly sharpened the effect of environmental factors. The lower the wealth indexes, the higher the risk of late entry to ANC seeking. For instance, women from the poorest wealth quintile were more than 2 times as likely to enter ANC late as women from the richest quintile.

Model IV shows that all the predisposing, environmental, enabling, and pregnancy-related variables are important determinants of time variation of ANC seeking. The inclusion of pregnancy-related factors increased the effect of maternal age and attenuated the effect of women’s education on ANC seeking, whereas the effect of environmental and enabling factors remained unchanged. Women who reported that their pregnancy was unintended sought ANC significantly later than those who reported their pregnancy to be intended. Women with third and fifth births were much more likely to attend ANC later than recommended compared with those having a first birth.

Discussion and Conclusion

In this study, we examined ANC seeking and the timing of receiving ANC services among women of Bangladesh using cross-sectional and nationally representative BDHS data. The findings showed that 52% of the most recent pregnant women sought skilled ANC from medically trained personnel. Of the women who sought skilled ANC, 57% initiated attendance after the recommended time for first ANC visit, and 43% initiated ANC seeking in due time. When the months were broken down into trimesters, the figure revealed that 22.0% of the women received ANC in their first trimester, 20.0% received ANC services in the second trimester, and 9.7% received ANC in the third trimester of pregnancy. Of the women who did not receive skilled services for ANC, about three-fifths reported that “ANC services were not needed” and one-fifth reported that “the services were too expensive.” This suggests that in Bangladesh, maternal health care services are not yet easily accessible for financial reasons.

Consistent with earlier studies^21,22 in developing countries, our findings revealed that although young women were more likely to receive ANC services, adolescents and young pregnant women attended ANC later than their older counterparts. A possible reason for this may be adolescent and young women’s lack of knowledge, power to make decisions, and financial stability as well as concepts of illness. Younger women, especially teenagers, are more likely to have unplanned pregnancies and hence face a scarcity of resources to access ANC services. In contrast, some studies found no significant differences in the timing of entry to ANC seeking among adolescents and adults.^7,23 The lack of studies in Bangladesh regarding the late entry to ANC among adolescents and the young as compared with their older counterparts points to the need for further investigation.

Predisposing factors, such as women’s education, have significant effects on both ANC seeking and the timing of ANC seeking. As expected, the lower the education women have, the more likely it is that ANC will be sought late or not at all. Some studies have reported an association between maternal education and receiving ANC services and the timing of ANC initiation. Our findings are in line with the findings of these earlier studies.^21,24 Contrary to these studies, there was no evidence of such an effect in some studies.⁷ Education is the viable means of gathering knowledge. It is plausible that educated women are better informed and, therefore, more aware of the importance of seeking ANC and other maternal health care services early compared with women with low or no formal education.

Environmental factors such as residence and region were also significantly associated with both the outcome variables. As expected, rural women were less likely to seek ANC services and more likely to enter ANC late compared with their urban counterparts. There was also regional variation in the entry to and receiving of ANC. It is likely that rural women as compared with their urban sisters are not as well educated and poorer, resulting in lower likelihood of entering or late entry to ANC. Variation in the likelihood of ANC in the regions and rural-urban residence are also associated with accessibility, socioeconomic status, and culture and tradition. Many of the sociodemographic and psychosocial characteristics are also associated with late ANC among rural women. Poverty, time management, difficult access, and negative perceptions may be resulting in delay in entry to ANC among rural women and women from some regions of the country.

The enabling factor—the wealth quintile—was found to be significantly positively related to receiving ANC and early entry to ANC. The explanation of this finding is straightforward. Resource scarcity, lack of adequate knowledge regarding the advantage of maternal health care seeking, and lack of autonomy are closely associated with poverty, resulting in lower likelihood of receiving and late entry to ANC services among the poor. Financial barriers posed by high out-of-pocket costs for consultations may be an important determinant of utilization of maternal health services, such as ANC, particularly in areas where poverty is high and for services that are more costly.²⁵ Thus, financing of health service utilization through insurance or other schemes may reduce these barriers.²⁶

Our findings showed that unintended pregnancy and multiparity were associated with ANC seeking and also late entry to ANC services utilization. Studies showed that unintended pregnancies are higher among women who were of lower economic status, at an early or late age of reproductive life, not using contraceptives consistently, and attending formal education.²⁷ Horror, excitement, or pleasure at the discovery of being pregnant for the first time lead to early ANC service utilization. Our findings confirm the findings of previous studies^28,29 that show that unintended pregnancies are associated with inadequate ANC, although our finding was attenuated in the multivariate analyses. Although pregnancy intention status was a significant factor in receiving ANC services, it was no longer significantly associated with late entry to ANC.

The study has several limitations. This was a retrospective, cross-sectional data analysis, and there might have been underreporting of age, times and timing of receiving ANC, and other independent exposures. First, in developing countries such as Bangladesh, where vital statistics are not properly maintained, there might have been errors in reports of maternal age. The underreporting may bias the findings of the study. Thus, the findings of this study should be interpreted with caution. The main limitation of this study is that several potentially important factors, such as distance to health facilities, were not included for analyses because of unavailability of information. Despite the limitations, the strength of the study is that it used a population-based and nationally representative large data set.

ANC is vital to achieving good maternal care outcomes in developing countries such as Bangladesh.³⁰ Financial scarcity and lack of knowledge regarding the importance of service utilization are important barriers to receiving and early entry to ANC. Our findings suggest that interventions that lower financial barriers to receiving and entering ANC are likely to have an impact. Barriers related to education have an independent effect on ANC seeking and late entry to ANC. These barriers should also be addressed as part of health interventions through information, education, and communication programs. Efforts should be made to design interventions that increase early entry to ANC seeking. Promotion of early and frequent ANC utilization through community-based interventions could potentially be scaled up at low cost and adapted to local needs. Research elsewhere reveals that quality of care influences women’s decisions to receive maternal health care services.⁷ Thus, at the same time, the quality of ANC services needs to be improved to attract women to use medical care throughout pregnancy, birth, and the postpartum period; outreach services should be offered on a regular basis to bring services closer to women living in very distant settlements.⁷

Footnotes

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

References

Carroli

Rooney

Villar

. How effectiveness is antenatal care in preventing maternal mortality and serious morbidity? An overview of the evidence. Paediatr Perinat Epidemiol. 2001;15(suppl 1):s1-s42.

Wehby

Murray

Castilla

Lopez-Camelo

Ohsfeldt

. Prenatal care effectiveness and utilization in Brazil. Health Policy Plan. 2009;24:175-188.

Heaman

Newburn-Cook

Green

Elliott

Helewa

. Inadequate prenatal care and its association with adverse pregnancy outcomes: a comparison of indices. BMC Pregnancy Childbirth. 2008;8:15.

Robinson

Comino

Forbes

Webster

Knight

. Timeliness of antenatal care for mothers of Aboriginal and non-Aboriginal infants in an urban setting. Aust J Prim Health. 2012;18:56-61.

World Health Organization (WHO), UNICEF, UNFPA, The World Bank. Trends in Maternal Mortality: 1990 to 2008. Estimates Developed by WHO, UNICEF, UNFPA and the World Bank. Geneva, Switzerland: WHO; 2010.

Main

Lower

James

Rouse

. Changes in ante-natal care and family planning in Krakor, Pursat, Cambodia, 1996-1998. Asia Pac J Public Health. 2001;13:74-78.

Gross

Alba

Glass

Schellenberg

Obrist

. Timing of antenatal care for adolescent and adult pregnant women in south-eastern Tanzania. BMC Pregnancy Childbirth. 2012;12:16.

Mpembeni

Killewo

Leshabari

. Use pattern of maternal health services and determinants of skilled care during delivery in Southern Tanzania: implications for achievement of MDG-5 targets. BMC Pregnancy Childbirth. 2007;7:29.

Abou-Zahr

Wardlaw

. Antenatal Care in Developing Countries: Promises, Achievements and Missed Opportunities: An Analysis of Trends, Levels and Differentials, 1990-2001. Geneva, Switzerland: World Health Organization; 2003.

10.

Trinh

LTT

Rubin

. Late entry to antenatal care in New South Wales, Australia. Reprod Health. 2006;3:8.

11.

National Institute for Health and Clinical Excellence. Antenatal Care: Routine Care for Healthy Pregnant Women. London, UK: National Institute for Health and Clinical Excellence; 2003.

12.

American Academy of Pediatrics. Guidelines for Perinatal Care. Washington, DC: American College of Obstetricians and Gynecologists; 2002.

13.

Villar

Bergsjo

. WHO Antenatal Care Randomized Trial: Manual for the Implementation of the New Model. Geneva, Switzerland: World Health Organization; 2002.

14.

World Health Organization. Antenatal Care. Report of a Technical Working Group, 1994. Geneva, Switzerland: WHO; 1994. WHO/FRH/MSM/968.

15.

Beeckman

Louckx

Putman

. Determinants of the number of antenatal visits in a metropolitan region. BMC Public Health. 2010;10:527.

16.

Raatikainen

Heiskanen

Heinonen

. Under-attending free antenatal care is associated with adverse pregnancy outcomes. BMC Public Health. 2007;7:268.

17.

Krueger

Scholl

. Adequacy of prenatal care and pregnancy outcome. J Am Osteopath Assoc. 2000;100:485-492.

18.

Collin

Anwar

Ronsmans

. A decade of inequality in maternity care: antenatal care, professional attendance at delivery, and caesarean section in Bangladesh (1991-2004). Int J Equity Health. 2007;6:9.

19.

Andersen

. Revisiting the behavioral model and access to medical care: does it matter. J Health Soc Behav. 1995;36:1-10.

20.

National Institute of Population Research and Training (NIPORT), Mitra and Associates, and Macro International. Bangladesh Demographic and Health Survey 2007. Dhaka, Bangladesh: National Institute of Population Research and Training, Mitra and Associates, and Macro International; 2009.

21.

Magadi

Agwanda

Obare

. A comparative analysis of the use of maternal health services between teenagers and older mothers in sub-Saharan Africa: evidence from Demographic and Health Surveys (DHS). Soc Sci Med. 2007;64:1311-1325.

22.

van Eijk

Bles

Odhiambo

. Use of antenatal services and delivery care among women in rural western Kenya: a community based survey. Reprod Health. 2006;3:2.

23.

Atuyambe

Mirembe

Tumwesigye

Annika

Kirumira

Faxelid

. Adolescent and adult first time mothers’ health seeking practices during pregnancy and early motherhood in Wakiso district, central Uganda. Reprod Health. 2008;5:13.

24.

Ochako

Fotso

Ikamari

Khasakhala

. Utilization of maternal health services among young women in Kenya: insights from the Kenya Demographic and Health Survey, 2003. BMC Pregnancy Childbirth. 2011;11:1.

25.

Agha

Carton

. Determinants of institutional delivery in rural Jhang, Pakistan. Int J Equity Health. 2011;10:31.

26.

Wagstaff

. Poverty and health sector inequalities. Bull World Health Organ. 2002;80:97-105.

27.

Kassa

Berhane

Worku

. Predictors of unintended pregnancy in Kersa, Eastern Ethiopia, 2010. Reprod Health. 2012;9:1.

28.

Magadi

. Maternal and child health among the urban poor in Nairobi, Kenya. African Popul Stud. 2004;19(SB):s179-s198.

29.

Cheng

Schwarz

Douglas

Horon

. Unintended pregnancy and associated maternal preconception, prenatal and postpartum behaviors. Contraception. 2009;79:194-198.

30.

Shen

Chan

. Regional disparities in prenatal care services in rural China. Asia Pac J Public Health. 2011;23:682-689.