Maternal health care in vietnam demand for antenatal care and choice of delivery care services

MATERNAL HEALTH CARE IN VIETNAM: DEMAND FOR ANTENATAL CARE AND CHOICE OF DELIVERY CARE SERVICES By Nguyen Thi Hoai Trang A Thesis Submitted in Partial Fulfilment of the Requirements fo

MATERNAL HEALTH CARE IN VIETNAM: DEMAND FOR

ANTENATAL CARE AND CHOICE OF DELIVERY CARE

SERVICES

By Nguyen Thi Hoai Trang

A Thesis Submitted in Partial Fulfilment of the Requirements for the Degree of

Master of Art in Development Economics

Academic Supervisor: Dr Truong Dang Thuy

HO CHI MINH CITY, June 2016

DECLARATION

“This is to certify that this thesis entitled “MATERNAL HEALTH CARE IN VIETNAM:

DEMAND FOR ANTENATAL CARE AND CHOICE OF DELIVERY CARE SERVICES”,

which is submitted by me in fulfillment of the requirements for the degree of Master of Art in Development Economics to the Vietnam – The Netherlands Programme (VNP) The thesis constitutes only my original work and due supervision and acknowledgement have been made in the text to all materials used

HCMC, June 06th, 2016

Nguyen Thi Hoai Trang

ACKNOWLEDGEMENT

I would like to acknowledge my supervisor, Dr Truong Dang Thuy for his great

contribution to my thesis Without his support, my thesis would be not possible By his large knowledge and experiences, he gave me the informative comments and enabled me to understand

my work better I would like to express my sincere gratitude to his guidance and encouragement, which make me stronger to overcome the challenges and fulfill my work completely

By this chance, I would like to express my appreciation toward all lecturers of the Vietnam – Netherlands Program who have provided with valuable economic knowledge during my study

in this program Next, I wish to thank to all my friends here at VNP- MDE 19, who share unforgettable memories in studying together

Finally, I would like to express my deep gratitude to my family for their support and endurance when I pursue my postgraduate studies

ABSTRACT

This thesis research aims to analyze the impact of individual characteristics, household characteristic and communities in utilization of maternal health care services in Vietnam Using the latest data of Vietnam’s Multiple Indicator Cluster Survey 2013-2014, it employs the Negative Nominal Model for demand of prenatal care visits and Multinomial Logistic Model for the choice

of delivery facility With respect to the demand of prenatal care visits, the result shows that higher education, higher age, exposure to mass media and no religion increase the number of prenatal care visits while higher birth order, unmarried or separated status, ethnicity group and lower household wealth index decrease the number of prenatal care Moreover, living in rural, disadvantaged areas and the community with higher illiteracy rate decrease the demand of prenatal care visits while living in the community with higher proportion of women giving birth at health facilities increase the demand Concerning the choice of delivery facility, more prenatal care visits and exposure to mass media are positively associated with the choice of giving birth at public hospital In contrast, suffering the burden of taking care more children, lower household wealth index, living in rural and the community with higher illiteracy ratio adversely affect the choice of public hospital delivery The results suggest the improvement of maternal health program in rural and underdeveloped areas as well as universal education over the country, especially for the ethnic minority group

Keywords: prenatal care visits, the place of childbirth, individual characteristics, household

characteristics, community characteristics, Vietnam

Contents

DECLARATION i

ACKNOWLEDGEMENT ii

ABSTRACT iii

LIST of TABLES and FIGURES vii

ABBREVIATION viii

CHAPTER I 1

INTRODUCTION 1

1.1 Problem statement 1

1.2 Research objectives 3

1.3 Research questions 4

1.4 Structure 4

CHAPTER II 5

LITERATURE REVIEW 5

2.1 The role of maternity health care 5

2.2 Overview of maternal health and health care in Vietnam 6

2.2.1 The culture 6

2.2.2 The two-child policy 6

2.2.3 Maternal mortality ratio and maternal health care in Vietnam 7

2.3 The demand for health care 11

2.3.1 Theoretical background 11

2.3.2 Empirical Literature Review 13

2.4 The choice of health care provider 19

2.4.1 Theoretical background: 19

2.4.2 Empirical literature review 20

CHAPTER III 23

METHODOLOGY AND DATA DESCRIPTION 23

3.1 Conceptual framework 24

3.2 Empirical framework 25

3.2.1 Demand for Prenatal care 26

3.2.2 Choice of birth delivery facility 27

3.3 Data 28

3.4 Variables definition 28

3.4.1 Dependent variables 28

3.4.2 Independent variables 29

RESULTS AND DISCUSSIONS 31

4.1 Descriptive Results 32

4.2 Analysis of Demand for prenatal care 34

4.2.1 Bivariate analysis 34

4.2.2 Analysis of Negative Binomial Model 37

4.3 Analysis of Choice in the delivery care providers 41

4.3.1 Bivariate analysis 41

4.3.2 Analysis of Multinomial Logistic Model 44

CHAPTER V 47

CONCLUSION, RECOMMENDATION and LIMITATION 48

5.1 Main findings 48

5.2 Policy Recommendation 49

5.3 Limitation and Further Research 50

REFERENCE 51 APPENDIX 56 STATA RESULTS 71

LIST of TABLES and FIGURES List of Tables

Table 1: Description of Variables 30

Table 2:Descriptive Results – Numeric Variables 33

Table 3 : Descriptive Results - Dummy Variables 33

Table 4: Bivariate analysis in the demand of prenatal care visits 35

Table 5: Negative binomial regression for the demand of prenatal care visits 40

Table 6 : Bivariate analysis in the choice of delivery care providers - numeric independent variables 41

Table 7:Bivariate analysis in the choice of delivery care provider – dummy independent variables 43

Table 8: Multinomial Logistic Regression for the choice of delivery care provider 46

Table 9: Marginal effects for the choice of delivery care provider 47

List of Figures Figure 1: MMR in Vietnam in the period of 2000 – 2015 8

Figure 2: MMR of the Asian countries in the period of 2000 – 2015 8

Figure 3: Percentage of women having at least 1 visit and at least 4 visits during pregnancy 9

Figure 4: The percentage of the women taking antenatal care visits by residence in 2011 and 2014 10

Figure 5: The percentage of the women taking antenatal care visits by ethnicity in 2011 and 2014 10

Figure 6 The association between individual level, household level and community level characteristics with the utilization of maternal health care services 25

Figure 7: The association between the demand of maternal care visits and numerical independent variables 37

ABBREVIATION

GSO General Statistics Office

IMR Infant Mortality Ratio

MMR Maternal Mortality Ratio

WHO World Health Organization

CHAPTER I INTRODUCTION

1.1 Problem statement

There is a growing concern about the maternal health care globally, especially in low income countries World Health Organization (WHO 2014) reported that the global maternal mortality ratio (MMR) in 2013 was 210 maternal deaths per 100 000 live births, decreasing from 380 maternal deaths per 100 000 live births in 1990 However, the ratio in developing regions was 14 times higher than in developed regions Even though maternal death is generally decreasing worldwide, it has yet to achieve the target of Millennium Development Goal 5 by reducing the MMR by three quarters between 1990 and 2015 (WHO 2014)

The maternal death has direct causes and indirect causes The direct cause results from arising complications during pregnancy, delivery and postpartum, or improper treatment such

as hemorrhage, infection, obstructed labor, unsafe abortion, ectopic pregnancy and related deaths while the indirect cause results from the disease which previously exists or be not due to indirect obstetric causes like hepatitis anemia, malaria, heart disease and tetanus

anesthesia-(WHO 2005) It was reported that direct causes made up the higher number of maternal death than indirect causes with 80% of the total MMR (WHO 2005)

These complications could be preventable thanks to the intervention of health care such

as antenatal care and delivery care, which was introduced by WHO in the safe motherhood package in 1994 (Tran 2012) Antenatal cares provide the opportunities to pregnancy women and their family to be informed of their health and the growth status of unborn baby Low birth weights could be prevented if the pregnant women are well acknowledged about their unborn baby’s weight and height during the antenatal care and then improve their diet In addition, antenatal check-ups detect the danger signs and risks of pregnancy and delivery and make timely interventions For example, tetanus immunization in the antenatal care period is vital to save the life of the women and their baby The management of high blood pressure during pregnancy ensures the maternal health and increase the infant survival (WHO and UNICEF 2003) Furthermore, delivery care also plays an important role in reducing maternal deaths WHO recommended the child birth at health facility or attended by skilled health staffs to ensure to the safe delivery and give birth to healthy baby With good hygiene and adequate medical equipment, the delivery at facility could decrease the complications arising from the

labor such as hemorrhage, obstructed labor In addition, skill health professionals are available

in the facilities ensure safe delivery and provide proper emergency management (Tran 2012)

In pursuit of Millennium Development Goal 5 “Improving maternal health”, Vietnam also is making progress in improving the maternal health with the drop of maternal mortality ratio The World Bank shows that MMR in Vietnam has remarkable improvements in last 15 years in decrease from 81 deaths per 100,000 live births in 2000 to 54 per 100,000 in 2015 The access to antenatal care, an important period for health of pregnant women and their baby and delivery service has also increased Multiple indicator cluster survey in 2014 (MICS 5) shows that the percentage of women aged 15-49 with a live birth in the last two years who received antenatal care at least once is 95.8 per cent nationwide However, there are considerable disparity in maternal mortality ratio and utilization of maternal health care among ethnicity group, place of residence and the regions where the pregnant women are living MPI

2015 reported that maternal mortality in mountainous areas is more than three times higher than in lowland areas Furthermore,MICS5 indicates that the times of prenatal care visits differs among the women living rural and urban area, especially regarding having more than 4 visits In addition, the ethnic minority groups get more disadvantage of access to maternal health care with 79% of those having 1 visit and 32.7% of those having at least 4 visits compared to 99.2% and 82.1 % of the Kinh group as shown Therefore, growing disparities in health outcomes and health care utilization have posed a great challenging in recent years

The above challenges lead to several studies in the utilization of maternal health care

in Vietnam Most of them focused on the influence of demographic and socioeconomic factors

(Sepheri et al 2008, Tran et al 2011, Goland et al 2012, Malqvist et al 2012, Malqvist et al

2013) Demographic factors which were shown to increase the probability of the health services are younger age, low birth order while the factors reported to decrease the probability are separated or unmarried status, unintended pregnancy In addition, socio-economic factors make greater influence on the use of the maternal health care services Higher education level of a

woman is the most important determinant reported in the previous studies (Sepheri et al 2008, Tran et al 2011, Goland et al 2012, Malqvist et al 2012, Malqvist et al 2013, Wakabayashi

2014) Lower household income is also shown to be a strong factor in the likelihood of using

maternal health care (Sepheri et al 2008, Goland et al 2012) Some studies emphasized the

disparity in the maternal health care utilization among ethnic majority and minority groups

(Malqvist et al 2012, Malqvist et al 2013) On the other hand, major equity in rural and urban

areas also was identified by Tran et al (2011) and Sepheri et al (2008) pointed out the regional

disparity in the availability and accessibility to the maternal health care in Vietnam

However, most of them overlooked the community factors, except Sepheri et al (2008)

estimating the impacts of poverty rate The omission of the factor may bias the influence on maternity health care utilization, especially for the disadvantaged women The omission of

community effects may result in biased estimates of the roles of factors (Singh et al 2014)

Community beliefs and norms encourage or prevent health care seeking behaviors of the women In addition, community economic development may influence the access to health services and indirectly increase decision making power of women and positive attitudes toward

health service use (Stephenson et al 2006) The community level factors reported to be strong

indicators could be the poverty rate, the proportion of women in the community with higher education and the proportion of women delivering their child in health facility The poverty rate was negatively associated with the probability of taking antenatal care (Gage & Calixte

2006, Sepehri et al 2008, Ononokpono et al 2013, Singh et al 2014,) and facility delivery

(Gage & Calixte 2006, Sepehri et al 2008) whereas the high proportion of women with higher education and higher proportion of women choosing facility for delivery were positively

associated with the utilization of maternal health care (Stephenson et al 2006, Gage 2007, Ononokpono et al 2013, Singh et al 2014)

Therefore, there is a need to properly investigate the determinants on health care service utilization including individual level, household level and community level characteristics Using the latest dataset from Vietnam Multiple indicator cluster survey in 2014 (MICS 5), this study applies the Poisson Model to estimate the impact of social determinants on the demand for prenatal care visits, and the Multinomial Logistic Model to measure the association between social factors and choice on delivery care providers

1.2 Research objectives

The overall objective of this thesis research is twofold First, we analyze the demand of prenatal health care Particularly, we examine the determinants of the number of antenatal care visits of women taking during the last two years using data from MICS 5 Second, we investigate the choice on service facility for delivery of pregnant women in Vietnam

1.3 Research questions

To investigate the above objectives, the following questions need to be answered thoroughly:

Question 1: What are the determinants of the demand for antenatal care visits?

Question 2: What are the determinants of the choice on delivery care provider?

1.4 Structure

The paper is organized as follows In Chapter 2, the general theories related to the demand of health care and the choice of health care provider are discussed and the previous studies on the impact of social determinants on utilization of maternal health care are reviewed Chapter 2 presents the conceptual framework and the methodology with two applied methods using the dataset namely Multiple Indicator Cluster Survey 2013-2014 (MICS5) In the same chapter, methods investigating the determinants of the demand of prenatal health care and the choice of service facility for delivery will be presented The results and discussion are presented

Chapter 4 Chapter 5 concludes the paper and discusses policy implications

CHAPTER II LITERATURE REVIEW

This chapter presents the theoretical review and empirical review regarding the demand for prenatal care visits and the choice of facility for delivery The first part is to provide the role of maternal health care and the overview of maternal health care in Vietnam The next part

is to present the theoretical background for the demand for health care services, and the choice

of health care facility and their determinants The final section reviews the determinants affecting the demand of prenatal care visits and the choice of delivery care providers reported

in the previous studies in developed countries and developing countries, especially in Vietnam

2.1 The role of maternity health care

Motherhood is a positive experience which a women encounters; however, there are some health problems happening during pregnancy, childbirth, and the postpartum period The consequences impact seriously not only on the women’s health but also on the babies Three quarter of maternal deaths is reported to occur during childbirth and the postpartum period However, antenatal care and delivery care can prevent these complications

The introduction of antenatal care (ANC) began in the early 1900s with the aim of helping pregnant women and the unborn healthy and detecting adverse conditions in order to take timely interventions For example, the antenatal care will enable the women to enhance the understanding of the fetal growth and her health status Adverse outcome such as low birth weights can be avoidable by improving the women’s nutritional status In addition, the women will be informed about the risks during pregnancy and delivery The World Health Organization (WHO) recommended at least one visits to skilled health providers or at least four ANC visits to any providers According to the guidelines of WHO, the ANC program includes assessment of mother and fetus like body weight and height measurement, blood pressure, urine and blood tests; medical provisions like tetanus vaccination and supplement of iron and foliate and health consulting and education

The delivery at health facilities also plays an important role to ensure women to go through childbirth safely and deliver healthy infants In fact, proper medical technology and hygienic conditions during delivery can prevent complications and infections leading to morbidity and mortality of mother and her child In addition, skilled birth attendants are available in most health facilities According to WHO, a skilled birth attendant is defined as a midwife, doctor or nurse, who is well skilled to ensure normal childbirth and the postnatal

mediation, and who can detect complications and provide necessary emergency management WHO recommended that in countries with very high MMR, at least 60% of child deliveries should be assisted by skilled birth attendants by 2015 During 2005-2010, it was reported that 69% of women giving birth were attended by skilled birth staffs (Tran, 2012)

2.2 Overview of maternal health and health care in Vietnam

2.2.1 The culture

Vietnam culture is highly influenced by Confucianism, especially in the north region

of the country According to the tradition of Confucianism, the son of the family will inherit the family resources and worship the ancestors In addition, he has responsibility for taking care of the family members and maintains the continuity of the family line Therefore, giving birth to a son brings a proud to the family and improves the status of the women in the family However, it will put high pressure of having a son on the women in the family, especially for the women with daughters The strong preference for sons is the main reason for increasing sex ratio at birth

Moreover, male members are considered as the breadwinners and main income earners

in the family and have strong decisions on the family affairs whereas female members are considered as vulnerable members when their life is determined by her parents After getting married, they will usually live with their husband family and the income will probably be controlled by the parent-in-law and their husband The strong Confucianism and existing hierarchism limits their autonomy and their independent decisions for their lives, especially their health For example, the childbirth experience from the mother and the mother-in-law highly influence on the maternity care of young women and could prevent the women to seek essential maternal health care

2.2.2 The two-child policy

The two-child policy in Vietnam was issued in late 1980s with restriction of the maximum number of children per household Vietnam government practiced some family planning measures to reach the goal such as providing the free birth control devices as well as prosing the facilities for those who was allowed for abortions In addition, the family who did not comply with the two-child policy, they would be penalized in different ways like paying high fee For the government staffs who broke the policy, they would be prolonged the salary raise or reducing to lower position Therefore, due to fear of penalization and discredit, some women conceal their pregnancy and do not take adequacy of maternal health cares In addition,

more children put heavy burden on shoulders of women and create time and financial constraints to utilization of maternal health care services After many revision of the policy, the revised 2009 Population Ordinance is the effective prevailing policy in Vietnam According

to the Ordinance, couples have right to decide the time of having children and the birth spacing but are allowed to have one or two children Thanks to the population policy, the total fertility rate dropped from 2.55 in 2001 to 1.99 in 2011 (GSO 2013) suggesting the two-child policy was successful to ensure the steady population growth However, there are some challenges for the government to take account for The contraceptive methods had been not conducted effectively IUD is the most popular method applied in most families; however, it comes with the side effects so many women hesitate to use it In addition, Vietnam has faced the high abortion rate, especially among the young age It could be mainly attributed to lack of knowledge about the contraceptive methods and son preferences beside other reasons such as financial problems or health status

2.2.3 Maternal mortality ratio and maternal health care in Vietnam

In Vietnam, the government recommends pregnant women should have at least three prenatal visits during the pregnancy to detect and prevent the risks negatively affecting the health of mother and baby The content of prenatal care includes blood pressure measurement, urine testing, blood testing and measure of weight and height In addition, the national guidelines suggest that pregnant women should deliver the baby at health facilities Proper medical attention and hygienic condition at health facilities reduce the complications occurring

in and after the childbirth In addition, for the complications, Caesarean section is required but should be performed by the skilled obstetric doctors to ensure safe childbirth During the postpartum period the guidelines recommend at least two health checkups for both mother and child

The maternal mortality ratio is the ratio of women who die from causes related to pregnancy and childbirth in the period from pregnancy until 42 days after delivery, per 100,000 children born during the study period The data from the World Bank shows that MMR in Vietnam has remarkable improvements in last 15 years in decrease from 81 /100,000 in 2000

to 54/100,000 in 2015 The target of Millennium Development Goal 5 with MMR of 58.3 per 100,000 live births by 2015 has been achieved completely However, when comparing to other Asian countries, Vietnam still left behind the developed countries such as Singapore, Malaysia

and Thailand Therefore, Vietnam should make greater effort to reduce the MMR and ensure the persistent population growth

Figure 1: MMR in Vietnam in the period of 2000 – 2015

Source: The World Bank

Figure 2: MMR of the Asian countries in the period of 2000 – 2015

81

76

72 68 64

Myanmar Philippines Vietnam

Source: The World Bank

In order to reduce MMR and IMR, WHO recommends each pregnant woman should

have at least four prenatal care visits or at least one prenatal care attended by professional health

staffs It is widely agreed that antenatal care is vital to provide the information to women and

her families about potential risks during pregnancy and childbirth Vietnam has made

significant improvements in antenatal care coverage over the last years The figure 3 shows that 95.8 % of pregnant women have at least 1 prenatal care visits in 2014, which is higher than

the ratio in 19997 However, the percentage of those receiving more than 4 visits is still low at 73.7% This is a challenge against which the country should take more measures In addition,

there are considerable disparity in utilization of maternal health care among ethnicity group,

place of residence and the region where the pregnant women are living Figure 4 indicates that there is greater difference in the times of prenatal care visits among the women living rural and

urban area, especially regarding having more than 4 visits In addition, the ethnic minority

group get more disadvantage of access to maternal health care with 79% of those having 1 visit

and 32.7% of those having at least 4 visits compared to 99.2% and 82.1 % of the Kinh group

Figure 4: The percentage of the women taking antenatal care visits by residence in 2011 and 2014

Sources: MICS3 and MICS4

Figure 5: The percentage of the women taking antenatal care visits by ethnicity in 2011 and 2014

Sources: MICS3 and MICS4

Antenatal care visits by ethinicity

Kinh Non Kinh

2.3 The demand for health care

2.3.1 Theoretical background

Economists became interested in the health seeking behavior in the late 1960s and investigating the factors influencing the behavior The major contributions were by Grossman (1972), Rosenstock (1974), Thaddeus and Maine (1994) and Andersen (1995)

Grossman (1972) argued that what individual demand when purchasing health care is not health care but good health Firstly, he set up the model of demand for health, in which the individual has positive utility of consumption goods and negative utility of sick time 𝑡𝑠(𝐻)

(Zweifel et al 2009)

In addition, the health stock H changes overtime Therefore, the health capital

decrease at the rate 𝛿 However, the individual can increase health capital by investing in I In

a two-period model, the current health could be specified as

Where, U is utility, H is stock of health capital, 𝛿 is the rate of the depreciation of the heath stock, w is wage rate, X is consumption goods, M is the consumed amount of medical services, I is investment in health, 𝑡𝑠 is sick time and 𝑡𝐼 is time invested in favor of health Grossman (1972) argued that the individual demand for health for two reasons: first, as investment commodity and second, consumption commodity

Basing on the function 𝐼(𝑀, 𝑡𝐼) and 𝑡𝑠(𝐻1) from (1.1) and (1.2), Grossman constructed the demand function for medical services in investment model:

ln 𝑀 = 𝑐𝑜𝑛𝑠𝑡 − (1 + 𝛼𝑀(𝜀 − 1))𝑙𝑛𝑝 + (1 + 𝛼𝑀(𝜀 − 1))𝑙𝑛𝑤 − (1 − 𝜀)𝛼𝐸𝐸 (1.3) Where, 𝛼𝑀 is the production elasticity of medical services and 𝛼𝐸 is the effectiveness

of education E; 𝜀 is the marginal efficiency of health capital 𝐻1

From the utility function (1.1) including sick time and consumption good, he constructed the demand function for medical services in consumption model:

ln 𝑀 = 𝑐𝑜𝑛𝑠𝑡 − (1 + 𝛼𝑀(𝜅 − 1))𝑙𝑛𝑝 + (1 − 𝜅)(1 − 𝛼𝑀)𝑙𝑛𝑤 − (1 − 𝜅)𝛼𝐸𝐸 − 𝜅𝑙𝑛𝜆

(1.4)

From the investment model (1.3) and consumption model (1.4), the demand of health services depends on the price of medical services, the wage rate, education and wealth It could

be seen that the price of medical services decrease the optimum quantity of 𝐻1 whereas the wage rate increase the quantity of 𝐻1 In the multiple period model, both of demand functions also are added the age because the depreciation rate 𝛿 is positively correlated with age (Zweifel

et al 2009)

In addition to the Grossman Model, some researches brought out some model to explain the individual’s health seeking behavior Rosenstock (1974) introduced health belief model to explain the health-related behavior The model suggested that individual’s belief about health problem, perception regarding benefit and barrier and cues to action combine together to explain the health-promoting behavior The modifying variables such as demographic and psychosocial characteristics made the indirect effect on the perceptions In the model, the perception regarding serious consequences and risks from developing a health problem most highly increase the likelihood of engaging in health-promoting behaviors The next step was the perception of benefits received from taking action and barriers to taking action such as inconvenience, side effect If the perceived benefits outweigh the barrier, it leads to health-promoting behavior or otherwise The following is cues to action including internal and external cues Internal cues are pain, symptoms while external cues are information obtained from friends, mass media The model was applied to develop intervention to change health-related behavior by targeting the parts in the model

Thaddeus and Maine (1994) developed the three delay theory to identify the barriers to the timely and adequately utilization of maternal health care They viewed that the first phase was delay in seeking care The factors effecting the phase included decisions of individual and family, the status of women, previous experience regarding health care system, financial and opportunity cost The next phase was delay in getting access to the health facility due to the availability of facilities, distance to facilities, transportation cost and transportation infrastructure The last phase was delay in taking adequate care The relevant causes were lack

of equipment and skilled health staffs, qualification of health staffs

The behavioral Model of Health Services Utilization by Andersen (1995) has been used

in the studies on utilization of health services in both developed and developing countries

(Thind et al 2008) In the Andersen’s Model, one individual getting access to and using health

care services was a function of three factors, namely predisposing, enabling and need factors The predisposing factors were classified into three groups such as demographic characteristics,

social structure and health beliefs Demographic characteristics represent the tendency of individuals to use services, including age, gender, marital status while social structure reflect the ability of individuals to seek health care services, including education, occupation and ethnicity Health beliefs were perception and attitudes regarding the health care systems that influence the utilization of the services The next was enabling factors, both personal and organizational, which measure the actual ability to obtain health services Personal enabling factors included income, health insurance, travel and waiting times while organizational enabling factors were the availability of health facilities The last one is need factors, the direct causes to utilization of health services The need was the self-assessment of health status and evaluation from health staffs

2.3.2 Empirical Literature Review

Theoretically, prenatal care is considered a vital factor in maternal and infant mortality Measuring and counseling during prenatal care help women understanding health status of herself and her baby, detecting risks in order to ensure safe pregnancy, childbirth as well as decrease post-partum problems While some researches were conducted to estimate the association between maternal health care and health outcome, some studies focus on determinants effecting utilization of maternal health care The below is to present the factors influencing the demand of prenatal care visits and the choice of delivery location from the previous studies

The determinants are categorized into three characteristics including individual level, household level and community level The individual variables include education attainment, maternal age, marital status, religion, ethnicity while household characteristics are household size and household wealth Further, community level consists of place of residence, region dummies, poverty rate and illiteracy rate

Individual level characteristics

Mother’s education

Most studies highlighted the importance of mother education on the use of maternal health care The higher level of education women has, the more likely to use adequate ANC (Arthur 2012, Bbaale 2011, Navaneetham & Dharmalingam 2002) It was explained that educated women had more decision-making power on health-related matters and seek the better health care outside their home (Navaneetham & Dharmalingam 2002) However, there was no

significant difference in the utilization of maternal health care among educated women by level

of education, especially between primary level and second level (Navaneetham & Dharmalingam 2002) In addition, when measuring the effect of the introduction of National

Health Insurance (NHI) in Taiwan, Chen et al (2003) found that education attainment did not

have insignificant impact on antenatal care before NHI but was significantly positive after NHI The reason for this difference was ambiguous In general, educated women are well informed

of the importance of maternal care and will increase more frequencies of antenatal care visits than less-educated mothers

Marital status

Marital status also is a key determinant of the use of maternal health care Sepehri et al

(2008) suggested that the marital status impacted the use of any prenatal care greater than the number of visits but has no significant influence on decision of delivery place It could be explained that single mothers unwillingly get access to the health care due to face of stigmatization because the childbirth is widely considered the responsibility of mother as well

as her husband in Vietnam (Sepehri et al 2008) Researches in other countries were consistent with the study of Sepehri et al (2008) In the study in Taiwan, Chen et al (2003) found that

thanks to the support from their husband, married women more likely to get access to maternal care visits than unmarried women

Mother’s age

The age of expectant mother is other factor in the utilization of maternal health services The higher age the mothers get, the lower the probability they will seek health care (Arthur 2012) It could be explained that experience and knowledge related maternal health may

influence on the maternal health seeking behavior (Chen et al 2003) Tsawe & Susuman (2014)

showed that the women in the age of 15-39 more likely to take health check-ups frequently than those in the age of above 40 However, there were some researches showing that women’s age was insignificantly associated with the attendance of ANC for example, in Turkey (Celik

& Hotchkiss 2000)

Birth order

As the same as the age of expectant mother, most studies found that the increasing number of children ever born had the same negative effect on the use of maternal health care services The higher order births will decrease the likelihood of using the services for some reasons It may be due to time and resource constraints for women having larger families (Navaneetham & Dharmalingam 2002) In contract, those who have the first pregnancy will more likely to use the antennal care due to lack of experience Arthur (2012) argued that those who underwent the child birth may undertake ANC visits less because she had bad experience with the service Similarly, Tsawe & Susuman (2014) agreed with Arthur (2012) that the use

of maternal health care was influenced by the experience over the provided service If a woman

is provided the better service, she willingly utilizes the service more frequently On the other hand, due to the two-child policy and fear of penalties, those with more two children may

undertake less maternal health care (Sepehri et al 2008)

Unintended pregnancies

Studying in the maternal health service in Southwestern Ethiopia, Wado et al (2013)

found that pregnancy intention was significantly related to the utilization of antenatal care but insignificantly to the delivery care The reason why the unwanted pregnant women less likely

to receive antenatal care adequately is ambiguous Wado et al (2013) hypothesized that women

with unwanted pregnancies did not prepare well in the respect of emotion and finance for childbirth and childbearing so they less take care of their health and their unborn One of the argument is the women with unintended pregnancies recognized their pregnancy late so they

miss the first timing antenatal care visits Wado et al (2013) observed that women with

unwanted pregnancy detect the pregnancy around one month later compared with ones with pregnancy intention In general, the study implied that the pregnancy intention was highly associated with the utilization of maternal care; however, its association with the delivery care was still unclear

Media exposure

Some studies resulted that frequencies of using mass media positively related to the utilization of ANC The higher is the level of exposure to mass media like radio and television the higher is the likelihood of using ANC Navaneetham & Dharmalingam (2002) found that women watching television and listening radio regularly increased the opportunities of her seeking ANC It can be explained that the mass media will provide the information regarding

to the maternal health and enhance the awareness of the available maternal health services Tsawe & Susuman (2014) reported that those who had more knowledge of the maternal health service more likely to receive the service timely and adequately In fact, lack of information poses the challenges to the utilization of maternal health

Working status

The importance of working status in maternal health care utilization also is well documented in some previous studies Navaneetham & Dharmalingam (2002) found that non-working women more likely seek maternal health care services than earning women It could

be explained that non-working women were relatively richer in comparison with those working In addition, women’s work in the developing countries does not come with well-paid salary, which leads lower likelihood of using maternal health care services Another study by Bbaale (2011) pointed out that the type of job which pregnancy women related to impacts the utilization of antennal care in different way It could be attributed to conditions of jobs and income they earn, which has great effect on the affordability of health care service Women who working in farm or factory have more tendency to use antenatal care than housewives as well as women who working in government more likely seek antenatal care after being covered

by national health insurance (Chen et al 2003)

Household-level characteristics

Ethnicity

The significant impact of ethnicity in the utilization of maternal health services is found

in some studies Sepehri et al (2008) showed that women belonging to ethnic majority more

likely give birth at health facilities than those in ethnic minority in Vietnam It could be explained that ethnic group find difficulties to communicate with health provider due to language barrier In addition, lower attainment of maternal health care among ethnic group

women is attributed to lower level of education and poor socio-economic condition (Singh et

al 2014) In other countries, the results are consistent with this study Navaneetham &

Dharmalingam (2002) pointed out that women in “Schedules” castes and tribes had lower ANC check-up in India Similarly, Kudish women were less likely to take ANC service in Turkey (Celik & Hotchkiss 2000)

Religion

Similar to Ethnicity, religious difference is one of the most important determinants of health service utilization because religion affects the way people live and behave as well as their belief (Bbaale 2011) If women believing in the traditional birth attendants more likely give birth at home rather than going to hospital or clinic (Tsawe & Susuman 2014) Studying

in India, Singh et al (2013) found that practices in Muslim restrict women among this religion

to enjoy health care benefits Similarly, Muslim women in Kerala tend to give birth at home and refuse to be assisted by skilled health providers; however, the probability of antennal care visits in the first trimester is higher among them (Navaneetham & Dharmalingam 2002)

Household wealth status

Previous researches show that household wealth status has positive impact on the use

of ANC (Bbaale 2011; Tsawe & Susuman 2014) There are many indicators regarding household income index showed in the previous studies For example, Navaneetham & Dharmalingam (2002) showed that the women enjoy higher living standard three times more likely to get adequate prenatal care visits rather than those with lower living standard Owing

a car found to be positive associated with utilization of maternal health care (Celik & Hotchkiss 2000) Even though some countries provided the service freely, the matter remained unchanged It is implied that the health care services may come with another costs directly and indirectly such as transportation cost (Arthur 2012) Women who find difficult to afford traveling fee to the hospital is less likely to undertake checkup of ANC than those who do not find difficult to get money to go to hospital (Tsawe & Susuman 2014) Generally, financial problems were also major hindrance on the utilization of maternal health care

Community-level characteristics

Place of residence

Place of residence is key concern in ANC utilization The previous studies resulted that women in rural areas less likely use maternal health services than urban women due to long distance to the health facilities In addition, financial problems also were a barrier for them to complete use of maternal health care because they could not afford the transportation cost as well as ANC cost (Tsawe & Susuman 2014) The shortage of skilled attendant also is the reason why rural women less frequently undertake antenatal care than urban area In fact, there is lack

of qualified health provider in rural area in comparison with the cities (WHO, 2006)

Understanding these matters, some countries applied the program to minimize the problems For example, when studying four states in the south Indian in 2002, Navaneetham & Dharmalingam (2002) found that there is no significant difference between rural and urban area in terms of the demand for antenatal care The underlying reason was that multipurpose health workers provide antenatal care to pregnant women by visiting their home In contrast, Celik & Hotchkiss (2000) found that urban and rural difference did not significantly impact the prenatal care after holding the constant regional status and other variables in the study in Turkey

Regional differences

The difference in utilization of maternal health care is due to the implementation of health care program, the availability and accessibility of the health care service among the

regions Studying in Vietnam, Sepehri et al (2008) pointed out that women living in the

disadvantaged regions like Central Highlands less likely undertake any maternal health care than those living in Red River Delta However, there was no significant regional difference in

the frequency of prenatal visits (Sepehri et al., 2008) Similarly, in Turkey, Celik & Hotchkiss

(2000) also found that living in developed regions of the country had significantly positive

association with attendance of prenatal care service

Poverty rate, illiteracy rate and ratio of facility delivery

Beside individual level and household level characteristics, community level characteristics have been paid greater attention in the previous studies in recent years (Gage

2007, Sepheri et al 2008, Ononokpono et al 2013 and Singh et al 2014) Stephenson et al

(2006) argued that community could influence health outcomes through several ways For example, community beliefs and norms may have great effect on positive or negative attitudes

on health care services, which in turn lead to health care seeking behavior In addition, economic development of a community will attract the investment of health care infrastructure, social support as well as improvement of knowledge in benefits of health care services Studying in rural Mali, Gage (2007) suggested that community with higher portion of well-educated women was highly associated to likelihood of getting access to maternal health care

In Vietnam, Sepehri et al (2008) pointed out that women living in community with higher

poverty rate have fewer tendency to deliver their baby in health facilities in comparison with those living in community with lower poverty rate One of the community level characteristics,

which were also found to be positively associated with the demand of prenatal care visits is the

proportion of women in same community giving birth at health facility (Ononokpono et al

2013) The research showed that the women living in the community with the high proportion more likely 5.8 times to take adequate prenatal check-up compared to those in the lower community level proportion

2.4 The choice of health care provider

2.4.1 Theoretical background:

In the late 1980s, Gertler, Locay and Sanderson (1987) firstly developed the health seeking behavior in the choice of health care providers They argued that at the first stage, the individual decides whether to seek health care and at the second stage, he or she selects the health care providers with the maximum utility The utility function of individual i receiving health care from provider j is presented as follows:

In which, 𝑈𝑖𝑗 is the utility of the individual i after receiving health care from provider

j, ℎ𝑖𝑗 is expected health status of the individual after receiving health care from provider j and

𝐶𝑗 is other consumption expenditure after paying provider j

The health status after receiving health care from provider j for an individual i depends

on the quality of provider j’s health care

Where ℎ0 is the health status before receiving health care from provider j

The quality of health care differs among health care providers and individuals Therefore, the quality of health care depends on the characteristics of health care providers 𝑍𝑖and characteristics of individual 𝑋𝑖 as follows:

The other consumption expenditure 𝐶𝑗 after receiving the health care is the remaining

of the income 𝑌𝑖 after paying health care provider j The price 𝑃𝑖𝑗 of the alternative j is the payment for the health care including direct cost such as consultation cost, medicine cost and indirect cost such as transportation cost and waiting time

𝐶𝑗 = 𝑌𝑖 − 𝑃𝑖𝑗 (2.4) Assuming that the individual has j alternatives and would like to maximize the utility

so the utility maximization is expressed as:

In which, 𝑉𝑖𝑗 is observed characteristics and 𝜀𝑖𝑗 is unobserved characteristics

Combining (2.5) and (2.6) we get

The observed characteristics of the individual are gender, age, education, income, insurance whereas the unobserved characteristics are perception of quality of the health care provider, the preferential medical administration The observed characteristics of the health care provider are the price, the distance from house of the patient to the health care provider whereas the unobserved characteristics are the fame, prestige of the health care provider

2.4.2 Empirical literature review

It is widely acknowledged that childbirth at health facilities is associated with lower rate of maternal mortality and morbidity However, there are still women choosing the child birth at home, especially in developing countries Therefore, some previous studies have further conducted to investigate the determinants of the choice of delivery care provider They have focused the characteristics of individual of the women, household and community where they are living

Individual level characteristics

Most studies highlighted the important role of prenatal care visits in the decision to seek

health care Stephenson et al (2006) argued that the prenatal care during pregnancy was highly

associated with the choice of delivery at health facilities by informing the benefit of

institutional delivery and appropriate services Sepehri et al (2008) agreed this point that

timing and adequate prenatal care visits could raise the awareness of the need of care for delivery

Education attainment also is found to play the important role in the choice of child birth

place (Sepehri et al 2008) For example, Celik & Hotchkiss (2000) pointed out that the women

with higher education more likely to choose facility delivery rather than traditional home delivery It is likely that the highly educated women have the stronger and independent decisions to the best health care services Furthermore, the educated women have much more chance to be well informed of the benefit of facility-based delivery

With respect to the birth order, the previous studies reported that it has strong association with the choice of delivery at health facilities Navaneetham & Dharmalingam (2002) found that the women with first birth order more likely to give birth at health care institutions rather than those with second birth order There are some arguments regarding the matter Some researchers pointed out that the women with higher parity faced time and resource constraints to utilize the facilities services due to large families On the other hand, it could be the bad experiences from previous childbirth, which make them to underestimate the demand for facility-based delivery

Concerning the age of the pregnancy women, there are mixed result regarding its effect

in the choice of birth delivery alternatives Celik & Hotchkiss (2000) reported that the age of women at last child birth was not significantly associated with the choice of delivery location

On contrary, Stephenson et al (2006) argued that the age of the interviewed women had

significant association with the choice of facility delivery in the study of six Africa countries

He found that the women with age of 40-49 and 30-39 more likely to deliver their baby at health facilities than women aged 20-29 However, the effect was not significant in some countries such as Burkina Faso, Ivory Coast or Ghana

In addition, marital status showed mixed result in the decision in the birth delivery

location Stephenson et al (2006) pointed that the women in polygamous marriage or those

who was separated less likely to have their last childbirth at health institutions in all of six

Africa countries However, Sepehri et al (2008) found that marital status had no significant

impact in the choice of delivery location

Household level characteristics

The choice of the place of delivery is also highly influenced by the household level characteristics, namely household wealth index, ethnicity and religion With respect to the

household wealth index, Stephenson et al (2006) found that women with higher household

income index more likely to choose the delivery at health institution rather than those with lower income index The reason is that the cost related birth delivery care as well as transportation restricts the poor women from utilizing the health services Some previous studies identified that ethnicity has great influence in the choice of the facility delivery Celik

& Hotchkiss (2000) agreed that the effect is true in urban women and rural women He said that it could be due to the cultural and economic barriers or the poor of health care service Similar to the ethnicity, religion also is considered as the key determinant in the women’s

decision in the place of childbirth When studying six Africa countries, Stephenson et al (2006)

found that Muslim women in Ghana are less likely to deliver their baby in health facility compared to Catholic women, while Protestant women more likely to utilize the service than Catholic women

Community level characteristics

There are also remarkable regional differences in the choice of delivery location Celik

& Hotchkiss (2000) found that urban women were more likely to seek the facility delivery rather than rural women Furthermore, Celik & Hotchkiss (2000) said that the Eastern regions

in Turkey is less advantaged than Western and Northern regions, which caused more difficulties for the women in Eastern regions to choose facility delivery compared to those living in other regions Gage (2007) argued that the regional differences partly indicate the unequal accessibility and availability of health care facilities

In addition, some previous studies investigated the impact of community level characteristics in the decision to delivery at health facilities There are several ways though

which the community affect the delivery facility choice of women Stephenson et al (2006)

reported that the community of higher concentration ratio of women with higher education improve the greater autonomy level and provide more opportunity for the women to seek health

care during pregnancy and delivery Furthermore, he also found that higher concentration ratio

of women in the community delivering their baby at health facilities had a significantly positive impact on the decision in the delivery location It suggests that the practices of others in the community could affect the health seeking behavior of the individual in the same community

One of the important community level factors as reported in the study of Sepehri et al (2008)

in Vietnam is community level poverty rate He pointed that the community poverty rate affected negatively the probability of giving birth at health facilities Women in the community with higher poverty rate above 50% less likely to deliver their baby at health institutions in comparison to those in the community with poverty rate of equal and less than 10%

CHAPTER III METHODOLOGY AND DATA DESCRIPTION

This chapter presents the methodology and data description used to investigate the determinants of the utilization of maternal health care The first part is to introduce conceptual

24

framework describing the associations between the determinants and the demand of prenatal care visits and the choice of delivery care providers respectively Next, two methods related to the objective of the study, namely the Negative Binomial model and the Multinomial Logistics model will be discussed to analyze the associations as mentioned in the previous part The final section presents the data used in the study

3.1 Conceptual framework

Based on the theoretical reviews and empirical reviews, the demand of prenatal care visits and the choice of delivery care provider are influenced by the individual, the household level and the community level characteristics Individual characteristics include the education level, the frequency of getting access to mass media, working status, marital status, pregnancy intention and the birth order while the household characteristics include household wealth index, household size, ethnicity and the religion of the household head Besides, the study supplements the community level characteristics, which also affects the individual maternal health seeking behavior through the residence place, poverty rate, illiteracy rate and the ratio

of women giving birth in health facilities The conceptual framework is expressed in Figure 6

Figure 6 The association between individual level, household level and community level characteristics with the utilization of maternal health care services

3.2 Empirical framework

The objective of the study is to measure the social determinants on two types of maternal health care services such as number of prenatal care visits and the choice of delivery facility The Negative Binomial Model is applied to analyzed the number of prenatal check-ups and the Multinomial Logit Model for the decision on the delivery facility

3.2.1 Demand for Prenatal care

Poisson Model is applied for dependent variables which are non-negative integer In the model, Poisson distribution is used to describe the likelihood of events occurring k times

in a given period of time The distribution function is

model the over-dispersion The command nbreg in Stata is implemented to estimate the

Negative Binomial regression and provide the test of over-dispersion

In the study, the dependent variable is number of maternal check-ups that a pregnant woman takes The independent variables are divided into three groups: individual level, household level and community level characteristics Different from the theory background, the prices of health care and the income are not included due to limitation of the data The detail will be introduced in next section

3.2.2 Choice of birth delivery facility

Multinomial Logit Model is applied to measure the relationship between categorical variables and other explanatory variables Mothers are assumed to choose the facility that maxime utility And the choice of facility j of individual i coded as 𝑌𝑖 is 1, 2 ,3 … J The probability for choosing each alternative is 𝑝𝑖1, 𝑝𝑖2, … , 𝑝𝑖𝐽 The logit function is described as follows:

𝑝𝑖1 = 𝑒𝑋𝑖𝛽1

∑ 𝐽 𝑒 𝑋𝑖𝛽𝑘 𝑘=1

𝑝𝑖2= 𝑒

𝑋𝑖𝛽2

∑𝐽 𝑒𝑋𝑖𝛽𝑘𝑘=1

𝑝𝑖𝐽= 𝑒

𝑋 𝑖 𝛽 𝐽

∑𝐽 𝑒𝑋𝑖 𝛽 𝐽 𝑘=1

The command mlogit in Stata is employed to estimate the multinomial logistic model

The independent variables are divided into three groups: individual level, household level and community level characteristics The details will be introduced in next section

3.3 Data

MICS was carried out in Viet Nam by Viet Nam General Statistics Office in collaboration with UNICEF The sample for Viet Nam MICS was designed to provide estimates for a large number of indicators on the national level situation of children and women

in urban and rural areas as well as six geographic regions Red River Delta, Northern Midlands and Mountain areas, North Central area and Central Coastal area, Central Highlands, South East and Mekong River Delta The survey conducts three set of questionnaires: first, the questionnaire on households to collect information regarding household members and economic status; second, the questionnaires on female household members in the productive age (15-49 ages) and the last one is administered to children under 5 years and their caretakers MICS5 is based on a sample of 10,018 interviewed households, with 9,827 women and 3,316 children For the purpose of the study, the sample is narrowed down into the women who gave birth to a live child within two years before the survey in order to reduce recall errors Therefore, there are 1,479 women in the productive age applied in the study

3.4 Variables definition

3.4.1 Dependent variables

The study measures two dependent variables such as amount of antenatal care coverage and place of delivery as indicators of maternal health care service utilization In MICS, antenatal care coverage is defined as the percentage of women aged 15–49 years – among those who had a live born during the two years before the survey – who received antenatal care from skilled health personnel at least once However, the World Health Organization recommends pregnant women should have at least four ANC visits during the pregnancy It is crucial for pregnant women to attend antenatal care visits early to prevent and detect conditions that could affect their health and baby Antenatal care should continue throughout the entire pregnancy

“Skilled personnel” includes accredited health professionals such as midwives, physicians and nurses, but not traditional birth attendants Therefore, the antenatal care visits are integer variables

Place of delivery is defined as delivery in a health facility (private or public) or outside the health system (home delivery) The increased proportion of births delivered at health facilities is an important factor to reduce health risks to mother and baby Proper medical attention and hygienic conditions during delivery can reduce risks of complications and

infections that can cause morbidity and mortality to mother or baby Therefore, the delivery place is categorized into three groups: 1 if women give birth at home, 2 if women give birth in public hospitals and 3 for those who deliver at private hospitals or clinics

3.4.2 Independent variables

The individual-household level, community-level explanatory variables were used based on the theoretical and empirical literature, considering the use of maternal healthcare services and their availability in the dataset Because two regressions for the two research objectives will be analyzed, they will share the same set of independent variables The detailed description of selected variables is given below

Regarding individual level factors, the study considers a number of potential individual factors such as the mother’s age at childbirth, birth order, education, material status, exposure

to mass media and pregnancy intention The mother age at childbirth is continuous variables from age of 15 to 49 Maternal education is divided into five dummy variables: no education, primary, lower second, upper second and tertiary The birth order, which means the number of children a woman ever giving birth is used as a continuous variable Material status is used as dummy variables: 1 if the women used to be married but no longer in union and the women are never married and 0 otherwise The exposure to mass media is divided into four dummy variables to measure frequency of getting access to mass media: mobile phone, newspaper, radio and television The last one is pregnancy intention measured as 1 if the women did not want to be pregnancy and 0 otherwise

Household level factors include ethnicity and religion of household head, household wealth status Ethnicity was defined on either Kinh or Hoa (= 1) or non-Kinh/Hoa (= 0) Although they are the sixth largest minority group in Viet Nam, their living standards are on a par with those of the Kinh majority The religion variable is used as a dummy variable: 1 if the household has no religion and 0 otherwise The household wealth status also is dummy variable: 1 if the household is in the poorest or poor quintiles and 0 otherwise

In addition to individual and household level factors, community level factors are added

to measure any effect in community level First, place of residence is divided into two categories: urban coded as 1 and rural as 0 Moreover, six regional dummies represent six socioeconomic regions such as Red River Delta, Northern Midlands and Mountain areas, North

Central area and Central Coastal area, Central Highlands, South East and Mekong River Delta Second, illiteracy rate is based on percentage of illiterate women in the community Third, poverty rate is calculated as percentage of women in the commune n the lowest wealth quintile Finally, the percentage of the women in the community delivering their baby at hospitals is applied

Table 1: Description of Variables

Individual-level Characteristics

NOEDU Dummy variable indicating the individual has not finished primary school Yes = 1, No = 0

LOWSECOND Dummy variable indicating the individual finished lower secondary school Yes = 1, No = 0 UPSECOND Dummy variable indicating the individual finished upper secondary school Yes = 1, No = 0 TERTIARY Dummy variable indicating the individual finished the college above Yes = 1, No = 0

Household-level Characteristics

Community-level Characteristics

POVERTY Percentage of women with poorest and the 2nd quintile in the commune percentage (%) ILLITERACY Percentage of women in the commune with no education certificates percentage (%) HOSPDELIRATIO Percentage of women in the commune giving birth the last child at hospitals percentage (%)

CHAPTER IV RESULTS AND DISCUSSIONS

The chapter presents the results of study to report the relationships between determinants and the demand of prenatal care visits and the choice of delivery care providers respectively The first part is to present descriptive statistics of dependent variables and independent variables The following part is to analyze the bivariate associations between each dependent variable and independent variables The final is to demonstrate the regression