There are a limited number of studies on ANC content (refer to Table 3.4). The identified studies are grouped according to methods of reporting information: proportions of each procedure or advice delivered, proportions of women receiving all or part of ANC services and mean number of procedures andor advice delivered. In each group, studies on women are presented before studies on health care providers. The common conclusion is that biomedical assessment receives the most attention, and that other components of health promotioneducation and care provision do not. This might be due to the fact that health assessment was the only component of ANC when first introduced and even now, assessment is still perceived as the most important part. Counselling on health behaviour was added later and is often perceived by both health care workers and women as less important
CHAPTER ONE: INTRODUCTION
Pregnancy and childbirth related mortality and morbidity
Pregnancy and childbirth are important events for societies and families They should be happy events, but in developing countries, these processes can result in suffering and death for many women In Vietnam, risk of pregnancy-related death used to be so high that there was a proverb saying: “Gai chua cua ma”, that meant being pregnant was equivalent to entering a tomb The situation is much better nowadays but still many women are suffering
1.1.1 Effects and consequences of pregnancy
Pregnant women are at risk of many unpredictable life threatening pregnancy complications such as miscarriage, ectopic pregnancy, eclampsia, placenta abruption, obstructed labour, haemorrhage, infection, high blood pressure and unsafe abortion The female body has to adapt to bear the fetus, it becomes weaker and more vulnerable to many health problems which may appear or worsen during pregnancy, such as anaemia, malaria, hepatitis, tuberculosis, heart disease and HIV/AIDS (WHO 2000)
Complications not only adversely affect mothers but also the fetus Poor maternal health can result in miscarriage, still births, neonatal death, low birth weight (LBW), congenital infection with syphilis, HIV, tetanus, malaria, conjunctivitis and neonatal retardation (WHO 2000)
The death of the mother has seriously consequences for the life of the baby and other children in the family Without the care of its mother, a child is at higher risk of dying than other children The risk of death in children without a mother is ten times higher in the neonatal period, seven times during the post neonatal period and three times between the age of one to five years compared to children with mothers (Kolinsky, Conroy et al 2000)
Health during the foetal period has an important effect in adulthood Nutrition received from the mother during a pregnancy and any exposure to infection can influence susceptibility to diseases in later life (Barker 1998) Studies show that many adult diseases such as coronary heart disease, stroke, hypertension, diabetes and cancer have strong correlations to the health status of newborns and mother’s body composition and diet (Vietnam DHS 1999)
Women care for families and contribute to household income Their death can result in a decline in the general nutritional status of family members and this in turn results in poorer health, less care for older people, reduction in education for children and less household income (Kolinsky, Conroy et al 2000)
All of these effects can cause great distress and loss for families and communities, besides the immeasurable mental and emotional suffering of families, relatives and friends
1.1.2 Magnitude of maternal and infant mortality and morbidity
Worldwide, 600,000 women die every year from pregnancy and childbirth complications, that is 1600 deaths every day, or one death every minute (WHO 1998; WHO 1999) A further 50 million women may suffer many long-term health problems, resulting from pregnancy and birth related complications (WHO 1998) It is estimated that around USD 7.5 billion of potential productivity may be lost due to pregnancy complications and maternal deaths (Kolinsky, Conroy et al 2000)
Source: WHO, UNICEF, UNFPA, Maternal Mortality in 1995: Estimates developed by WHO, UNICEF,
Figure 1.1: Maternal mortality in the world
Fetus and newborn mortality and morbidity
Every year, six million newborns die Among them, three million are stillbirths and another three million die within the first week postnatal Poor maternal health contributes to half of these deaths (WHO 1999) In addition, an estimated 16 million babies per year have low birth weight as a result of poor maternal health The economic value of these effects on newborn morbidity and mortality is around eight billion United States dollars per annum (Kolinsky, Conroy et al 2000)
Maternal mortality is especially of concern in developing countries with 98% of maternal deaths occurring in the developing world The ratio of maternal deaths per live births in developing countries is 18 times higher than in developed countries with 480 deaths per 100,000 live births compared to 27 deaths per 100,000 live births In some parts of the world, one in every 10 women dies because of pregnancy and childbirth complications (WHO 1999)
In 1987 in recognition of the problem of maternal and infant health, WHO joined with United Nations Children’s Fund (UNICEF), United Nations Population Fund (UNFPA), the World Bank and other organisations to launch the Safe Motherhood Initiative which aimed to improve maternal and infant health Many efforts have been put in place but after 12 years of implementation, WHO admits: “decreases in maternal mortality globally have been limited” Given the current rate of progress the goal of reducing infant mortality to less than 35 deaths per 1000 live births will not be achieved by the year 2015 as initially planned (WHO 2000)
East Asia and Pacific region
Compared to the rest of the developing world, the East Asia/Pacific region has a low maternal mortality rate: 140 per 100,000 live births compared to the average of 440 per 100,000 live births for the developing world
However, due to the large population and high birth rate in the East Asia/Pacific region, the total number of maternal deaths is significant In 1995, 49,000 women in the East Asia/Pacific region died from pregnancy complications which accounted for about 10% of the world total number of deaths from these causes This proportion made the East Asia/Pacific the region with the third highest number of maternal deaths, after Sub- Sahara Africa (50%) and South Asia (30%) The remaining 10% of deaths were distributed across the rest of the world with less than 1% occurring in industrialised countries (WHO/UNICEF/UNFPA 2001)
Similar to the maternal mortality rate, the infant mortality rate in the East Asia/Pacific region is lower than most parts of developing world The rate was 35 deaths per 1000 live births compared to the average of 59 per 1000 live births for developing countries in 1998 (The World Bank Group 2001)
Compared to other parts of the world, Vietnam has a high fertility rate The birth rate estimated by the Ministry of Health and World Bank is around 2% (MOH 1999) (The World Bank Group 2003) There are 39 million females, of which 21 million are of child bearing age (15-49 years of age) It is estimated that there are more than one million babies born every year in Vietnam
Vietnam has high maternal mortality and morbidity Although there is no accurate figure for the maternal mortality rate in Vietnam and estimations can vary between different sources (MOH 2003) The maternal mortality rates in Vietnam were estimated to be 130 to 160 per 100,000 live births in 1998 (UNICEF 2000), or 165 per 100,000 live births in 2002 (MOH 2002) It is estimated that there are 1500 maternal deaths and up to 45,000 women become disabled as a result of pregnancy and childbirth complications annually (MNPI 2003) Maternal deaths occur more often in mountainous provinces than in non-mountainous areas (MOH 2002) The objective of the Vietnamese government is to reduce the maternal mortality rate to 70 per 100,000 live births by the year 2010 (UNICEF 2000) The maternal mortality rate for Vietnam is higher than in Thailand (44 per 100,000 live births) and Malaysia (39 per 100,000 live births), but lower than countries such as Laos (650 per 100,000 live births) and Indonesia (470 per 100,000 live births) (WHO/UNICEF/UNFPA 2001)
Role of ANC in preventing maternal and infant mortality and morbidity
1.2.1 Effect of ANC in preventing maternal and infant mortality and morbidity
The World Health Organisation has conducted extensive literature reviews of the effectiveness of ANC in preventing maternal mortality and morbidity The large review published in 2001 (Carroli, Rooney et al 2001) focused mainly on developing countries and included nearly 300 randomised controlled trials, intervention and observation studies ANC was dichotomised as either having any ANC or not The conclusion was that ANC was effective in preventing maternal mortality and morbidity
For infants, there have been studies showing that ANC is effective in preventing LBW and preterm delivery (Weiner and Milton 1970; Kessner, Singer et al 1973; Schrarmm 1980; Showstack, Budetti et al 1984; Covington 1988; Koroukian and Rimm 2002)
However, only the effect of ANC utilisation (having any ANC, number of visits and duration of pregnancy at entry to ANC) was investigated in these studies The effects of ANC content and overall ANC adequacy have been rarely investigated, not only in the developing countries but also in developed countries
In the review by Carroli et al, the authors have outlined how ANC could be effective in prevention of maternal mortality and morbidity (Carroli, Rooney et al 2001) (presented in Table 2.1)
Table 1.2: Antenatal care interventions known to be effective
Condition, stage Test, treatment Proven effect
Prevention of anaemia Routine supplementation with iron and folate Malaria chemoprophylaxis
Reduces or prevents haemoglobin fall Reduces percentage of anaemic women
Reduces percentage of anaemic women
Detection and investigation of anaemia
Detect haemoglobin level below chosen cut off points Estimates haemoglobin concentration
Diagnoses of type of anaemia Diagnoses of type of anaemia and malaria
Treatment of iron-deficiency anaemia
Can raise Hb by 0.4-0.7g/dl per week
Can raise Hb at same rate as oral treatment Avoids problem with compliance but need IM or IV equipment and trained staff Danger of anaphylaxis
Packed cell transfusion Raises Hb immediately Hazards of blood transfusion infection such as HIV and hepatitis, and fluid overload Equipment and trained staff needed
Measurement of blood pressure with sphygmomanometer using fifth Korotkoff sound
Detection and investigation of hypertensive disorder of pregnancy Urinalysis of clean catch urine Detect proteinuria Indicative of pre-eclampsia or presence of hypertension
Transfer to first referral level for expert care
Control of disease Reduces case fatality
Treatment of eclampsia Supportive first aid maintaining airway and preventing injury during fit
Magnesium sulphate IM or IV Reduces recurrent convulsions and mortality
Recognition and quickly transfer to equipped facility Reduce fatality Expedited delivery Only definite treatment
Prevention of obstructed labour External cephalic version at term Reduces Caesarean delivery
Serological screening and treatment for syphilis
Detect asymptomatic disease Together with effective treatment, contact tracing and follow up, reduce foetal loss, maternal and infant morbidity
Microbiological screening for gonorrhoea Detect asymptomatic disease Together with effective treatment, contact tracing and follow up, reduce foetal loss, maternal and infant morbidity
Screening for bacteriuria with quantitative culture of urine Detect asymptomatic disease Together with effective treatment prevents pyelonephritis and pre term delivery/LBW
Tetanus immunisation in pregnancy and/or women of child bearing age
Prevents maternal and neonatal tetanus
Induction of labour in uncomplicated pre labour rupture or membranes
Source: How effective is ANC in preventing maternal mortality and serious morbidity? An overview of evidence (Carroli, Rooney et al 2001)
First introduced in Europe and North America in the early 1900s, ANC is the routine care delivered to women throughout pregnancy to prevent, detect and treat conditions which could lead to mortality and morbidity in pregnant women and/or the fetus
Recently, care provision and health promotion have been added to ANC content All pregnant women are advised to visit health care workers for regularly check ups throughout their pregnancy, with more frequent visits during the late phases of pregnancy ANC is considered adequate when the number, timing of visits and the content of care provided meet recommended levels
Number and timing of visits
The number of recommended ANC visits varies between countries Developed countries recommend up to 13 visits for full term pregnancy, once a month until the 6 th month, once each two weeks for the 7 th and 8 th month and once a week until delivery (American Academy of Pediatrics 2002)
Recently, there have been a number of trials examing the impact of a reduced number of ANC visits in both developed and developing countries A review conducted by WHO of ten randomised control trials involving over 60,000 women suggested that a moderate reduction of the number of visits does not affect pregnancy outcomes but may reduce women’s satisfaction (Villar, Carroli et al 2001)
As the consequence of this finding, some developed countries like England has started to recommend a reduced number of visits than traditionally recommended (National collaborating for women's and children's health 2003) The recommended visits were reduced to ten visits for primiparous women and seven visits for multiparous women
WHO also conducted a multi country randomised controlled trial in selected developing countries, Argentina, Cuba, Saudi Arabia, and Thailand, involving 25,000 women in more than 50 clinics (Villar, Ba'aqeel et al 2001) The results showed that a reduction of visits from eight to five did not affect pregnancy outcomes and may reduce costs although for some women satisfaction was reduced
For developing countries, WHO recommends that women experiencing a normal pregnancy (i.e without complications) should have at least four ANC visits and these visits should be at the 4 th , 6 th or 7 th , 8 th and 9 th month to allow follow up and detection of conditions developed throughout out pregnancy (WHO 1994)
For many years, the Vietnamese government has recommended three visits, one visit during each trimester (MOH 2002)
The effects of components of ANC content on pregnancy outcomes have been thoroughly evaluated in a literature review involving 130 studies (Villar and Bergsjo
1997) Villar et al had summarised the proven effective components and these are presented in Table 1.2
Components of ANC contents may vary between countries but there are certain basic components should be provided everywhere WHO recommends that ANC should consist of three basic components: (i) assessment based on medical history, physical examination and laboratory tests; (ii) health promotion and (iii) care provision (WHO
1994) Assessment is evaluation of health and potential risks for the woman and fetus This includes taking a medical history, conducting a physical examination and laboratory tests Women at risk of pregnancy complication are detected and treated
Health promotion provides counselling to women on how to take care of themselves and their baby during pregnancy and after birth Examples are issues on resting, nutrition, hygiene, breastfeeding and safe sex Care provision includes discussion with the women about the timing of follow up visits, delivery plan and giving preventative treatment such as tetanus toxoid immunisation and iron/folic acid supplementation as required Tetanus vaccination should be given in two doses to be fully effective However if a woman was pregnant previously and received tetanus vaccination during that pregnancy, one dose of tetanus vaccination is enough (The Population and Family Health Project 2002)
The recommended content of care differs between the first ANC visit and subsequent visits The first visit is the most important because the women are assessed in all health aspects A comprehensive medical history is taken, general health is examined and most of laboratory tests are done during the first visit It is also the first time when the women are counselled on health promotion such as nutrition and rest The planned timing of follow up visits and a delivery plan is also discussed with the women at this time
Level of ANC adequacy
The use of ANC varies enormously between developed and developing countries Figure 1.2 illustrates the differences in proportions of women receiving ANC among different regions of the world In developed countries, most women receive ANC In contrast, only about two thirds of women in the developing world receive any ANC (UNICEF 2001) (WHO 2003) Among the developing countries, there are also large differences between regions The region with the lowest use of ANC is South Asia where only one in two women receives ANC The best among developing countries are East Asia/Pacific, CEE/CIS (Eastern Europe countries) and Latin America/Caribbean where around 80% of women receive any ANC (UNICEF 2001)
The number of visits is also much higher in the developed world than in developing countries For instance, the mean number of visits is 12 (Buekens 1993), median is 11 in the US (Backe, 2002) or 10 in Denmark (Buekens 1993) The mean number of visits is 5.3 in Malawian adolescent girls (Brabin, 1998) and four in Indian women (Bruun Nielsen 1998)
Antenatal care, by region (UNICEF, 2001)
EEC/CIS Latin America/Caribbean
Percentage of women aged 15-49 having any antenatal care delivered by trained health personel
*Data for East Asia/Pacific region excludes China
EEC/CIS: Eastern Europe countries
Industrialis Improvement in antenatal care*, 1990-200098
Sub-Saharan Africa Middle East/North
*Bases on 45 developing countries with trend data
Nevertheless, the use of ANC has been improved over the 10 year period between 1990 and 2000 The biggest improvement is in Asia with 31% and the least improvement is in Sub-Saharan African with only 5% The improvement of ANC by region is presented by UNICEF in Figure 1.3 (UNICEF 2001)
Figure 1.3 : Percentage improvement in ANC, 1990-2000
The East Asia/Pacific region is one of the regions that have better coverage of ANC among developing world with around 80% of women having ANC delivered by trained health personnel
However, due to the differences in economical development between countries, the use of ANC varies greatly across the region In more advanced countries such as Singapore and Brunei, 100% of pregnant women receive ANC There is also large coverage of
ANC in Malaysia (90%), Philippines (83%), Indonesia (82%) and Myanmar (80%) Thailand and Vietnam are within the medium range with 77% and 78% of pregnant women attending any ANC In the least developed countries such as Cambodia and Laos, the proportions are very low with only 52% and 25% of pregnant women attending ANC respectively (Erbayda 2003)
According to the Demographic and Health Survey (DHS) data of 1997 and 2002, the use of ANC in Vietnam is low but on the increase The proportion of women with at least one ANC visit increased from 71% in 1997 to 87% in 2002 The proportion of women having more four or more ANC visits increased from 15% in 1997 to 29% in
2002 There is a large variation between areas, for example only 60% of women in the Central Coast and the Central Highland had any ANC compared to 87% in the Red River Delta in 1997 (Vietnam DHS 1999; Vietnam DHS 2002)
Although the proportions of women who utilise ANC may be low, the proportions of women who receive recommended ANC content are often lower For example it has been found that not all of the women attended ANC in developing countries received tetanus vaccination, only 60% to 86% of women in 38 developing countries who attended ANC received tetanus vaccination (Buekens 1995)
The common results from studies on ANC content is that most health care providers concentrate more on the bio-medical assessment than health promotion and care provision (Kogan, Alexander et al 1994; Haas, Berman et al 1996; Peoples-Sheps, Hogan et al 1996; Gifford 2001) and it is commonly recognised that ANC content actually provided to women is far below recommended levels (WHO 2003)
There is one study that used the opinions of experts around the world to evaluate the maternal and neonatal care programs in 49 developing countries, the Maternal and Neonatal Programme Effort Index (MNPI) ANC was rated on six aspects: (i) iron/folate supplementation to prevent anaemia; (ii) detection and treatment of hypertension; (iii) detection and treatment of syphilis; (iv) vaccination against tetanus; (v) information on danger signs of pregnancy; and (vi) offer of counselling and testing on HIV
The average rating for all 49 countries was 47 out of 100 There is some variation between regions, ranging from 32 in South Asia to 51 in Francophone Africa (MNPI
The MNPI rated content of ANC program in East/South East Asia countries 39 out of
100, place it to be the second lowest rated regions, among six regions of developing countries (MNPI 2003)
Vietnam was rated 63 out of 100 for ANC content which is slightly better than average (MNPI 2003) (see Figure 1.4) Tetanus injection rated highest with 79 out of 100 Clinical procedure of blood pressure measurement received the second highest rating score of 76 In contrast, laboratory testing for syphilis received only 45 scores and testing and counselling for HIV infection received the lowest scores of 43
Some studies show that tetanus vaccination coverage in Vietnam is increasing from 72% of women in 1997 (Vietnam DHS 1999) to 85% in 2002 received two doses of tetanus vaccination (Vietnam DHS 2002).
Tetanus injection Blood pressure test
Iron folate Information on dangerous signs
Syphyis test HIV counseling and testing
Figure 1.4 : Content of Antenatal Care provided in Vietnam
There are a limited number of studies that incorporate both ANC utilisation and content
In a study in India (Bloom, Lippeveld et al 1999) 20 components of ANC were used to measure ANC adequacy The result showed that on average ANC adequacy in women was rated half of the optimum ANC
1.3.4 Indicator/s or index/es to measure ANC adequacy in Vietnam
There are several types of indicators currently used to measure ANC adequacy in
Vietnam They are not only different in type but also use different cut off points for the same type of indicators Except for any utilisation of ANC during pregnancy which has a clear cut off point and has been used widely (Vietnam DHS 1999) (Vietnam DHS
2002) (UNFPA 2000; MOH 2002), the cut off points for number of ANC visits and duration of pregnancy at initial visits vary widely between studies
Cut off point for the number of visits range from two visits (Vietnam DHS 1999)
(Vietnam DHS 2002), to three (Toan, Hoa et al 1996; Toan, Hoa et al 1996; UNFPA
2000; MOH 2002) and four (Vietnam DHS 1999) (Vietnam DHS 2002) Cut off points for duration of pregnancy at the initial visit ranges from three months (MOH 2002) to
20 weeks (Vu, Le et al 2000), or six months (Vietnam DHS 1999); (Vietnam DHS
2002) Some studies use one indicator while others use several indicators.
Factors associated with ANC adequacy
1.4.1 Factors associated with ANC utilisation
There have been numerous studies on the factors associated with utilisation of ANC all over the world (Petrou, Kupek et al 2001) (Delvaux, Buekens et al 2001; Navaneetham and Dharmalingam 2002) Several studies have been conducted in South East Asia, including the Philippines (Becker, Peters et al 1993), Thailand (Raghupathy 1996) and Vietnam (Swenson, Thang et al 1993) (Toan, Hoa et al 1996)
The common finding was that inequity existed in both developed and developing countries Women who belong to majority ethnic groups, who have higher levels of education, who have higher socio-economic status, who have fewer children, who have planned pregnancy and who reside in urban or more developed regions utilised more ANC services compared to women who belong to minority ethnic groups, who have less education, who have lower socio-economic status, who have more children, who have unexpected pregnancies and who reside in rural or less developed regions
However, few studies use the Andersen Health Behaviour Model, a widely used model in health care utilisation behaviour studies These studies were done in the United States (LaVeist, Keith et al 1995) (Perloff and Jaffee 1999), Russia (Ivanov and Flynn 1999) (Ivanov 2000) and Guatemala (Glei, 2002)
The majority of studies did not use any health seeking behaviour theoretical framework in selection and analysis of variables Therefore, factors mostly investigated were socio- economic characteristics and reproductive history Other important factors such as beliefs, availability and quality of health services, personal health behaviour and women satisfaction were not investigated
1.4.2 Factors associated with content of ANC
There have been few studies which explore the factors associated with the content of ANC services (Peoples-Sheps, Hogan et al 1996) The general finding is that there is no significant relationship between characteristics of women and the content of care provided, except that the women who begin care earlier have more laboratory tests (Peoples-Sheps, Hogan et al 1996) The more influencing factors are provider-related characteristics The public sector delivers better ANC content than the private sector (Peoples-Sheps, Kalsbeek et al 1991; Kogan, Alexander et al 1994; Gifford 2001), and physicians who are newly graduated and use commercially available ANC forms provided more services (Peoples-Sheps, Kalsbeek et al 1991)
ANC content as an outcome is defined differently between studies None of these studies used a theoretical framework for quality of care to investigate ANC content delivered to women No studies which explored the factors associated with content of ANC delivered to women in South East Asia, including Vietnam, have been identified.
Gaps on information about ANC in Vietnam
Information on ANC in Vietnam is seriously lacking on the level of ANC adequacy and on the factors associated with it This is due to the limited number of studies on ANC and to the methodologies that have been used
Although there have been several nationwide studies, DHS 1988, DHS 1997 and DHS
2002 (Swenson, Thang et al 1993) DHS, 1997 DHS, 2002), on a provincial scale (Vu,
Le et al 2000) and on a smaller district scale (Toan, Hoa et al 1996) (Toan, Hoa et al
1996), information on the use of ANC in Vietnam is still very limited These studies only use simple indicators of the number of visits and/or duration of pregnancy at initial visits to report the use of ANC None of these studies use any methods to incorporate these single indicators into a more comprehensive way to describe the utilisation of
ANC For example, proportions of women with a sufficient number of visits and have entered ANC early, the proportion of women with insufficient number of visits and/or have entered ANC late This information is important because sufficient ANC utilisation includes having the recommended number of visits and having them at the right time
Information on the content of ANC is scarce in Vietnam: with fewer studies conducted and studies only reporting the proportions of women in receipt of one or more individual procedures or items of advice None of these studies used an indicator or index to determine the proportion of women who receive all or most procedures and advice, and proportion of women who receive very few or none of the recommended procedures and advice This information is important because the number of procedures and items of advice delivered to a woman should meet the recommendations for ANC to prove effective
There has been no information on how women utilise ANC services and what ANC content is delivered to the same women during the ANC visits in Vietnam This information is important because ANC can only be effective when utilisation is conducted at the recommended levels and content of ANC is adequate
Suitable indicator/s or index/es to measure ANC adequacy
It is important for a country to have a suitable indicator or index to measure ANC adequacy A uniform indicator or index allows comparison between studies and the systematic monitoring of progress over time The indicator or index should reflect the true ANC adequacy level, and should not overestimate or underestimate the situation because of unfavoured consequences in both cases The indicator or index also should be sensitive to change to allow measurement of small improvements while monitoring progress It also should be less prone to bias and easy to use
Such indicators are lacking in Vietnam It is not only that different studies use different indicators and cut off points but also that these indicators only reflect single aspects of
ANC adequacy and that may well overestimate the true ANC adequacy level The recommendations of the Vietnamese government on ANC utilisation is to have a first visit within three months with three visits in total, while the WHO recommendation is to have a first visit within four months and four visits in total Neither of the recommendations may be optimum because the Vietnamese government recommendation stresses the important role of earlier entry to ANC while the WHO stresses the importance of the total number of visits There is a need to search for an indicator or index that balances both duration of pregnancy at entry to ANC and number of visits which is suitable for Vietnam
1.5.2 Factors related to ANC adequacy
Factors associated with ANC utilisation
There is limited information on factors associated with ANC in Vietnam There have been a few studies conducted which explored the factors associated with ANC utilisation, however they only provided information on factors related to any use of ANC (Swenson, Thang et al 1993; Toan, Hoa et al 1996) which is only one of the many aspects of ANC utilisation No information is available on factors associated with the duration of pregnancy at first ANC visit and continuing of visits In addition, as most of studies on ANC utilisation in the world, none of the studies used a health seeking behaviour model as a theoretical framework to select and analyse variables As a result, many important factors were not investigated
Factors associated with ANC content and overall adequacy
There have been no studies on factors related to the content of ANC, and also no study of factors related to the overall adequacy of ANC in Vietnam.
General Aims and research questions
The current study aimed to: a Identify and propose a set of indicators and indices to measure ANC adequacy level in all possible aspects, which have been adjusted for the situation in Vietnam b Apply the identified and proposed indicators and indices to measure ANC adequacy level in the three provinces of Long an, Ben tre and Quang ngai, using the data collected by the Vietnam Australia Primary Health Care Project (Vietnam Australia Primary Health Care Project 1999) to identify the patterns of ANC utilisation, ANC content and overall level of ANC adequacy in the three provinces c Compare results on ANC adequacy levels measured by different indicators and indices to select the most suitable indicator/s or index/es d Apply the Andersen Health Behaviour Model to investigate factors related to multiple aspects of ANC utilisation among women e Apply the Donabedian Health Quality Model to investigate factors related to content of ANC reported by women, and f Apply the Donabedian Health Quality Model to investigate factors related to overall ANC adequacy
There were several questions that the current study aimed to answer:
What were the levels of ANC adequacy in the three provinces, Long an, Ben tre and Quang ngai of Vietnam?
What was the best indicator/s or index/es to measure ANC adequacy in this setting?
What were the factors associated with ANC utilisation among women in the three provinces?
What were the factors associated with ANC content reported by women in the three provinces?
What were the factors associated with Overall ANC adequacy among women in the three provinces?
Significance
The results on pattern of ANC utilisation, ANC content and overall ANC adequacy will be helpful for future planning and intervention to improve ANC not only in the three provinces but also in similar provinces
The proposed indicators and indices can be applied to measure ANC adequacy levels in other studies in Vietnam and possibly in other similar developing countries
1.7.2 Factors associated with ANC adequacy
Factors identified to be associated with ANC adequacy will help to focus resources to improve ANC adequacy in the three provinces and other similar provinces of Vietnam
The current study is one of the very few studies on ANC to apply the Andersen Health Behaviour Model, originated from the developed world, in a developing and socialist country If the model works well, it should be applied in other studies to examine ANC utilisation and general health care services in Vietnam and in other developing countries If not, a model specific to developing countries may have to be considered
The methods used in this study can be replicated in other studies in Vietnam and in similar countries to determine factors associated with ANC adequacy The factors found to be related to ANC adequacy can serve as references for other studies to select potential variables.
Ethic approval
The study was approved for secondary data analysis by the Human Ethics and Research Committee of the University of Newcastle (the HREC number is H-705-1103) and the director of the VAPHC project provided a letter of approval for secondary data analysis (see Appendix 2.1)
CHAPTER TWO: SETTING AND DATA
Vietnam
Vietnam is located in the South East Asia region It has borders with China, Laos, Cambodia and the South China Sea, with a surface area of 332 thousand square kilometres (The World Bank Group 2003) The country is divided in to seven regions based on geography: North East, North West, Red River Delta, North Central Coast, South Central Coast, Central Highland, North East South and Mekong River Delta (Figure 2.1) Each region is divided into provinces and cities, then districts: there are currently 61 provinces and cities in the whole country, with an average of about 10 districts in each province Each district is then divided into about 10 to 20 communes The hamlet is the smallest administration unit under commune Beside the central government, there are levels of government administration from provincial to commune level At each level, there is the People’s Committee and other technical sections, including health Staffs are employed by the government Hamlet authorities, such as hamlet leaders, however are selected by villagers and not paid by the government
In 2002 the population of Vietnam was 80.5 million people (The World Bank Group
2003) The main ethnic group is Kinh allocated for about 90% of the population (2003) and mainly reside in the plain areas The remaining 10% of the population includes about 50 minority ethnic groups reside in the highland and mountainous areas Vietnam is considered one of the poorest countries in the world with Gross Domestic Product (GDP) of USD 430 per capita in 2002 (The World Bank Group 2003) For comparison, the equivalent figure in Australia is USD 27,000 (2003), which is about 60 times higher Most of Vietnamese live in rural areas and are farmers Due to economic reform since the early 1990s, the economy has fast and steady growth rates Private sectors have appeared and are growing People have more income, but the wealth is not equally distributed Gaps between rich and poor, urban and rural are widening over time
Dong Bac Bo: North East; Tay Bac Bo: North West; Dong bang song Hong: Red River Delta; Bac Trung bo: North Central Coast; Nam Trung bo: South Central Coast; Cao nguyen: Central Highland; Dong bac Nam bo: North East South; DB song Me Kong: Mekong River Delta
Source: UNICEF (2000) Vietnam: Children and Women, A Situation Analysis Hanoi, Vietnam
The Vietnamese government has installed a substantial health care system from the central to lower levels of provinces, districts, inter communes and communes in the whole country and in many areas, to the lowest level of hamlets At the central levels, there are general and specialised hospitals and institutions, mostly in big cities At province level, there is a provincial general hospital, mother and child’s health protection centre and other general health centres, such as nutrition, malaria or goitre, functioning independently with the hospital At district level, there is a district general hospital and mobile teams specialised in health prevention, work under the direction of the provincial health prevention centres and the district hospital The mobile teams form part of the hospital At inter commune level, there are poly-clinics in charge of several communes
At commune level, there are Community Health Centres (CHC) Most of the communes (95%) have CHCs (MOH 1999) A CHC has a staff of about five people including a doctor, several nurses and midwives and possibly a laboratory technician The CHCs are responsible for conducting all primary health care programs such as vaccination, ANC, malaria, diarrhoea and goitre program in addition to daily consultations For the majority of the population, CHC is the first and in some cases the only accessible health care service Staff at commune level and above are government employees and paid monthly salary
The lowest level is hamlet health workers About 43% of hamlets in the whole country have hamlet health workers (Ngo 1998) Hamlet health workers are local residents who undergo a minimum of three-months training and are supervised by commune health centres Their main work functions are health education and prevention They are involved in reproductive health, child health program, nutrition, vaccination, environment hygiene and first aid Before economic reform early in the 1990s, they were paid by the communes but due to the effect of economic market, many of them were no longer paid Therefore their functioning has been weakened
Since economic reform, there is a growing but still modest private health sector, mostly in the forms of drug sellers and government health workers working after hours There are a limited number of private facilities operating full-time and there are very few private hospitals
2.1.3 Health care costs and health insurance
Despite the huge number of personnel and infrastructures, the health budget per capita is low, at VND 62,000 or USD 5 annually (MOH 1999) To increase the budget for health, since economic reform, the government has introduced health care fees for which patients have to pay a portion of health care costs This amount varies between types of services and providers Generally, services at higher levels (eg Province compared to district) and more specialised are more expensive People with health insurance do not have to pay for basic health care For more expensive health care, insured people have to pay 20% of the cost Free basic health care is also extended to children younger than six years old, people with special circumstances or special diseases, people resident in mountainous areas, people within three years of migration to an economic zone, and very poor people (Tran 1998) Only 10.5 million people or 14% of the population had health insurance in 1999 (MOH 1999) Most of these people were government workers, for whom health insurance is compulsory or people who live in the urban areas The rest of the population, mostly living in the rural areas, do not have health insurance This group is generally poorer than the insured group
In addition to health care costs, there are costs of transportation, accommodation and food for accompanying family members Many average and poor families can not afford basic health care People may delay seeking health care and purchasing/attending treatments Some families incur debts with high interest to pay for health care and become poorer (Segall 2000)
At lower levels, most of the health care services are basic health care and many of them are free of charge, or only incur low fees Many health workers, especially at lower levels cannot live with the low government salaries As a consequence, they take other jobs or run private businesses and many of them are not present at the health care centres for their full expected working hours, limiting the true availability of care for poorer people
Health in Vietnam is much poorer than in developed countries, but is comparable to other developing countries in the region Life expectancy is around 70 years which is similar to the 69 years in Thailand and in Indonesia (2003), but much lower than the 80 year expectation in Australia Vietnam has many important health problems which are typical to a developing country, including infectious diseases such as malaria with more than 340,000 cases per year, tuberculosis with almost 90,000 cases per year and recently HIV with 17,000 cases per year (MOH 1999) In addition to infectious diseases, there are important non-infectious health problems such as iodine deficiency with 43% of population at risk (UNICEF 2003) and malnutrition, suffered by 40% of children under the age of five (UNICEF 1999)
The main reproductive health activity in Vietnam is the massive family planning campaign which was installed since 1960 The government encourages families to have one or two children and contraception use to slow down the population rate increases These resulted in a reduction of fertility rate from 3.8 in 1989 to 2.3 in 1999 (1996) (UNFPA 2004)
However, the government recognised the limitation of its population program which has a narrow focus on family planing and fertility reduction Other reproductive health activities such as maternal health appear to have received much less effort
Consequently, maternal mortality remains one of the main health problems with 165 per 100,000 live births in 2002 (MOH 2002)
The Vietnamese Ministry of Health identified the five most serious obstetric complications that cause maternal deaths in the country: haemorrhage, eclampsia, tetanus infection, uterine rupture and post partum infection (MOH 2002) A large proportion of those complications could have been prevented through skilled, assisted delivery and ANC with detection and treatment of complications, iron supplement, tetanus vaccinations and discussion of safe delivery plans, self care for mother and baby There were 96% of women who had deliveries assisted by trained health workers in 1999, but only 72% of women had any ANC in 1999 (MOH 1999)
Another problem is the high rate of abortion In 1999, there were approximately
800,000 women who had voluntary abortion/menstrual regulation or one voluntary abortion/menstrual regulation per two live births (MOH 1999) A lack of trained personnel and standard equipment can lead to mortality and morbidity for women due to unsafe abortion This high rate of abortion highlights the need for education on family planning ANC provides one opportunity to provide further education to women
Long an, Ben tre and Quang ngai provinces
The Vietnam Australia Primary Health Care Project for Women and Children was funded by the Australian Agency for International Development (AusAID) to improve primary health care for women and children in three Vietnamese provinces: Long an and Ben tre in the Mekong delta region, and Quang ngai in the Central Coast Table 2.1 provides general information, health system and reproductive health data from these provinces (MOH 1999)
Long an and Ben tre are both in the Mekong Delta region and both close to Ho Chi
Minh city, the largest economic centre of the country Quang ngai is located in the
Central Coast, one of the least developed regions and far from any big economic centre
Quang ngai is poorer compared to the other two provinces (GDP of VND 2.2 M per capita compared to VND 4 M for Long an and VND 3.5 M for Ben tre)
Table 2.1: Information on three provinces: Long an, Ben tre and Quang Ngai
Long an Ben tre Quang ngai
Health budget per capita (VND) 22.000 17,000 23,000
Number of communes without CHCs 7 0 33
Total number of women at reproductive age in the province
Fertility rate in the region 2% 2% 2.5%
Number of people using contraceptive methods 54,000 69,000 62,000
Source: MOH (1999) Vietnam Health Statistic Year Book
The three provinces have the similar number of communes but Quang ngai has a significantly higher number of communes without CHCs than in the two other provinces (33 compared to 7 in Long an and 0 in Ben tre) Quang ngai has a higher birth rate of 2.5% compared to 2% in the other two provinces There is much less numbers of abortion/menstrual regulations in Quang ngai than other two provinces.
Data from the Vietnam Australia Primary Health Care Project
At the beginning of the Vietnam Australia Primary Health Care Project, baseline surveys were conducted to gain information on health care for women and children (Vietnam Australia Primary Health Care Project 1999) The baseline surveys were conducted by staffs at the Ho Chi Minh City Institute for Hygiene & Public Health, Tu
Du Hospital for Obstetrics & Gynaecology, Paediatric Hospital No 1, University
Training Centre, Ho Chi Minh City Child Nutrition Centre and Health Centres of provinces where the survey was conducted, in conjunction with Vietnamese and
Australian consultants The author of this thesis was not involved in the design or development of these surveys, or the data collection
Information collected regarding women’s health included ANC, delivery, contraception, nutrition and general health Information collected regarding child health included nutrition, general health, diarrhoea, respiratory infection and utilisation of health care services Basic information from the surveys was reported to AusAID, but much of the data had not been analysed in depth The data on ANC analysed in this thesis is only one part of the whole data set Permission to further analyse the data was obtained from AusAID, refer to Appendix 2.1 for a copy of the permission letter
The following information about the surveys was adapted from the report of the
Vietnam Australia Primary Health Care Project Only parts relevant to ANC has been presented
Cross sectional household surveys were conducted over a two week period in December
1998 in Long an and Ben tre and in January 1999 in Quang ngai Since the aim of the survey involved collecting information on women and children’s health, the primary respondents were women of reproductive age (15-49 years) and women with pre-school aged children living at home Each of these surveys used a standard “probability proportionate to size” cluster sampling method Since the project only operated in parts of the provinces, a representative self-weighted sample was selected from the six project districts and another from the remaining non-project districts in the province The sampling method did not take account of “project status of communes” within the project districts to avoid complication of sampling procedures and because of the potential expanding of the project
Many indicators of women and children’s health were collected As the assessment of children’s health was not an a-priori aim of the current study, only women’s indicators are presented
Women were asked about the following information:
• Having any antenatal checks for last pregnancy during the three years prior to interview
• Number of ANC checks during last pregnancy
• Duration of pregnancy at first antenatal check
• Taking iron / folate supplements during last pregnancy
• Usual places where ANC was delivered
• Practitioners as usual provider for ANC
• Numbers of previous pregnancies and deliveries
• Food availability in household during pregnancy
• Night blindness during last pregnancy in the last three years
Head of household were asked about the following information:
• Total number of household members
• Number of people with health insurance
• Level of total household expenditure
• Level of household expenditure on food
• Water source and sanitation facilities in house
• Time and distance to usual health care provider
Representative of local health centres were asked about:
• Types and number of reproductive health services offered
Communes and province level factors
Surveyors recorded the following information:
The reference population for the survey were households from the three project provinces, Long An, Ben Tre and Quang Ngai In the absence of provincial household sampling frames, a multi-stage cluster sample was constructed starting with primary stratification by project versus non-project districts and “probability proportionate to size” (PPS) selection of 30 clusters within each of these groups of districts The samples for each group of districts were self weighted Provincial level estimates of program indicators required the application of population size weights for each group of districts Thirty-three districts in three provinces were included in the survey (14 in Long an, seven in Ben tre and 12 in Quang ngai)
Stage List Used Sampling Method Total
One Groups of districts within each project province
Stratification by project supported versus non- project supported districts 6 groups (2 groups in each province)
Two Communes from stage 1 30 communes by systematic random sampling with probability proportionate to population size within each group of districts & stratified by district or 60 clusters in Long an and Quang ngai Ben tre had only one non project district, the provincial town, and so only five communes was collected from this district made up the total number of 35 clusters selected
155 (60 communes in each province of Long an and Quang ngai, 35 in Ben tre)
2 hamlets (sub-communes) by simple random sampling within each commune
Four Households from stage 3 12 households by systematic random sampling All women of reproductive age (15-49) and preschool aged children for survey
The survey aimed to measure several different health, nutrition and program indicators Sample size estimates were prepared for a selection of these program indicators assuming an average of five persons per household and 0.7 children younger than five years of age per household These estimates are reported in Table 2.3 They were calculated assuming a prevalence of the indicator (p), level of absolute precision (d) and a design effect (DEFF) as recorded in the Table, and a level of confidence of 95% (Z=1.96) The sample sizes were calculated using the following standard sample size formula for proportions in cluster surveys:
Table 2.3: Estimated sample sizes required for various survey indicators
Design effect ( DEFF) No of
No of HHs to contact
Proportion of women having > two antenatal visits
12.5% Women who have delivered in last three years
Proportion of women delivered by trained health professional
12.5% Women who have delivered in last three years
Proportion of women satisfied with
12.5% Women who have delivered in last three years
Proportion of women satisfied with delivery care services of health professionals
12.5% Women who have delivered in last three years
* Based on two week recall
Reproductive Health Survey (VNRHS-95) National Committee for population and Family Planning of Vietnam & GTZ, reported 55% of women in four provinces having two or more antenatal visits, pp113
Based on these calculations a total sample size of 678 household interviews per group of districts would be required to examine all but one of the program indicators This number was increased to 720 to allow for non-consent and missing data of 5% to 7.5% The total sample size for Long an and Quang Ngai provinces would be 1440, and 840 for Ben Tre as there is only one non-project district The sample size required for those indicators based on mean values (e.g BMI or number of antenatal visits) was considerably lower than that required for proportions
The survey was conducted as a collaborative effort between staff from selected institutions in Ho Chi Minh City, the provinces to be surveyed, the local consultants and an Australian epidemiologist
The project institutions participating in the baseline survey included:
• Institute for Hygiene & Public Health in [lead organisation]
• Tu Do Hospital for Obstetrics & Gynaecology
The following provinces are involved in the baseline survey:
• Long An Province Health Services, Mother and Child Health/ Family Planning Centre and Community Health Centres (CHCs)
• Ben Tre Province Health Services, Mother and Child Health/ Family Planning Centre and Prevention of Diseases Centres
• Quang Ngai Province Health Services, Mother and Child Health/ Family Planning Centre and Prevention of Diseases Centres
Survey design and analysis team
The survey design team was responsible for preparing the survey questionnaires and other data collection instruments, the methods for assessing nutritional status, the design and procedures for sampling households, the methods for data entry and data cleaning
Survey training and monitoring team
This team was responsible for preparing and conducting the training of provincial survey teams and for monitoring survey field activities to ensure the data is collected according to the protocol Survey monitoring was conducted by both the project field supervisors and provincial field supervisors These field supervisors stayed in the field area for most of the survey except for the anthropometry team who were present for the first three days of the field work
The provincial survey teams each consisted of one interviewer and one anthropometrist who were responsible for interviewing and measuring (respectively) the respondents from the selected households in four communes over the 16 days of the survey There were 17 teams with 34 field interviewers in both Long An and Quang Ngai provinces, and nine teams with 18 field interviewers in Ben Tre province
This team was responsible for the design the of the data entry system and data cleaning procedures
In preparing the questionnaire, the team reviewed existing forms for suitable items, especially when an analysis of these questions was available An important source of questionnaire items for the baseline survey were those questionnaires used in the Vietnam Reproductive Health Survey 1995 (NCPFP and GTZ 1996) The Health Survey used a modified set of the health and demographic items which were translated into Vietnamese All forms were pre-coded and structured in lay-out to facilitate easy recording of data in the field and data entry by computer operators
The draft questionnaires were pilot tested by two focus group discussions involving 10 mothers, and through 30 individual interviews of mothers in communes in Long An
Based on information gathered during these discussions, difficult-to-understand questions were revised and interviewer instructions were added to the questionnaires The questionnaire design team also sought input from representatives of the provinces and senior members of collaborating institutions from Ho Chi Minh City in a workshop
A record of all discussions and suggestions was translated into English, reviewed and considered prior to finalising the survey forms Table 2.4 outlines the type of questionnaires used in the survey to collect information on ANC in women
The survey teams were trained over one week for practical experience in using survey forms and using anthropometry measure equipments
The clusters were selected based on information provided by each province
Information requested included a list of the commune names organised by district, the total population of the commune, the total number of households and the cumulative population for the province
This information was used to select the sample of communes by the PPS method described previously in the sampling design section The lists of selected communes for both project and non-project districts was returned to the respective provinces for the preparation of lists of hamlets in these communes These lists of hamlets were then returned to the baseline survey team for the random selection of two hamlets per commune The province survey team then listed all the households in these hamlets The households for the survey were identified by the baseline survey team in Ho Chi Minh City using the listing of households in each hamlet and systematic random sampling to select 12 households and eight stand-by households
Table 2.4: Questionnaires or forms used in the surveys
KTC Cluster Control Sheet A form for monitoring the questionnaires collected in one hamlet
PB Dispatch Control Form This form was completed for each package of forms sent to
A form to explain the purpose of the survey and seek respondent’s consent to participate
This form recorded information about all members of the household such as age, gender and education and assisted the interviewer identifies which members needed further information to be collected
K Socio-economic characteristics of the household
This form collected information about the socio-economic status of the household
P Women's health questionnaire This form collected health information about the female respondents aged 15 to 49 years, including reproductive history, antenatal and delivery care, contraceptive methods and family planning and nutritional status
S Birth History This form recorded information about all births of the female respondents and the survival of these infants
C Health Services Survey This form collected information about health services provided by public and private health providers in the hamlet
CHAPTER THREE: LEVEL OF ANC ADEQUACY
Background and literature review
The background and literature review describes methods previously used world-wide to measure ANC utilisation, ANC content and ANC adequacy, in addition to a description of studies that utilised these methods
Although there are widely reported methods in the literature which measure ANC utilisation, there is no commonly agreed “gold standard” Even if there were such a standard, it may still not be suitable for Vietnam, a country with low ANC utilisation
In contrast to the numerous number of methods used to measure ANC utilisation, there are very few methods to measure ANC content and none of these methods has been applied widely
Measurement of overall ANC adequacy which combines both utilisation and content has been reported even less in the literature In general there is no simple way to measure ANC adequacy
Methods available to assess ANC utilisation
Over time different methods to measure of ANC utilisation have been developed around the world: ranging from simple indicators of month of pregnancy at first ANC visit or number of visits, to more complicated methods which combine these indicators into indices, which are then modified and extended in different ways The simple methods are mostly utilized in the developing world while the complicated ones are mostly utilized in the United States (US) and other parts of developed world Following are brief descriptions of these methods:
ANC utilisation assessment is based on any of the following indicators:
• Whether or not women receive any ANC that includes any type of ANC or restricted ANC delivered by modern health care providers (doctors, midwives, nurses) rather than traditional healers
• Duration of pregnancy at initial contact
Except for receiving any ANC for which the cut off point is clear, the cut off points for the two other indicators can vary The recommendation for the time of first ANC visit varies from two months (Mortimer 1991; UNFPA 1998), to three (MOH 2002), or four (WHO 1994) or even six months (DHS 1997; DHS 2002) The total recommended number of visits also varies from three visits (MOH 2002) to four (WHO 1994) to 13 visits (Greg and Donald 1987; Greg and Kotelchuck)
Methods that combine indicators into indices
There have been several indices developed in the US, based on a combination of two indicators: the duration of pregnancy at initial contact and the total number of ANC visits The differences between these indices are not only on the number of categories, the cut off points but also on the ways of combining indicators
• The Kessner Index or the Institute of Medicine Index of Utilisation of Care
(Kessner, Singer et al 1973): This index was first introduced in 1973 based on the recommendations of the American College of Obstetricians and Gynaecologist (ACOG) The index combines the duration of pregnancy at first ANC visit and the number of ANC visits adjusted for gestational age at delivery, to classify ANC as adequate, intermediate and inadequate Another element of this index is the type of carer, for which only private care is considered good enough However this element is widely ignored by researchers A woman who has reached full-term in her pregnancy is defined as having “adequate” care if she has nine or more visits and begins care during the first trimester; “inadequate” if she has four or less visits and begins care during the third trimester; or “intermediate” for any other combination This index has been widely used in many studies over several decades
• The Graduated Index of Prenatal Care Utilization or R-GINDEX Index (Greg and Donald 1987; Greg and Kotelchuck): In 1987 Alexander expanded the Kessner Index from three to six categories The added categories are “no care”, “missing” and “intensive care” The “no care” group is for women who have no ANC visits at all The “missing” group is for women with missing data The “intensive care” category is for women with more care than the recommended standard (16 visits or more) Women belonging to this “intensive care” group may have abnormal pregnancies and need more care
• The Adequacy of Prenatal Care Utilisation or APNCU Index (Kotelchuck 1994): In
1994, Kotelchuck criticised the Kessner Index for not considering the expected number of visits after the first visit A woman may enter ANC late but have more visits than expected after this point Kotelchuck proposed an index which calculates the proportion of actual visits compared to the expected number of visits for a given pregnancy duration after the initial visit This index also added the “adequate plus” category to the three existing categories ANC is considered to be “adequate” if the first initial contact is made within the first four months and the number of actual visits is equal to a proportion in the range of 80% to 109% of the expected number after the first visit “Adequate plus” is when the number of visits makes up 110% or more of the expected numbers
However the APNCU Index has been criticised as not appropriate for the assessment of the relationship between ANC and LBW by Koroukian and Rimm (Koroukian and Rimm 2002) The reason for this criticism is that the ACOG recommends that nearly one third of the total visits occur during the last four to five weeks of gestation Thus the lower the gestation age, the lower the number of visits expected
Women who have a short gestation period are likely to have less than the expected number of visits and thus are classified as “adequate plus” However, short gestational age is associated with LBW Therefore there is more LBW in the adequate plus category, which goes against the common theme that increasing ANC reduces LBW There has been some more debate over this issue between authors (Koroukian and Rimm 2003; Kotelchuck 2003)
• US Public Health Service Expert Panel on Prenatal Care (PHS-REC) (Mortimer 1991): While other indices are based on the recommendation of the ACOG, this index is based on the opinion of an expert panel This index differs from those described earlier in that it recommends a pre-conception visit, fewer but earlier visits and more importantly it incorporates the women’s risk of developing complications or poor outcomes Women are considered to be at higher risk if they have not previously given birth to a child The level of recommended ANC starts before conception to two months after conception, and the adequate number of visits is seven for multiparous women and nine for primiparous women
• The Prenatal Care Evidence-Based Index (PEI) (Koroukian and Rimm 2003): This index uses percentiles of the number of visits by gestational age to classify ANC utilisation Thirtieth (30 th ), 60 th and 90 th percentiles are the cut off points to classify
“poor”, “fair”, “average” and “above average” care This index was developed specifically to investigate LBW
The simple methods for assessing ANC utilisation based on single indicators of any use of ANC, duration of pregnancy at initial contact and total number of visits, are widely used in the developing world due to their ease of use and interpretation However, these methods do not provide the most comprehensive measure of ANC utilisation For example, a woman may enter ANC early but have a very low number of visits It is difficult to determine the level of utilisation that the woman receives
Method
3.2.1 Proposed indicators & indices to measure ANC in Vietnam
Methods to measure ANC utilisation and content will be proposed separately They will then be combined in a method to measure adequacy The reason for using several indicators/indices instead of only one index is that in the search for the most suitable ones, it is necessary to use all of the available methods and compare them with each other The cut off points for the classifications were decided based on the available recommendations (WHO and Vietnamese government) and results of preliminary analysis of the data
To compare indicators/indices between studies the total number of pregnant women was used as a common denominator for all outcomes, instead of number of women with any use of ANC for certain outcomes
Indicators and index to measure ANC utilisation
This indicator measures the proportion of women receiving any care at any time during the pregnancy There were two categories:
• Yes, have any ANC of any kind
• No, have not had ANC of any kind
Duration of pregnancy at first ANC visit
This measure classified duration of pregnancy at first ANC visit WHO recommends initial visits during the first four months (WHO 1994) while the Vietnamese government recommends a visit within the first three months (MOH 2002) Many women may not know about their pregnancy during the very early stages, especially in many rural areas of Vietnam where knowledge of women about pregnancy is poor and where access to pregnancy tests and ultrasound are not always possible In addition, with the low ANC utilisation in Vietnam, four months seems to be less conservative and more reasonable Duration of pregnancy at entry to ANC is classified into five categories:
• Early: first visit within four months of pregnancy
• Intermediate: first visit occurs after four months but before seven months of pregnancy
• Late: first visit occur after seven months of pregnancy
• Missing: the woman could not remember the duration of pregnancy at first visit or the information was not recorded
• No care: the woman did not have any ANC
This measure is of the total number of ANC visits during pregnancy The Vietnamese government recommends three visits (MOH 2002) while WHO recommends four visits Given the low ANC utilisation in Vietnam the less conservative guideline of three visits from the Vietnamese government was utilised, but the proportion of women with four or more visits as recommended by WHO (WHO 1994) was also reported Number of ANC visits was classified into five categories:
• Enough: three or more visits
• Missing: the woman could not remember the total number of ANC visits or the number was not recorded
The utilisation index is a composite measure of the number of visits and the duration of pregnancy at the time of the first visit The proposed classification incorporates the time of initial contact and the total number of visits, adjusted for Vietnam
The following five categories were proposed:
• Sufficient: first visit within the first four months of pregnancy and three or more visits in total This classification is consistent with the WHO guideline for a first visit within four months and with the Vietnamese guideline for minimum of three visits
• Intermediate: first visit between five and six months and/or two visits in total
• Insufficient: first visit between seven and nine months and/or one visits in total
• Missing: any missing value on duration of pregnancy and/or number of visits
• No care: no ANC visit
The category “adequate plus” was not included because the small proportion of women who had more than four visits in this study (8%) and in Vietnam (15% from DHS
1997) The maximum number of visits in this study was nine and there was no women with 16 visits or more to be defined as adequate plus (Greg and Donald 1987; Greg and Kotelchuck) Refer to figure 3.1 for a summary of the proposed index
Figure 3.1: Proposed index to measure ANC utilisation among women in Vietnam
Values Missing Missing Missing Missing
Late (7-9 months) Insufficient Insufficient Insufficient Missing
Intermediate (5-6 months) Intermediate Intermediate Insufficient Missing
Duration of pregnancy at first visit
Indicators and index to measure ANC content
The ANC procedures and information investigated in the surveys did not fully comply with the WHO or the Vietnamese government’s guidelines because they did not include items such as height measurement, clinical examination of anaemia, advice on hygiene and breast feeding that are recommended in both guidelines However the procedures and information investigated make up the basic content of ANC recommended by both the WHO and the Vietnamese government, although some items were only recommended by the Vietnamese government such as testing of urine protein and examination of vagina Ultrasound is not recommended in both guidelines but was asked during the surveys
Based on the available data, 13 items were included in this analysis There were seven items on bio-medial assessments, four items on care provision and two items on health promotion/education Except ultrasound, all of the items selected are recommended by the Vietnamese government (MOH 2002) Most of them are commonly provided to women
The order of the visits when the items occurred was not specified Since the number of visits was not high (three as median), the presence of each item was assessed regardless of the order of visits or the timing of when the services were delivered
• Monitoring of foetal heart rate
Ultrasound was chosen because it has been used in large cities, although it is not a common practice, especially in rural areas due to resource constraints Due to the later classification of ANC, which did not require receiving all procedures as having fair ANC, the inclusion of ultrasound would not greatly affect the overall results
Vaginal examination was chosen because it is recommended by the Vietnamese Government, although it may not be the case in developed countries This study aimed at comparing what the women received according to government guidelines therefore the inclusion of this variable was justified
Fetal movement is a simple and effective way to monitor health of the babies but was not included because women were asked about what they received during ANC visits rather than about their own observations of their unborn babies’ health
• Provide tablets or advice on Iron/folate supplement
Results were reported using the following indicators:
- Percentage of women who reported they had received each item
- Percentage of women who reported they had received all or part of each procedure
- Percentage of women who reported they had received all or part of each item of advice
- Percentage of women received all or part of the total set of procedures and advice
In addition, ANC content was reported in terms of the total number of procedures and information provided to the same women ANC content was classified into five categories:
- Fair: ten or more procedures and items of information out of 13 (>75%)
- Intermediate: seven to nine procedures and information (50-75%)
- Poor: zero to six procedures and information (< 50%)
- Missing: unable to recall or not recorded
- No care: no ANC visit
The information on ANC content was not reported in terms of the proportion of women with all procedures or advice, due to recall bias potential The women may be more likely to remember most of the procedures/information than all of them Some women might report many of the procedures but very few items of advice, or vice versa It is hard to classify the level of ANC content for these women
Proposed index to measure overall adequacy
The ANC adequacy measurement was based on ANC utilisation and the content of services (refer to Figure 3.2) ANC adequacy was classified into five categories:
• Adequate: sufficient utilisation (three or more visits, initial visits within 4 months) and fair ANC content (10-13 services)
• Intermediate: intermediate utilisation (two visits and/or initial visits after 4 months to 6 months) and intermediate ANC content (7-9 services)
• Inadequate: insufficient utilisation (one visits and/or initial visits after 6 months) and poor ANC content (0-6 services)
• Missing value: any missing value on ANC utilisation or ANC content
• No care: no any ANC
Figure 3.2: Classification of ANC adequacy
Missing Missing Missing Missing Missing
(0-6 items) Inadequate Inadequate Inadequate Missing
(7-9 items) Intermediate Intermediate Inadequate Missing
(10-13 items) Adequate Intermediate Inadequate Missing
Intermediate (>4 to 6 months or 2 visits)
Results on ANC adequacy levels were reported by province plus the total for the three provinces overall ANC levels in the three provinces was measured by different methods and compared with each other
3.2.3 Selection of the most suitable index to measure ANC adequacy
ANC adequacy levels, measured by different indicators and indices, were compared to each other An indicator/s or index/s was chosen based on the following criteria:
• Provide more comprehensive information on ANC adequacy
• Reflect the true ANC adequacy level in Vietnam, not over or underestimate
Data was analysed using the “survey commands” in STATA version 8.0 (StataCorp
Results
There were 1335 eligible women who responded to the survey Among these women,
917 women had at least one or more ANC visit The variable for the duration of pregnancy at first ANC visit had 40 missing values (3%) Among them, 17 women with no response and 23 women who did not remember the time when they had the first visit Since duration of pregnancy was used to construct ANC utilisation index, there were also 40 missing values on ANC utilisation outcome These numbers were small, and would not affect the internal validity of the result
The approach used to reduce the number of missing values for classification of ANC had resulted in 86 missing values for ANC content classification instead of 171
The overall adequacy index was constructed based on the utilisation and ANC content and therefore had a larger number (120 cases or 9.7%) of cases with missing values, compared with utilisation index and categorical content
Due to the large number of women with missing values on procedures and information provided for ANC content (n1), there were in total 198 women (16%) with any missing value on duration of pregnancy or on the number of services reported for ANC content Among them, 93 women (7.6%) had missing values on both outcomes, 27 women (2%) had missing value on duration of pregnancy only and 78 women (6.5%) had missing value on number of ANC content reported only
Table 3.8: Missing values for different indicators/indices for ANC adequacy n %* Any use of ANC Categorical 0 0%
1st ANC visit Continuous (month)
Number of visits Continuous (number of visits)
Content Continuous (number of procedures/information) Categorical
*Weighted percentage using survey commands
Demographic characteristics of all women and women with missing values
Demographic characteristics of (a) all women and (b) women with any missing value/s are presented in Table 3.9 Since the survey was conducted in three separate provinces by three different field work teams, the information on ANC adequacy is presented according to province, plus total
Respondents to the survey were young, with a mean age of 29 years More than half of the women aged between 20-29, one third were aged between 30-39, a small proportion aged younger than 20 or older than 40
The majority of women were married (97%) Quang ngai had slightly more non-married women (6% compared to 3% and 0.5% in Long an and Ben tre respectively, p