INDICATORS OF MONITORING HEALTH EQUITY NASHID KAMAL, Professor, Department of Population-Environment, INDEPENDENT UNIVERSITY,BANGLADESH INTRODUCTION Equity encompasses notions of fairness, justice and equality and implies that everyone should have an equal opportunity to attain their full potential for health. Health refers to both physical and psychological health status. Health determinants include both proximate factors, with directly and relatively immediate links to health outcomes, and factors that may occur and act quite distantly from the outcome, near the beginning of what may be a long and complex causal chain (Braveman,2003). Monitoring equity in health requires comparing indicators of health and its social determinants among social groups with different levels of underlying social advantage, ie groups who occupy different positions in social hierarchy. Monitoring is a particular kind of research that involves study of a question over time, requiring the ongoing collection of data. Monitoring is explicitly action oriented, with the primary purpose of keeping policies or programmes on course in relation to an explicit or implicit set of criteria (ibid). In addition to meeting standard scientific criteria for research in general, a monitoring system has other more specific requisites, including clear relevance for policy; simplicity, so that local personnel can use the techniques for data collection and analysis in an ongoing routine fashion; affordability, considering the 1
resources required for all phases of the monitoring process,not only data collection, sustainability, allowing the data collection and analytic activities to be repeated routinely over time and timeliness, considering the time lag between data collection and availability of findings to inform policy. These criteria can be difficult to fulfill, particularly in combination. In the late 1980s, Braveman and colleagues worked on developing a system to monitor social inequalities in health and health care in USA. Their experience may be summarized as follows: `Soon it became clear that rather than acquiring new equipment or technology or creating new data sources, what was needed first was to more fully use the existing data sources; only then would the creation of the new sources make sense (ibid). MONITORING HEALTH EQUITY During the later 1990s,the WHO supported demonstration projects in three lower-income countries with the goal of developing systems for monitoring health equity. The experience in these countries, led the policy makers to believe that existing data could be fruitfully utilized and analysed in an effective manner by exploring all their capabilities. In one of these countries for example, relevant data had been disaggregated to both province level and district level for some time, but there was no routine process of presenting and considering the implications of this information. A monitoring system which focuses specifically on equity in health has several prerequisites ;these include appropriate research questions, adequately identified social groups to be compared, relevant indicators of health and its determinants, appropriate 2
estimates of disparities in those indicators between the different groups, and an effective process for interpreting and applying findings (Braveman,2003). The figure given below illustrates schematically several steps in an ongoing process for monitoring equity and is intended to call attention to several key issues that are specific to monitoring health equity Fig 1;Eight steps in policy oriented monitoring of equity in health and its determinants Step 1 Identify the social groups of a priori concern. In addition to reviewing the literature, consult representatives of all social sectors and civil society, including advocates for disadvantaged groups. Step2 Identify general concerns and information needs relating to equity in health and its determinants. Again, in addition to the literature, consult representatives of all social sectors and civil society, including advocates for disadvantaged groups. Step 3 Identify sources of information on the groups and issues of concern. Consider both qualitative and quantitative information. Step 4 Identify indicators of (a) health status (b) major determinants of health status apart from health care, and (c) health care financing,resource allocation, utilization, and quality) that are particularly suitable for assessing gaps between more and less advantaged social groups. Step 5 Describe current patterns of avoidable social inequalities in health and its determinants. Step 6 Describe trends in those patterns over time. 3
Step 7 Generate an inclusive and public process of considering the policy implications of the patterns and trends. Include all the appropriate participants in this process (eg all relevant sectors, civil society, NGOs) Step8: Develop and set in motion a strategic plan for implementation, monitoring, and research, considering political and technical obstacles, and including the full range of appropriate stakeholders in the planning process.(who,1998). INDICATORS OF HEALTH AND ITS DETERMINANTS Choosing indicators of health needs serious judicious approach,since all the measures are not straightforward and ethical criteria must be met before firm decisions are made. According to Braveman (2003) `the indicators must be included in data sources that can be expected to be accessible over time and across the social groups of interest; information on the health measure must be disaggregated at the appropriate level (individual, household, neighbourhood, municipality, province) for the questions being asked. Monitoring health equity requires using indicators to reflect not only health status itself, but also major determinants of health, such as economic resources. Apart from wealth, education, status of women, supply of clean water, sanitation, food security, housing and health care are examples of other social determinants of health that would be important to monitor in most settings. In many cases, a significant improvement in capacity to monitor health equity can be made by introducing modest modifications into the existing data sources. Most countries have population based national household surveys characterizing either living conditions or health. The introduction of one or two socioeconomic measure(s) in such health 4
surveys and addition of health measures in economic surveys,can make these sources powerful resources for monitoring equity. THE CONTEXT OF BANGLADESH For Bangladesh, the data on equity is extremely limited and most of the studies that have been conducted are based on secondary analysis of existing data. Very few studies have actually been conducted to solely study the indicators of health equity. Bangladesh has achieved significant progress in the area of primary health care and has been successful in raising average longevity of the population. However, the situation of female health continues to remain poor and girls and women are worse off than men and boys. Health care for women is restricted to their reproductive health and their general health remains largely neglected. Poverty, illiteracy and limited access to health services coupled with early marriage, absence of premarital counseling, pressure for early child bearing and poor nutritional status are major contributing factors for the inequity in health status of women compared to men (Khan et al.,2004). Discrimination against girls mostly due to son preference and limited access to nutrition and health care services have resulted in inter generational consequences on women and endangered their current and future well being. Wide gaps exist between the health indicators of men and women. Nearly 70 percent of the women suffer from caloric deficit including pregnant and lactating women. One recent study however, finds that among boys and girls ranging in age from 6 months to 71 months, there is no gender difference in nutritional status which is encouraging (Equity Dialogue,Vol 1).The study uses data from the Child Nutritional Survey (CNS) 2000,Bangladesh Bureau of Statistics (BBS) and finds that the percentage of stunted and underweight children in rural areas of Bangladesh is higher compared to 5
children in urban areas. The percentage of stunted and underweight children is higher in households with lower per capita annual expenditure. Out of six administrative divisions in Bangladesh, Barisal had the highest percentage of malnourished children, while Khulna had the lowest (Equity dialogue, Vol 1 No 3 and 4).Mother s education was found to be inversely related to malnutrition of children. Mother s who had higher levels of education had lower proportion of malnourished children in the household. 6
Fig 1 Malunutrition of children (6-71 months) by area of residence, 2000 Rtural Urban % 60 50 40 30 20 10 0 51 53 42.3 40 12 10 Wasting [WHZ<-2SD] Stunting [HAZ<-2SD] Underweight [WAZ<-2SD] Fig. 2 Malnutrition of child by mother's education, 2000 Never attended school Passed class I-V Passed class VI-IX Passed SSC or above 70 % 60 50 40 30 20 10 13.3 13 11 8.7 56.4 57.7 47.3 38.2 22.2 49.4 42.7 26.9 11.7 10.7 8.3 5.7 51.6 51.6 35.1 26.4 45.3 28.3 19.5 18.8 0 Wasting [WHZ<- 2.00]Rural Stunting [HAZ<- 2.00] Rural Underweight [WAZ<-2.00]Rural Wasting [WHZ<- 2.00]Urban Stunting [HAZ<- 2.00]Urban Underweight [WAZ<-2.00]Urban 7
However, life expectancy of women and men is almost equal now (Khan et al.,2004).evidence from Matlab area where the International Centre for Diorrheal Disease Research, Bangladesh has been operational finds that improvements have occurred in child survival rates (children under 5 years age). The data comes from a 1983-85 cohort and is compared with a 1993-95 cohort. This study finds that in the ICDDR,B area the increase in poor-rich ratio of under five mortality was mainly because of the increase in poor rich ratio for infant mortality (1.12 to 1.54).In the government area, the increase in poor rich gap of under 5 mortality is explained by the increase in poor rich gap in both infant mortality and the 1-4 year olds mortality (Razzaque et al.,2004). The sex bias in child mortality seems to have disappeared in Matlab. This study shows that the earlier cohort had the difference but it disappeared for the recent cohort in both ICDDR,B and the government areas. It may be noteworthy to mention that the sex differential during infancy is more tilted towards the male child (more male children die compared to females 1.09 changed to 1.05 for recent cohort).there has been a reduction in boy-girl mortality of the 1-4 year olds which has been reflected in the elimination of sex differentials among the under fives mortality in the Matlab area. Though the infant mortality rate has reduced, the neonatal mortality rate remains a concern and this is associated with low rate of institutional delivery (9percent),low attendance of deliveries by skilled personnel (12 percent,bhsmms/2001),high incidence of low birth weight (LBW) estimated between 40-50 percent and low utilization of antenatal care (48 percent,bhsmms/2001). 8
Poor health seeking behaviour and practices is an important contributing factor to the high rates of child mortality. Appropriate care seeking for emergency situations remains unsatisfactorily low at five percent for obstetric emergencies,17 percent for pneumonia with discriminatory practices towards women and children contributing to the poor situation. Additionally, contacts with target groups are not used to its optimum. For example, TT coverage of pregnant women is around 80 percent, yet ANC coverage is as low as 48 percent (HNPSP). Health programmes for poor people, in particular for poor women, are extremely limited. Various studies, surveys and interviews indicate that there is serious dissatisfaction over the quality of health services available to women at Upazila and District Hospitals. Besides the public hospitals, Upazila Health complexes and Union Health Centres are the only available facilities for the treatment of rural people. Wide spread absence of doctors, the absence of female doctors, lack of sensitivity towards female patients, distance and poor access, rising hidden costs and little value attached to female health issues, discourage women to avail health services for themselves and their daughters. These differences have been found to be very acute as presented in a study by Rani and Lule (2004) conducted among 15-19 year old girls in Bangladesh and eleven other countries. This study shows that among the richest quintile of the population,72 percent of the girls are married by age 18 while for the poorest quintile it is 94.1 percent. Among women aged 20-24, those who had a birth by age 18,the richest have a figure of 43.7 percent as opposed to 73.9 percent among the poorest, indicating that the poor are married early and by age 18 almost 74 percent have had at least one childbirth. Among women aged 15 19 those who used a modern contraceptive ranged from only 23 percent among the 9
poorest quintile to 41.3 percent among the richest quintile. Only 3.4 percent poor adolescent girls used a trained attendant during their last delivery, while the figures are 36.8 percent among the richest quintile, on the average percentage was only 10.8 New health problems are also emerging as imminent threats due to the feminisation of poverty and discrimination against women and girls. Due to socio-cultural beliefs and practice there is reluctance among women and adolescent girls to consult health professionals, having serious consequences on their sexual and reproductive health including vulnerability to HIV/AIDS. Government statistics indicate that there are 188 positive cases (81 percent male and 19 percent female) (UNAIDS) at present in Bangladesh. Due to lack of agency of women it is feared that women may constitute the bulk of newly infected persons, whose only `high risk behaviour is being married (UNAIDS). One study has found that the use of condoms is only 2 percent among the brothel based CSWs which compounds the matter further. A recent study has indicated that 70 percent of women and half of men have never heard of AIDS in Bangladesh. Rani and Lule (2004) finds that among girls aged 15-19, one third women in the richest quintile are aware of AIDS, among the poorest quintile the percentage is simply nil. Among the population with average wealth, the level of knowledge is around 7 percent. This study summarizes that` Data from all countries reveal widespread socio-economic differentials in reproductive health outcomes and service utilization. Young women in the poorest households were less likely than those in the richest households to be enrolled in school, to use modern methods of contraception, to have given birth with a trained attendant present and to know at least one way to avoid contracting HIV; they were more likely to have gotten married by age 18 and to have had a child by that age (ibid). 10
URBAN RURAL DIFFERENCES Fig 3 Health status and behaviour by area of residence Rural Urban % 80 70 60 50 40 30 20 10 0 14.9 15.5 21.7 18.6 79.2 77.5 Chronic illness Recent illness Sick persons received treatment 71.6 62.5 1-4 year children immunized One study conducted on the nationally representative sample of the household income and expenditure survey of 2000,conducted by BIDS, finds wide variation among rural and urban sections of the population. They find that the health status of the population is poorer in rural compared to urban areas in terms of prevalence of both chronic and recent illness. The difference is more pronounced for recent illnesses. 11
Health seeking behaviour also varies by area of residence. Rural children are less likely to be immunized and the sick living in rural areas are less likely to receive treatment when ill. The urban gap is more evident for immunization. And for all indicators, both health status and health seeking behaviors, there is a visible gender difference in favor of males, but it is interesting to note that the gender difference in health status is less for Fig. 4. Gender differences in health seeking behavior by area of residence Male Female 80 60 79.9 78.6 78 77.1 71.8 71.3 64.2 60.7 % 40 20 0 Urban (Sick persons received treatment) Rural (Sick persons received treatment) Urban (1-4 year children immunized) Rural (1-4 year children immunized) rural than urban areas. There is reason to suspect that these differences may have been caused by more frequent visits to the health centers by the females. The health centers in the rural areas are mainly offering reproductive health services which may be the cause for better health status among women in the rural areas. In the same breath, lack of appropriate services for general health of the population may have rendered the male population in a worse situation compared to their urban counterparts. The probability of suffering a chronic or recent illness has an inverse relationship with head s wife s education (Fig. 5). Also, the probability of the sick receiving treatment 12
when ill and children 1-4 years of age being immunized is more with the extent of head s wife s schooling level (Fig. 5). The difference, however, is visible when SSC or higher schooling level is attained. Another study also found mother s education to be related with her child s nutritional status (Mahmud,2004).As before, the gender difference remains irrespective of head s wife s schooling level. The only exception was in the group of 5-9 years schooling where more likely to receive treatment than their male counterparts. Figure 5. Health status and behavior by head' wife's education 0-4 years 5-9 years SSC+ % 100 80 60 40 20 0 15.6 15 14.1 21.2 21 19.3 77.3 77.3 Chronic illness Recent illness Sick persons received treatment 81.5 82.1 61.3 71 1-4 year children immunized Total Fertility Rates Many studies have addressed the issue of the TFR and its differentials in Bangladesh. While substantial gains in increasing CPR and lowering TFR has been noted by demographers, sufficient evidence has been found in the literature to indicate that both urban rural differentials, regional differences and differences among the socio-economic quintiles exemplify lack of equity among these social classes. The most current figures show that on the average rural women give birth to 0.7 more children compared to the 13
urban women (UNFPA,2005).Table 1 produces some figures of the current demographic surveys classified according to the wealth quintile Table 1: TFR by socio-economic quintiles Socioeconomic group BDHS 93-94 BDHS 96-97 BDHS 99-2000 BNMS 2001 1(Poor) 3.8 3.6 4.1 4.2 2 3.6 3.7 3.5 3.7 3 3.6 3.4 3.3 3.2 4 3.3 3.3 2.9 2.9 5(Rich) 2.9 2.5 2.5 2.4 Shows that the richest class has declined from a TFR of 2.9 to 2.4 in a period of five years, while the poorest class has actually risen from a TFR of 3.8 to 4.2 in the same period. In some districts of Bangladesh, particularly in the western side of the country, replacement level fertility, or close to it has already been achieved. While such districts may be more economically advanced and social progressive compared to most other districts, their population is nevertheless poor and largely illiterate (Islam et al.,2005). CONCLUSION Although, in the context of Bangladesh, common stratifiers have been used by some researchers to identify certain social groups which need more attention in health and health seeking behaviour, a clear causal pathway has not yet been identified. For all purposes, women s education has had clear association with increased levels of contraceptive use, reduced desire for more children, child s immunization, health seeking 14
behaviour. Urban rural differences in various health indicators such as infant mortality, fertility, child mortality, immunization rates, sanitation, etc has been presented in details; yet in some cases rural figures have paradoxically been better than the urban ones (Mahmud,2004). Research has also found that within the urban poor the child survival rates are different for native urban poor versus urban migrants (Islam and Azad,2005),indicating that many more stratifiers may be needed to locate the exact population which is most vulnerable. A comparison between the rural population and poor urban migrants may yet again render results in a startling manner. Similarly, the socioeconomic indicators which have been chosen for the BIDS study may not tally with the data presented in Table 1. Several researchers have pointed out at the need for utilizing various socioeconomic indices to stratify the data and the ensuing debate that follows from the use of various techniques has not yet been tested for the Bangladeshi context, neither has there been a fruitful analysis which will point out to the inequitable groups. For this, Bangladesh needs to work further in disaggregating national data into many more subgroups and analysing the data more extensively to obtain more insight into this complex issue. 15
REFERENCES Braveman,P. (2003) Monitoring Equity in Health and healthcare: A conceptual Framework in Journal of Health Population and Nutrition,Vol 21 No.3 ICDDR,B Bangladesh Equity dialogue, Vol 1 No 3 and 4 `Socioeconomic inequities in childhood malnutrition in Bangladesh. Islam,M (2005) `Possible Reasons for the Apparent Stagnation of the TFR paper presented in south South Centre Workshop, Dhaka, Bangladesh, April 2005. Islam,M and Azad,A.K. (2005) `Rural urban migration,poverty and child survival in urban Bangladesh accepted for presentation in the annual conference of IUSSP. Khan,S.; Bhatia,K.; Khan, M. I. And Pervin,S. (2004) Gender Thematic Review UNFPA,Dhaka,Bangladesh Mahmud,S. (2004)`Inequalities in health: Evidence from the Bangladesh household income and expenditure survey 2000 Bangladesh Institute of Development Studies (BIDS). Rani,M. and Lule,E. (2004) `Exploring the socio-economic Dimension of Adolescent Reproductive Health: A multicountry analysis in International Family Planning Perspectives. Vol 30 No. 3 16
Razzaque, A; Streatfield, P.K.; Ahsan, K.Z. (2004) `Health intervention and health equity: Evidence from Matlab, Bangladesh in Equity Dialogue, Vol 2 No. 2 17