ISSN (Print) - 0012-9976 | ISSN (Online) - 2349-8846

A+| A| A-

Untreated Morbidity and Demand for Healthcare in India: An Analysis of National Sample Survey Data

This paper studies the problem of poor health outcomes in India from the demand side, and using the unit level data from the 60th round of the National Sample Survey analyses the determinants of not accessing medical care. This analysis is confined to persons who have reported being ill within 15 days of the survey but have not sought either public or private professional medical services. There are systematic variations in accessing healthcare between urban and rural areas, as well as between males and females in each sector. While in the rural areas, the demand for healthcare increases significantly with the education level of the head of the household, in the urban areas the evidence is mixed. Richer economic sections constitute a larger proportion of sick persons who do not access medical care, especially in urban areas. Paradoxically, among poor households, which cite financial reasons for not accessing healthcare, women are less likely to be discriminated in rural than in urban areas.

SPECIAL ARTICLEEconomic & Political Weekly EPW november 15, 200871Untreated Morbidity and Demand for Healthcare in India: An Analysis of National Sample Survey DataAnit N Mukherjee, Krishanu KarmakarThis paper studies the problem of poor health outcomes in India from the demand side, and using the unit level data from the 60th round of the National Sample Survey analyses the determinants of not accessing medical care. This analysis is confined to persons who have reported being ill within 15 days of the survey but have not sought either public or private professional medical services. There are systematic variations in accessing healthcare between urban and rural areas, as well as between males and females in each sector. While in the rural areas, the demand for healthcare increases significantly with the education level of the head of the household, in the urban areas the evidence is mixed. Richer economic sections constitute a largerproportion of sick persons who do not access medical care, especially in urban areas. Paradoxically, among poor households, which cite financial reasons for not accessing healthcare, women are less likely to be discriminated in rural than in urban areas.The linkage between health, nutrition and economic devel-opment has been extensively discussed in literature (Strauss and Thomas 1998). The construction of the human development index (HDI) includes life expectancy – the broadest measure of health of the population in a country. Developed nations without exception have low maternal and infant mortality, as well as lower rates of malnutrition.1 IntroductionThe current health scenario in India is often described as “dismal” or “disturbing” (Bose 2008).1 Except a few states like Kerala, Goa and Tamil Nadu which have done relatively well, the situation in most parts of the country is a cause for worry. Levels of infant and neo-natal mortality, child malnutrition, female anaemia, non-institutional delivery, etc, are higher in some states in India than countries of sub-Saharan Africa. Going by present trends, India is in danger of missing the health targets set by the Millen-nium Development Goals (MDGs).One major reason put forward for this low level of achievement of health in India is the systematic lack of investment by the government, which adversely affects the poor. Public expendi-ture on health stands at less than 1% of gross domestic product (GDP), with state governments sharing most of the burden. In the light of economic reforms in the 1990s, the squeeze on public expenditure in health has been aggravated further especially at the state level (Mooij and Dev 2002). There have been a series of studies documenting the precari-ous situation vis-à-vis the provision of public health facilities in India. Most of them focus on access and quality issues, which deter people from utilising government health services. Recent papers have also investigated the links between non-utilisation and administrative factors such as absenteeism among the health staff in the rural areas, as well as the presence of alternative informal sources of medical care (Banerjee, Deaton and Duflo 2004). They also find instances of untrained and uninformed quacks being frequently consulted in the rural areas. Even in the fee-paying urban private sector, one particular study reports a lack of knowledge among medical practitioners in Delhi for common symptoms of diseases such as tuberculosis (Das and Hammer 2007).However, the demand for healthcare has received relatively little attention, particularly because of the non-availability of representative household-level datasets. The National Sample Survey (NSS) data is the most suitable for this kind of analysis. Over the last two decades, there have been only three rounds Anit N Mukherjee ( and Krishanu Karmakar (kkarmakar1@student. are both at the National Institute of Public Finance and Policy, New Delhi.
Yes Medical Advice Sought? Went to Government Service? Yes No Why? Yes No Whom Consulted? Any Other Measure Taken for Recovery? No Why?
SPECIAL ARTICLEEconomic & Political Weekly EPW november 15, 200873(ii) the major diseases for which persons reporting illness did not seek medical advice; and (iii) the proportion of reported cases of illnesses that were left untreated arranged in descending order of magnitude. Fever of unknown origin is the most frequently reported illness in the sample. It is also the most frequently neglected, along with known ailments, orthopaedic and respiratory conditions. It is interesting to note that the top four reported illnesses and most frequently neglected diseases are the same albeit with a slightly different ranking. However, in terms of the proportion of reported cases for a particular disease for which medical attention is not sought, the ranking varies considerably. The most neglected ailments are also those that occur in old age, indicating that these may be accepted as a normal condition by the person. Therefore, data in Table 1 points to a decrease in the demand for healthcare as the age profile increases. This will be verified later in Section 3.The reason for not seeking treatment for illness spells across all the three health rounds is given in Table 2. It is interesting to note that over the last two decades, the proportion of rural respondents citing lack of health infrastructure as a reason for not accessing medical care has gone up significantly from 3 to 12%. There has been a very dramatic decline in the proportion of people stating illness not serious as the reason for not treating morbidities. This is true both for urban as well as for rural areas. What this signifies is that there is a better awareness of the medical condition, but demand for healthcare is constrained due to financial and “other” reasons. The last category is interesting in itself and merits more atten-tion. It may reveal an increasing trend towards getting treatment informally, either from family members, or from the untrained and unregistered informal sector. Overall, more than one-third of the cases of illness are not considered serious enough to merit a visit to a medical professional. Therefore, apart from supply and economic constraints, increase in the demand for healthcare would depend on the individual’s perception of their medical condition.The gender and sector disaggregation is presented in Table 3. The proportion of male and female reporting untreated morbi-dity is nearly equal in rural, while there is a large difference in the urban areas. Rate of untreated morbidity in urban females is about 20% higher than the reported untreated cases for males. Also noteworthy is the fact that among the reasons cited for not accessing healthcare, there is very little difference among males and females in the rural areas. However, for the urban sample, there is a 3 to 4% gender gap for respondents citing finan-cial reasons and self-assessment of their illness. This may point to the persistence of gender bias in the utilisation of healthcare as already noted in studies using the 52nd round data (Sen, Iyer and George 2002).As we shall see subsequently, this gender disparity among the two sectors persists even if we stratify the data by demographic characteristics, income class or education levels. One important lesson from a policy perspective is that urbanisation may aggra-vate, rather than mitigate, the gender disparity in health seeking behaviour in future.3 DemographicCharacteristicsDemand for healthcare has been shown to have a non-linear relationship with demographic factors – principally age (Musgrove 2004). Newborn babies and children need greater care, hence the emphasis on neo-natal and child health in policy formulation. Lower infant mortality is positively correlated with better standards of healthcare and is also an indicator of the general level of development of a country or region. The example of Kerala has often been cited – high standards of healthcare has reduced infant mortality and malnutrition to levels seen in advanced countries. On the other hand, Uttar Pradesh has infant mortality and malnutrition rates comparable to the least developed nations of the world (Bose 2008).The demand for healthcare reduces during the productive years of adolescence and middle age, and then rises again when a person gets old. In the years between 15 and 45, the perception of the gravity of the illness is also much lower. Individuals making rational decisions about their health status may sometimes discount the long-run cost and prefer not spending time and resources in accessing healthcare during the productive age. This can very often result in increasing the probability of developing serious ailment in old age due to untreated morbidity. What it also means is that in case where the public provision of health-care is deficient, untreated morbidity may actually raise the economic dependence of the old on the younger generation. In the context of declining fertility rates, demographic composi-tionthat is now favourable for India may in the long run turn out to be a significant challenge for health policy, as is evident in ageing societies such as Japan, western Europe, US and even China. Our analysis of the 60th round data supports the points raised above. We disaggregate the data on untreated morbidity by age-groups and by gender – the rationale for dividing the Table 2: Distribution of Untreated Ailments by Reason of No TreatmentReason for Not Seeking Medical Treatment Rural Urban 42nd 52nd 60th 42nd 52nd60thNo medical facility 3 9 12 0 1 1Lack of faith 2 4 3 2 5 2Long waiting 0 1 1 1 1 2Financial problem 15 24 28 10 21 20Ailment not serious 75 52 32 81 60 50Others 5102461225Total 100100 100 100100100Estimates for Others include the cases where the reason is not reported.For 42nd and 52nd rounds estimates refer to untreated persons.Source: Table 4.9 ofNSSO 52nd round Report and Statement 16 ofNSSO 60th round Report.Table 3: Sector and Gender Distribution of Untreated Cases RuralUrbanReason for Not Seeking Medical Treatment MaleFemaleMaleFemaleNo medical facility 6.00 5.52 0.41 0.79Lack of faith 1.67 1.36 0.76 1.13Long waiting 0.51 0.36 0.29 1.35Financial problem 13.15 15.12 8.07 12.47Ailment not serious 14.86 17.38 24.68 25.60Others 11.8412.2411.5412.91Total 48.02 51.98 45.75 54.25Source: NSSO 60th round data.
SPECIAL ARTICLEEconomic & Political Weekly EPW november 15, 200875presumably their son. It is therefore unclear whether prefer-ence for male children within the household is reflected in health outcomes for the elderly.4 EducationLevelThe connection between education level and health status is well established. In the case of Kerala, for example, studies have pointed to the fact that there is almost universal school educa-tion, and also a much higher standard of health compared to other states in the country – especially in terms of maternal and child mortality. Paradoxically, the NSS report also indicates that the proportion reporting ailment is also the highest in India. This may be due to causal linkages between better education stand-ards and greater health-seeking behaviour. From another standpoint, number of years of education is one of the determinants of income level of an individual. We would therefore expect to find that for persons who are either illiterate or with a few initial years of school education, financial factors as well as their self-assessed health condition are the primary reasons for not accessing organised medical service. If access to health centres is difficult, the opportunity cost of transport and mandays lost can also become critical. This will disproportion-ately affect persons with lower rather than higher education – and consequently income – levels.To describe the data on the demand for healthcare by varying levels of education, we first stratify the data using the level of schooling of the head of the household as a proxy for the general standard of education of the household. If the household head is over 60, there is a higher probability that he/she would be less educated than the next generation. In case of a representa-tive household of middle-aged parents and young children, the level of education of the head of the household would be better correlated to the demand for health services for the entire family. For the purposes of our analysis, we do not separate out the two cases mentioned here, and report the results using the full sample of those who reported ill but did not seek medical advice.Table 7 describes the data for the whole sample, while Table 8 presents the picture in the rural and urban areas separately. We divide the sample into four classes as far as the education of the household head is concerned – illiterate, up to primary, up to secondary, and higher secondary and above. As expected, the share of untreated morbidity for those cases where the head of household is illiterate is over 45% which drops progressively to below 7% in the highest education level. Therefore, overall, there is a distinct inverse relationship as far ashouseholdhead’seducation level and demand for healthcare is concerned. The data in Table 7 also confirms the hypothesis that having lower levels of education in a household increases the chances of financial difficulty in accessing healthcare. Financial reason is cited by 16.3% of the household heads who are illiterates. On the contrary, the major reason for not accessing healthcare for other education groups is “ailment not considered serious”. What is clear is that the cost of medical care restricts access for those with low education levels. Furthermore, health infrastructure is also more of a hindrance in accessing care for persons with lower education levels. The linkages between supply and demand sides, therefore, need to be looked at more carefully in multiple contexts of human development.The overall picture masks significant differences between the rural and urban areas as far as education level is concerned, as seen from Table 8. For rural areas, the broad conclusion of Table 7 still holds, except for the fact that the “return to education” is higher. That is, the decline in the rate of untreated morbidity decreases faster across education levels in the rural area compared to the general picture. From a policy perspective, it points to the importance of adult education, which may have a first-round impact on reducing untreated morbidity. In the long term, demand for healthcare would improve alongside improve-ment in the general level of education.For the urban area, the picture is very different. The propor-tion of untreated morbidity is the highest for those instances where the head of household has at least secondary level educa-tion. It is nearly 6% higher than those where the head of house-hold is illiterate. This result is counter-intuitive, but may be explained by the fact that most households with low levels of education are likely to be daily-wage manual labourers, who cannot afford not to treat temporary ailments such as fever, injuries or respiratory problems. Given the greater availability of private healthcare facilities in urban areas and higher wage rates, the proportion of illiterates who reported not accessing health-care due to infrastructure or financial reason is much lower than rural areas. In fact, 75% of untreated morbidities are either due to the individual perception of their illness not being serious enough or “other” reasons given by the respondent. As we shall show later, most of those who gave “other” reasons have taken medical Table 7: Reasons for Not Accessing Treatment by Education Level of Household Head (in % )Reason for Not Seeking Education Level of the Head of the FamilyMedical Treatment Illiterate Primary SecondaryHigherSec TotalNo medical facility 4.70 2.61 1.79 0.41 9.52Lack of faith 1.31 0.96 0.49 0.05 2.81Long waiting 0.19 0.26 0.49 0.08 1.02Financial reasons 16.35 6.76 3.04 0.62 26.77Ailment not serious 12.88 10.68 8.38 3.79 35.73Others 10.226.385.641.9224.16Total 45.6527.6519.836.8710.00Higher Sec: Higher Secondary.Source: NSS 60th round.Table 8: Demand for Outpatient Care Classified by Education Level of Head of Household(in %)Reason for Education Level of the Head of the FamilyNot Seeking Rural UrbanMedical Treatment Illiterate Primary Secondary HigherTotal IlliteratePrimary SecondaryHigher Total Sec SecNo medical facility 5.70 3.21 2.18 0.42 11.52 0.53 0.12 0.18 0.35 1.19Lack of faith 1.40 1.07 0.52 0.03 3.03 0.90 0.49 0.36 0.13 1.88Long waiting 0.14 0.27 0.38 0.09 0.87 0.41 0.21 0.96 0.06 1.64Financial reasons 17.75 7.04 2.83 0.64 28.26 10.50 5.59 3.93 0.52 20.54Ailment not serious 13.35 10.78 6.40 1.70 32.24 10.93 10.26 16.6 12.49 50.28Others 11.35 6.645.081.0124.09 5.515.297.985.67 24.45Total 49.70 29.01 17.39 3.90100.0028.79 21.96 30.02 19.23 100.00Higher Sec: Higher Secondary.Source: 60th round NSS.
SPECIAL ARTICLEnovember 15, 2008 EPW Economic & Political Weekly76advice from their friends, family members, or medical shop attendants. This implies that compared to rural areas, healthcare demand being met by informal sources such as medical shop attendant, friends and family, etc, in urban areas may actually be higher.5 EconomicGroupsThe proportion of untreated morbidity are expected to vary across various economic groups as well. However, the direc-tion of causality is more difficult to hypothesise. Higher income groups may have a lesser degree of financial constraint, but that may not necessarily translate into higher demand forhealthcare. The interplay between the various socio- economic factors – two of which have already been examined – may determine the proportion of reported illnesses that go untreated.The NSS questionnaire enables us to stratify the sample by monthlyexpenditure level of the household. We divide the whole sample into five expenditure groups on the basis of increasing monthly household expenditure and tabulate the results for untreated morbidity for each quintile. As per the expenditure group classification, Table 9 provides the summary for the entire sample which reported ailing within the survey recall period of 15 days. We find that for rural and urban areas combined, there is little variation in the distribution of untreated cases across expenditure quintiles. However, two points emerge from the table: (i) in the poorer quintiles, financial reasons dominate, and (ii) in the higher quintiles, the perceived health condition is the major reason for not seeking care. Interestingly, access to health infrastructure as a constraint to seeking care in the fourth quintile is the highest among all expenditure groups. Table 10(a) and 10(b) present the data disaggregated by rural and urban sectors. Within each expenditure quintile, we present the share of the reasons for not accessing healthcare as a proportion of total untreated morbidity. The contrasts are very clear: (i) the proportion of untreated morbidity due to financial reason within each expenditure quintile shows a secular decline from poorest to richest expenditure levels both for rural, but not for urban areas; (ii) access to healthcare facility is not a constraint in urban areas, but is still reported in the rural sector across all expenditure quintiles; (iii) from the second quintile onward, more than half the share of untreated morbidity is due to self-assessment of the disease as well as “other” reasons; (iv) even for the richest quintiles in the rural areas, issues of access, quality and financial constraints cannot be disregarded; and (v) among the poorest urban quintile, the proportion of reported untreated morbidity due to financial reason is nearly 8% more than the poorest rural quintile.Supplementary questions were asked whether those reporting no treatment for ailments took any other measure, for example, consulting friends and family, a medical shop, etc. In Table 11, we tabulate the data by expenditure groupwhere the respondent gave the reason for no treatment as “other”. For the highest ex-penditure quintile in the urban area, this reason itself consti-tutes half of the cases of untreated morbidity. The distribution is more even in the case of rural areas. To summarise, we find that economic condition of the house-holds can partly explain the non-treatment of morbidity in terms of the reasons asked during the survey. A relaxation of the financial constraint, i e, moving from lower to higher expenditure groups, is accompanied by a greater proportion of untreated morbidity being due to the individual’s own health perception, as well as due to access to informal sources of healthcare. This contrast is greater for urban as compared to rural areas. In the latter, physical access to health facilities and the quality of treatment still remain important – something which needs urgent attention at the policy level.Table 9: Distribution of Untreated Illness (according to Family Expenditure Category and Stated Reason for Not Seeking Medical Advice) (% in sub-sample) Reason for Not Seeking Treatment Expenditure No Medical Lack of Long Wait at Financial Ailment Not Others Total Quintile Facility Nearby Faith the FacilityReasons Serious Very poor 2.17 0.49 0.15 7.41 5.04 3.96 19.22Poor 1.770.810.077.256.514.4120.81Middle 1.750.570.216.006.993.5419.06Richer 2.470.470. 1.360.470.392.108.986.5119.80Total 9.52 2.811.0226.77 35.73 24.16 100.00Source: NSS 60th round.Table 10(a): Rural India Reason for Not Seeking Medical TreatmentExpenditure No Medical Lack of Long Financial Ailment Not Others Total Quintile FacilityFaith Waiting Reasons Serious Poorest 2.610.560.178.135.484.5321.49 (12.16)* (2.61)(0.81)(37.84)(25.51) (21.07) (100.00)Poor 2.180.970.088.266.744.8923.12 (9.42) (4.20) (0.35)(35.74) (29.15) (21.14) (100.00)Middle 2.150.580.156.267.433.8720.44 (10.50)(2.86)(0.74)(30.61)(36.37)(18.92)(100.00)Richer 3.070.510.123.767.235.7020.39 (15.04) (2.51)(0.60) (18.45)(35.46) (27.94)(100.00)Richest 1.510.400.341.855.355.1114.56 (10.38) (2.74)(2.35) (12.71) (36.74) (35.08) (100.00)Total 11.52 3.03 0.87 28.26 32.24 24.09 100.00* The values within brackets are proportions within each expenditure quintiles. Source: NSS 60th round.Table 10(b): Urban India Reason for Not Seeking Medical TreatmentExpenditure No Medical Lack of Long Financial Ailment Not Others Total Quintile FacilityFaith Waiting Reasons Serious Poorest 0.300.210.074.393.201.579.74 (3.06)* (2.17)(0.67)(45.11) (32.89) (16.10) (100.00)Poor 0.07 0.12 0.03 3.02 5.55 2.41 11.19 (0.59)(1.04)(0.23)(26.99) (49.58) (21.58) (100.00)Middle 0.100.500.484.945.152.1613.33 (0.79) (3.75) (3.58)(37.08) (38.60) (16.21) (100.00)Richer 0.000.300.505.0512.255.9724.06 (0.00)(1.23)(2.08)(20.97)(50.91)(24.81)(100.00)Richest 0.730.760.583.1424.1312.3441.68 (1.74)(1.82)(1.38)(7.53)(57.91)(29.61)(100.00)Total 1.191.881.6420.5450.2824.45100.00* The values within brackets are proportions within each expenditure quintiles. Source: NSS 60th round.Table 11: Distribution of Cases of Untreated Morbidity due to “Other Reasons”Expenditure Quintile Rural Urban TotalVery poor 18.80 6.41 16.37Poor 20.299.8718.25Middle 16.068.8414.64Richer 23.6524.4123.80Richest 21.20 50.47 26.93Total 100.00100.00100.00Source: NSS 60th round.
SPECIAL ARTICLEEconomic & Political Weekly EPW november 15, 2008776 ConclusionAnalysis of unit-level data from theNSS 60th round throws up interesting set of issues related to the demand for healthcare in India, and the challenge faced in formulating public policy towards the health sector. In this paper, we investigated the health-seeking behaviour of the respondents of the survey, and explored three avenues through which health and human devel-opment outcomes may be related – demographic characteris-tics, education level of the head of the household, and expendi-ture groups. We find that theNSS data points to large differ-ences in the demand for healthcare when we disaggregate according to gender and geographical location. The intra-family relationship as well as the level of education of the head of household exert considerable influence on health-seeking behaviour. As expected, for lower expenditure groups, financial reasons play an important role in the lack of demand for health-care. Interestingly,theproportion of cases within the lowest quintile citing lack of resources is higher in urban than in rural areas.The paper has several implications for the future direction of health policy in India. It is clear that health-seeking behaviour in rural and urban areas is different across demographic and socio-economic groups. The two sectors, therefore, need to have differ-ent strategies for improving health conditions. Infrastructure deficit still persists in rural areas affecting all economic classes. Adequate resources need to be allocated for expanding the network of public health facilities in rural areas. Financial problems continue to prove a major constraint for lower expendi-ture quintiles both in the urban and rural areas. Reducing the cost of access to the formal healthcare system through better quality standards in public facilities and spreading the cost through social security programmes need to be looked at carefully. Age and gender increase the inequity in access to healthcare, which can only be addressed through mass campaigns and outreach by dedicated community health workers. A multi-pronged strategy is needed to address the lack of demand for healthcare that will, in the long run, lead to a betterment of the overall health status of the population.Note1 See Bose (2008) for a discussion on child and maternal health in the light of the latest National Family Health Survey (NFHS-3) data. ReferencesBanerjee, A, A Deaton and E Duflo (2004): “Wealth, Health, and Health Services in Rural Rajasthan”, American Economic Review94, No 2, pp 326-30.Bose, A (2008): “India’s Disturbing Health Card”, Economic & Political Weekly, Vol XLII, No 50, pp 10-13. Das, J and J Hammer (2007): “Money for Nothing: The Dire Straits of Medical Practice in Delhi, India”,Journal of Development Economics83, No 1, pp 1-36. Mooij, J and M Dev (2002): “Social Sector Priorities: An Analysis of Budgets and Expenditures in India in the 1990s”, Development Policy Review 22, No 1, pp 97-120.Musgrove, P (2004): Health Economics in Development (Washington DC: World Bank).Sen, G, A Iyer and A George (2002): “Structural Reforms and Health Equity: A Comparison of NSS Surveys, 1986-87 and 1995-96”, Economic & Political Weekly, Vol 37, 6 April, pp 1342-52.Strauss, J and D Thomas (1998): “Health, Nutrition and Economic Development”, Journal of Economic Literature, 36, (2), June, pp 766-817.SAMEEKSHA TRUST BOOKSInclusive GrowthK N Raj on Economic DevelopmentEssays from The Economic Weekly and Economic & Political WeeklyEdited by ASHOKA MODYThe essays in the book reflect Professor K N Raj’s abiding interest in economic growth as a fundamental mechanism for lifting the poor and disadvantaged out of poverty. He has also been concerned that the political bargaining process may end up undermining growth and not provide support to those who were excluded from access to economic opportunities. These essays, many of them classics and all published in Economic Weekly and Economic & Political Weekly, are drawn together in this volume both for their commentary on the last half century of economic development and for their contemporary relevance for understanding the political economy of development in India and elsewhere.Pp viii + 338 ISBN 81-250-3045-X 2006 Rs 350Available fromOrient Blackswan LtdMumbai Chennai New Delhi Kolkata Bangalore Bhubaneshwar Ernakulam Guwahati Jaipur LucknowPatna Chandigarh Hyderabad Contact:

Dear Reader,

To continue reading, become a subscriber.

Explore our attractive subscription offers.

Click here

Back to Top