The National Sample Survey: A Preliminary Look at Health Practices in India

This blog post is the first in a series on the Indian National Sample Survey. This post was coauthored by Arun Nair. 

The Indian National Sample Survey is a multisubject, multipurspose, and multistage household survey. The survey has been conducted every year since 1950. The current topics of the survey include employment, unemployment, and consumer expenditure; unorganized enterprises in nonagricultural sectors; a wide range of healthcare indicators, such as population, births, deaths, disability, morbidity, fertility, maternity and child care, and family planning; landholdings and livestock enterprises; and debt, investment, and capital formation.[ref]Katyal, A., Singh, P. & Samarth, A., 2013. Using the Indian National Sample Survey data in Public Health Research. National Medical Journal of India, 26(5), pp. 291-294.[/ref]

The healthcare portion of the National Sample Survey is conducted every ten years and was last conducted in 2014. In this series of blogs, we look at the findings of this healthcare round, compare these findings to the earlier healthcare round of 2004, for both financially better off and financially disadvantaged states in the India. This series of blogs will look at preliminary findings and offer a comparative analysis in terms of out of pocket expenditures, access to care, and which episodes of illness cause catastrophic health expenditures. We will look at changes for both the overall group and for various subgroups organized by religion, rural vs. urban, social caste, and education levels.

In this first blog post of the series, we discuss the summary findings for the states where ACCESS Health has experience working on health seeking behavior, for both primary care and hospital care. Our goal is to whet your appetite for the more detailed analysis in the subsequent blogs.

The Andhra Pradesh survey included 2,448 households. Madhya Pradesh included 3,613. Rajasthan included 2,912 households. The Uttar Pradesh survey included 7,921 households. Data from these state specific surveys were compared to data from 65,932 households nationally. Each of these states has different characteristics. Andhra Pradesh has a long coastline. Rajasthan is essentially a desert. Uttar Pradesh is a part of the Gangetic plains and Himalayas. Madhya Pradesh has dense forests and scantily inhabited areas.

In terms of wealth, Uttar Pradesh has the highest gross state domestic product. Madhya Pradesh has the lowest. Rajasthan has the highest monthly per capita expenditure. Again, Madhya Pradesh has the lowest.

Basic Findings

One interesting aspect of the National Sample Survey is that the survey records “self reported illness” (See Charts 1 and 2). For this survey, self reported illness is defined as one’s own perception of health in the last fifteen days. In Madhya Pradesh and Rajasthan, the highest contributors to child death, the reporting of child illness (the age group 0-14) is the lowest in these two states. Andhra Pradesh, on the other hand, has higher reporting of ailments across all age groups among these four states. Higher literacy leading to higher awareness, as well as better access to facilities, may explain the high rate of reported illnesses. In states that contribute less to child death, self reported illness per thousand is higher. For Tamil Nadu, this figure was 121 per one thousand. In Kerala, 221 per one thousand.

Based on our own studies in these states, we know that information asymmetry is high. People are unaware of the benefits that they are entitled to receive. Another issue could be low socioeconomic status, as determined by low gross state domestic product. This is further worsened by the low literacy rates, causing low disease awareness. Low socioeconomic status and lack of awareness together contribute to low reporting of diseases.

 

Another interesting aspect to note is the percentage of males and females in rural and urban areas seeking contemporary medical care, here defined as allopathic medicine, versus those seeking care in alternate forms of medicine. Alternate forms of medicine such as Ayurveda, Unani, and Siddha are popular in both rural and urban India. In rural Madhya Pradesh, the percentage of women not seeking care is higher than men. It is noteworthy that in Rajasthan and Uttar Pradesh, the percentages for women seeking care compared to men are equal and lower, respectively. These states have poorer economic indicators. Consistently, throughout these states, women prefer alternate forms of medicine more than men do. This pattern is most pronounced in Uttar Pradesh and Rajasthan. Even though men overall consistently prefer the allopathic system of medicine, in urban areas, the preference for alternate forms is higher than for rural men.

In terms of seeking outpatient care, Andhra Pradesh shows a higher preference for seeking care in private hospitals. The other three states show a preference for private doctors for outpatient care. This trend is consistent with most of the other states in India, including more developed states like Kerala and Karnataka. Findings in Tamil Nadu, however, are consistent with those in Andhra. Both Andhra Pradesh and Tamil Nadu have large scale health insurance programs. The insurance program in Andhra Pradesh, specifically, has been examined in published literature[ref]Reddy, S. & Mary, I., 2013. Aarogyasri Scheme in Andhra Pradesh, India: Some Critical Reflections. Social Change, 43(2), p. 245–261.[/ref] for encouraging people to seek care in more specialized facilities and hospitals and bypassing primary care and community healthcare centers, as shown in Chart 3.[ref]Selvaraj, S. & Karan, A., 2012. Why Publicly-Financed Health Insurance Schemes Are Ineffective in Providing Financial Risk Protection. Economic and Political Weekly, 17 March, 47(11), pp. 60-68.[/ref]

 

For hospital care, the National Sample Survey showed a clear preference for private hospitals. This trend was more pronounced among the males, both rural and urban. The exception being rural Madhya Pradesh where the survey data showed a greater preference for public hospitals.

Given that this preference for private hospitals is a trend across India – and a trend that has been extensively quoted in the literature – it is worth probing for any noteworthy exceptions. States in northeastern India, along with Himachal Pradesh, Chandigarh, Andaman and Nicobar Islands, Lakshadweep, and Jammu and Kashmir, are among the few states that showed a preference for public sector providers in both rural and urban areas. The availability of private hospitals is lower in many of these states, which may be a contributing factor. Himachal Pradesh is an exception. Himachal Pradesh has the best health indicators and high public healthcare spending, despite being a due to its mountainous terrain.[ref]http://www.lse.ac.uk/internationalDevelopment/pdf/WP/WP133.pdf[/ref] Public healthcare facilities in Himachal Pradesh are well functioning, and general social awareness and literacy is high throughout the state.

 

Further Research

The difference in reporting healthcare ailments as reported by the National Sample Survey dataset certainly warrants better awareness generation about general ailments, as well as about public health programs and entitlements. There is a disparity between the health seeking behavior of the better off states, especially those with large scale government sponsored health insurance program. The differences in health seeking behaviors may also be due to the presence of better private healthcare facilities in these states. Increased health seeking behavior in these states verifies, to a certain extent, the extensive studies of these health insurance programs that found that health insurance medicalizes and privatizes healthcare.

To substantiate these findings and comment on their statistical significance, we need to compare this data with earlier healthcare rounds of the National Sample Survey. Even though there are many differences in the datasets, many indicators remain the same.

Stay tuned for more blogs in this series as we begin to analyze the data in more detail. Click here to see the next post in the series.