Health Investment Priorities in India: A New Methodology-based Analysis of State Budgets

Definitions of Type of Care: Brief Review

One early definition of primary, secondary and tertiary health care classifies interventions to prevent a problem from occurring as primary, interventions at a stage before problem is manifested as secondary and remediation to reverse manifestations of problem as tertiary (Starfield, 1996).

UK’s National Institute for Health and Care Excellence (NICE) defines primary health care as health care which is delivered outside hospitals, but secondary care is to be provided in hospitals. Tertiary care is defined as care that requires complex treatments as stated in (NICE, 2018).

The American Academy of Family Physicians (AAFP) defines primary care as care that is provided by ‘physicians specifically trained for and skilled in comprehensive first contact and continuing care for persons with any undiagnosed sign, symptom, or health concern’. This type of care includes health promotion, disease prevention, health maintenance, counselling, patient education, diagnosis, and treatment of acute and chronic illnesses in a variety of settings (AAFP, 2018).

The first international declaration underlining the importance of primary health care is known as the Declaration of Alma-Ata, which was adopted in 1978. The emphasis on primary care was renewed with the World Health Report (WHO 2008) on primary health care, which advocated for a more balanced way of spending health resources and re-focusing on preventive and primary care as cited in the WHO report on primary health care. The idea that prevention and primary care can actually reduce expenditure on health spending for households is now widely accepted.

The other aspect—which is also related to the definitional issue—is that while levels of care is a clinical concept, government budgets that report annual allocations and expenditures comprise mostly uniform financial/accounting line items. Categorisation of health expenditure by levels of care would, therefore, require expertise in both clinical and financial concepts, as well as a thorough understanding of the health system. Researchers have thus largely stayed out of this categorisation issue due to the absence of any standard acceptable procedure of classification. Understandably, isolated attempts at classification could lead to very different results across studies, depending on the researchers’ understanding of the health systems and the resultant assumptions used for classification.

In India, the approach has been to define different kinds of care based on the level and type of facility at which the care is provided. The Five-Year Plans of the erstwhile Planning Commission in India defined primary health care infrastructure as those that provide the ‘first level of contact between the population and health care providers up to and including primary health care physicians and forms the common pathway for implementation of all the health and family welfare programmes in the country’ (Planning Commission, 1999). Only those services requiring specialised care were to be referred to secondary or tertiary care.

An important point in this context is whether definitions of levels of care should be based on the type/level of the facility, or on the nature of services offered. The possibility of hospitals providing primary and secondary care has been discussed in the literature, and it is recognised that tertiary care hospitals often end up giving primary and secondary care due to availability and accessibility issues at lower tiers of hospitals and facilities in many countries according to Jamison et al. (2006). In India, the district hospitals and also urban hospitals were to serve as secondary care facilities, but the definition of secondary care always remained somewhat imprecise in all the official documents. In fact, the line between secondary and tertiary care has not been very clearly articulated in most major documents of the government. The commission on macroeconomics and health cited in Ministry of Health and Family Welfare (MoHFW 2005) defined secondary care to mean sub-district hospitals for every 500,000 population and district hospitals at district level. It estimated the cost of an essential package with only primary and secondary care facilities included in it and visualised that much of the services for prevention and treatment of non-communicable disease such as vascular diseases, cancer and mental illnesses could be done in the secondary facilities.

The High Level Expert Group (HLEG) on Universal Health Coverage (UHC) also seemed to indicate that tertiary care facilities are those that require super specialty services (Planning Commission, 2011).

Spending on Type of Care in India: Recent Estimates

The National Health Accounts of India provides health care expenditures in India based on the System of Health Accounts 2011 which is a global standard framework for producing health accounts. It divides current health expenditure of the country into five categories: primary (P), secondary (S), tertiary (T), governance & supervision and ‘not classified elsewhere’ (NHSRC, 2016, 2017). Table 1 presents the shares for each of these for these years.

Slightly less than half of total government expenditure went to primary care, followed by secondary care, according to these estimates. There was not much change between the two years, 2013–2014 and 2014–2015. However, the NHA numbers are based only on current expenditures (capital expenditure not included).

OPEN IN VIEWER

Table 1.

National Health Accounts Estimates

Categories	2013–2014	2014–2015	2015–2016
Primary	45.5	45.1	45.1
Secondary	34.8	35.6	35.2
Tertiary	16.1	15.6	15.2
Governance and supervision	2.5	2.6	3.1
Not elsewhere classified	1.0	1.1	1.5

Source: National health accounts estimates for India, respective years.

A recent exercise in resource tracking for primary health care for Uttar Pradesh (UP) and Bihar shows that for both these states, the share of P was much higher than those reported in Table 1, while the shares of S and T were much lower. For Bihar, these were 69, 10 and 6 respectively and for UP the shares were 61, 17 and 3 respectively (Bhawalkar et al., 2016).

Spending on Type of Care: A New Methodology

Unlike the NHA classification, the aim here was to classify total expenditure (current + capital), not merely current expenditure, across six broad heads: primary (P), secondary (S), tertiary (T), education (E), social security (SS) and administration (A). SS is a notable addition to the standard classifiers, the rationale for its inclusion being the plethora of schemes—insurance, assurance, conditional cash transfers, etc.—that have been initiated by states.

A detailed analysis of the budget data for Bihar and UP—from the Empowered Action Group (EAG) of states—West Bengal and Himachal Pradesh—from among states that are doing fairly well on some of the health indicators—and Tamil Nadu—which is leading in almost all the health indicators, was undertaken.

The main document used was the Detailed Demand for Grants (DDG), which is presented to the state legislative assembly by the finance departments of each state. Each demand contains totals for ‘voted’ and ‘charged’ expenditure, the ‘revenue’ and the ‘capital’ expenditure and the gross total expenditure for which the demand is presented. Expenditure which under the provisions of the Constitution is subject to the vote of the Legislature is shown as ‘Voted’ and is different from expenditure which is ‘Charged’ (on the Consolidated Fund of India or of a State or Union Territory Government). It also presents estimates of expenditure under different units (levels) of classification, namely ‘major head’, ‘minor-head’, ‘sub-head’ and ‘detailed head’.

In this analysis, primary care was taken to be inclusive of prevention and promotion, and largely pertaining to the principle of first point of contact between patients and the health care system.

Secondary care included services requiring diagnostics and non-specialty hospitalisation, whereas tertiary care was all major hospitalisations including those requiring super speciality services.

Alternate systems of medicines were mainly classified under primary and sometimes under secondary, if hospitals were mentioned.

The need for a separate set of algorithm for classification can be better understood if one is familiar with the basic structure of the DDG. One such example from the state of Tamil Nadu state is given in Appendix A. For the state health departments, the DDG is useful for arriving at not only the total expenditure on health and family welfare but also on various pre-defined heads of expenditure. However, it is not very useful for further separation of expenditures across the six categories, especially P, S and T, which is relevant in the context of discussions on health policy priorities.

In the DDG, the budget items are divided in four rows across ‘major heads’, ‘sub-major heads’ and ‘minor heads’, with the title or functional description of the actual item of expenditure given out separately in another row labelled as sub-head or detailed head. Based on these four rows, it is possible to understand the exact nature of the item of expenditure in about half of the cases. But there remain cases that are not easily classifiable or are ambiguous and require further investigation and/or assumptions to separate these into logical categories. The exercise requires going through each line item of the budget to assign P, S, T, E, SS or A. For example, there are 688 line items in the West Bengal budget in a typical year.

For the straightforward cases the classification was done easily based on budget heads, but for the remaining ones, a set of algorithms were devised based essentially on knowledge about the nature of these services, sometimes facilitated by more disaggregated keyword searches based on stated line items. For administration, the classification given in the DDG was followed; in other words, wherever the reported classification mentioned ‘direction and administration’ as a minor head, those items were accepted under this category. For social security, the separation was not problematic since it was easy to separate out expenditures that pertained to health coverage schemes or programmes. For medical education, there were instances when expenditures of teaching colleges and hospitals were given together and had to be separated and apportioned to the relevant categories. The category medical education also included research. For the remaining items, the assumptions and algorithms adopted are given in Appendix B.

Some of the important methodological decisions are around expenditures incurred by general hospitals. These were separated into P, S and T based on shares derived from expenditures and volume of patient load based on a teaching hospital of national importance—All India Institute of Medical Sciences (AIIMS)—for which physical and financial details were available from their Annual Reports which are in the public domain. The estimates indicated that of the total expenditure incurred by (AIIMS), 61 per cent went for primary care, 14 per cent for secondary care and 25 per cent for tertiary care, respectively. This will be referred to as the ‘hospital ratio’ in the rest of the article, and has been used in cases where it was not possible to separate the expenditure of medical colleges and their attached hospitals.