Tracking SDG Targets on Health and Nutrition in India: Challenges and Opportunities for Streamlining Surveys

I.  Introduction

On 25 September 2015, the General Assembly of the United Nations (UN) adopted the Sustainable Development Agenda and secured commitments from member-nations to a “comprehensive, far-reaching and people-centred” set of universal and transformative development goals.^[1] India, being a signatory, has to devise a vision with actionable strategies and targets, keeping national realities and timelines in mind. In this regard, the Ministry of Statistics and Programme Implementation (MoSPI) has drafted a National Indicator Framework (NIF) that will lay the foundation for India’s SDG tracking.^[2] A complementary exercise of mapping of policies and programmes under various nodal ministries associated with each of the goals and targets was conducted earlier by NITI Aayog.^[3] The draft NIF called for inputs from stakeholders for thorough revision, effectively negotiating data availability within health and nutrition sectors. A robust NIF will help continuous tracking of the national targets to enable mid-course correction in policy.

Rolling out the NIF requires a robust data-collection framework that furnishes quality data on a regular and timely basis. This makes it imperative to scrutinise and steer all potential sources of data generation towards facilitating target-tracking, including the need for disaggregated data to meet the equity agenda of SDGs.^[4] Taking into consideration the fact that Civil Registration System (CRS) has scanty coverage barring few states,^[5] and Health Management Information Systems (HMIS) as well as surveillance systems largely exclude healthcare provisioning by the private sector,^[6] India will largely be depending upon population surveys for health and nutrition targets, in both short and medium terms.^[7] It is necessary to harmonise the data collected by these surveys, keeping in mind the requirements of the equity agenda of the SDGs. This will obviate the dependency on suboptimal sources of data.^[8] However, these surveys are commonly conducted and managed by various agencies under the aegis of different government ministries addressing various policy objectives. The fragmentation of data collection apparently results in suboptimal information sharing between ministries as well. At the 48^th session of the UN Statistical Commission on 15 March 2017, MoSPI made an unambiguous statement that “[the] Government of India (GoI) is not aware of any Food Insecurity Experience Scale (FIES) survey in India,”^[9] even as FIES is being canvassed as part of the Comprehensive National Nutrition Survey’ (CNNS) by the Ministry of Health and Family Welfare (MoHFW) and Unicef.^[10]

To begin with, India admittedly has the capacity to produce data for only 23 out of the 231 indicators,^[11] and will soon have to either devise an exhaustive system of data generation to track progress or to adapt indicators based on existing data sources. At the culmination of MDGs in 2015, the MoSPI, responsible for tracking and monitoring MDGs in India, highlighted that non-availability of data at the sub-state/district level prevented a comprehensive appraisal of the impact of MDGs.^[12]

The two largest sources of population-level data in India that remain primary for health and nutrition are the National Family Health Survey (NFHS), conducted by the MoHFW, and the National Sample Survey (NSS) conducted by MoSPI. Although the NFHS covers a wide range of indicators and will repeat every three years according to latest plans,^[13] doubts remain whether such frequency is even possible, given that the previous round happened ten years ago. Drafting NIF aligned with the SDGs provides an opportunity to streamline existing national surveys as the source for a comprehensive set of indicators around reproductive, maternal, newborn and child health (RMNCH), as well as nutrition.

Researchers have warned that finding reliable metrics to track India’s SDG progress is going to be a massive challenge, given the fragmented nature of the health and nutrition data infrastructure in the country.^[14] Keeping in line with the SDG theme of “leaving no one behind,” the inequalities within these themes that can be addressed require that each aspect is represented credibly in the policy formation processes and evaluation. Therefore, it is imperative that geography, gender, age, social groups, economic class and other dimensions be represented in the data.

The Observer Research Foundation’s (ORF) Health Initiative has organised various events and published a series of peer-reviewed papers, issue briefs, and special reports carrying policy recommendations on data limitations, challenges, and implications for the pursuit of SDG targets, with a particular focus on health and nutrition. This paper is part of the series, and the views presented are based on both a literature survey and analyses stemming out of the deliberations and recommendations from a technical focus-group discussion convened by ORF on the issue of streamlining existing national surveys to improve target-tracking in health and nutrition.^[15]

The paper builds on ORF’s existing work and outlines issues related to the feasibility of monitoring indicators through an examination of the available information base. As a starting point, the health and nutrition indicators listed in the draft NIF are classified into three categories using pre-defined criteria. The paper concludes with two sets of recommendations: the first set is specific to generation of better data by streamlining existing surveys through altering data construct and revising instruments for data collection, and the second is around the modifications to the NIF itself. Advocating for streamlining existing surveys, however, does not undermine the importance of improving upon other aspects of the health and nutrition data infrastructure, discussed elsewhere.^[16]

The rest of the paper is organised as follows:

With a focus on analysing the scope, availability and appropriateness of indicators in the NIF, section II presents a methodology for systematic classification of NIF indicators into categories of priority in terms of changes needed in the system to address gaps. Section III discusses the availability of data from existing national surveys to populate the NIF, based on the analysis of indicators under each category. Section IV distils the discussion on category 1, 2, and 3 indicators, and makes recommendations for streamlining existing surveys as well as revising NIF, acknowledging the limitations of the indicators and negotiating data constraints.

II. Methodology: Classifying the NIF Indicators

Under the UN General Assembly’s Resolution 70/1, an Inter-Agency and Expert Group (IAEG) was mandated to develop a global indicator framework for monitoring the goals and targets of a post-2015 development agenda that at its heart would entail equity and sustainability. The framework was expected to give a comprehensive set of indicators, which are simple yet robust and are reflective of inter-country diversities, preserving the ambition of bringing nations together irrespective of their levels of development. Thus, the development of the framework was in accordance with differing national realities, capacities and levels of development and respective national policies and priorities.^[17] In this context, the member-nations also recognise the importance of adequate data for the follow-up and review of progress, and agree to intensify efforts to strengthen capacity-building, especially for developing countries.

Following the broader mandate, the IAEG undertook a global assessment of the level of methodological development and overall data availability for the selected SDG indicators and grouped these into three different tiers; the global indicator framework currently contains 82 tier-I indicators, 61 tier-II indicators and 84 tier-III indicators. However, the status of data availability varies across countries.^[18]

The situation in India in the context of availability of data differs from the global picture that emerges from the tier classification. To assess the availability of health and nutrition data from existing national surveys, an exhaustive analysis of the data sources and their timeliness was undertaken which required mapping data sources for information and going into the nature of questions asked to collect specific information. The following section gives a detailed analysis of the indicators, based on the draft NIF.

Systematic Classification of the NIF Indicators

The following section presents a comparison of the NIF with the global framework, and a schematic classification. Building on the global schema, the indicators are divided into three categories, based on analysis at two levels. The first level pertains to analysing the availability of data through established sources in the country. The understanding of mapping current and potential data sources largely draws from ORF’s previous work on the issue.^[19] The availability of data for indicators is mapped concurrently with the level of disaggregation. In the case of unavailability of data for particular indicators, the possibilities of either modifying an existing question/biomarker in the household surveys (and/or in combination with some other sources of information) or adding a focused module within the survey tool is suggested in essence of the rationale and normative interpretation of indicators as laid by the UN. The second layer of analysis relates to the alignment of the NIF indicator with the specific SDG targets. Final classifications are arrived at based on the following norms:

Category 1 has indicators that are conceptually clear, has clear and direct questions/biomarkers covered and available through either NSSO (to be conducted in every five years^[20]) or NFHS (to be regularly conducted every three years^[21]) or other government sources that can offer state-level (for all states), and potentially sub-state-level indicators. Minor revisions are still suggested for certain indicators in this category, where availability of data at a further lower level, say district, was possible.

Category 2 has indicators that do not have conceptual clarity or ease of operationalisation through a methodology. Thus, some modification may be required in either the available data-generation methods through rephrasing questions, adding a focused module within the survey instrument, re-categorising the response codes or replacing them with information available through existing surveys. The replacement indicator was chosen based on availability, informed by reflection on the normative interpretation of the goal and respective targets. This category also includes those indicators that are conceptually clear, though the data is not available for all states and union territories (UTs).

Category 3 has indicators for which there are reliable data currently unavailable, or need to be revised completely, given that either there is no scope of including them in the current data-generation processes or they require a different mode of data generation through routine statistics, special surveys, rapid assessments, among other activities. Here, too, alternative indicators are suggested to be included in the NIF.

Based on this methodology, 52 indicators in the NIF directly related to health and nutrition were analysed. Seven of these indicators were repeated more than once. Two indicators, (Indicator 3.8.16, percentage of attributes of 13 core capacities, repeated once; and Indicator 3.b.2, total net official development assistance to the medical research and basic health sectors) were excluded from the analysis as they were beyond the scope of this exercise. Thus, in all, 43 indicators were categorised. The detailed matrix generated as part of the analysis is given in the annexure. The following session discusses, category-wise, the individual indicators that were analysed.

III.  Availability of Data for the NIF from the National Surveys: A Discussion

The analysis of the indicators from the draft NIF was done keeping in mind the availability at the smallest disaggregated level, which in the Indian context is the district. However, it should be noted that availability of some indicators is perceptibly poor at the state level and data at district disaggregation is highly limited. Therefore, those indicators that are available at the state level for all states and UTs, with some scope for sub-state estimates, were added in the first category. Currently, 15 out of the 43 indicators—after removing the seven repeat indicators—are in category 1, with reliable data available at the state level for all states, both big and small. Out of these, only a few have data available at the district level.

In the following section, each indicator is discussed under these categories:

1. Indicators Under Category 1:

Indicators 2.2.1 and 2.2.2 pertain to prevalence of stunting and malnutrition among children aged five and less than five, respectively. With a fivefold increase in the sample size of the survey, NFHS-4 can give robust estimates at the district level. The indicators 3.2.1, on under-five mortality rate, and 3.2.2, on neonatal mortality rate, are available through the NFHS survey, given that the survey has RMNCH as a focus. The mortality rates can be estimated using established techniques based on birth history and survival information obtained in the survey. None of these mortality rates can, however, be estimated beyond state level, given the sample size constraints. Strengthening of CRS alone can help provide regular sub-state mortality numbers.

The coverage indicators, such as indicator 3.1.2, on skilled birth attendance, is available at the district level from the NFHS. In addition, there is a group of indicators^[22] for age-related fertility and health-service utilisation for maternal and family planning services, which have disaggregated data available from NFHS. The indicator 3.7.2 on adolescent births is kept in the first category, despite the possibility of high relative sampling errors in general and specifically for the states where the proportion of child marriages are low. Even while estimation has challenges, NFHS is expected to offer district-level estimates, where marriage before legal age is a problem. As the age of marriage is going up across the country, the error margins will increase, and for this reason, attention can be given to indirect estimates that can offer a robustness check for the directly obtained estimates.

There are a few indicators for which the data through both NFHS and NSS can be triangulated for target tracking and progress appraisal. The indicator 6.1.2 on proportion of population using an improved drinking water by source as well as indicator 6.2.2 percentage of population using basic sanitation services are available with NSS and NFHS surveys. Description-wise, “Basic sanitation” connotes improved sanitation,^[23] and district-level numbers are readily available from NFHS-4. However, before proceeding with any exercise of triangulation, there is a need to carefully evaluate the definitional aspects and their match between NFHS and NSS. For example, there is a variation in the definition of what is labelled as “improved drinking water sources” across NFHS-4 and NSSO 69^th round.

 Indicator 3.8.14, dealing with human resources for health (HRH), can be made available from Census as well as NSS’ employment rounds based on occupational classification, which can be obtained at the state level. NSS will be the choice, as Census numbers will be available only after considerable delay.^[24] While it is possible to arrive at a composite number, having separate estimates for nurses and midwives is currently impossible as these categories are merged at the point of data collection. NSS samples are sometimes considered too small for arriving at robust state-level estimates.^[25] However, assigning population weights using the Census can help in arriving at robust employment estimates. The pooled data of state and central samples of NSS should enable generation of estimates at the sub-state level.^[26] Given the prominent role they play in the delivery system, a separate code for community health workers in NSS can help get better, policy-relevant estimates. Currently, the National Classification of Occupations, which NSS and Census follow, does not have a separate code for community health workers. In terms of analysis, this remains a major blind spot.^[27]

Data for the indicators 3.8.17 and 3.8.18 on out-of-pocket (OOP) expenditure on health and poverty headcount due to OOP expenditure on healthcare can be obtained from NSS morbidity rounds. However, there are limitations to the conventional measures of catastrophic financing that systematically excludes people who are unable to seek necessary care due to lack of purchasing power.^[28] To mitigate such deficiency, a complementary ‘health need-based’ analysis with an indicator on the proportion of people not accessing medical care for financial reasons will furnish a better picture. Moreover, given the focus on insurance-based interventions, harmonising the data on health insurance collected by NFHS and NSS is necessary. While NFHS collects data on ESIS, CGHS, and schemes for BPL populations separately, NSS treats all “government-funded insurance schemes” as a single entity, reducing scope of analysis. However, before modifying survey instruments, sampling adequacy for each of these separate categories will need to be kept in mind.

Lastly, for the indicator 3.8.12 pertaining to tobacco use among individuals, NFHS provides information on the “use of any kind of tobacco” from men and women, aged 15–49, at the district level.

 2. Indicators Under Category 2:

Out of the total of 43 indicators analysed, nine belong to category 2, which represents the set with data sources that have some scope for improvement or where possible alternatives offering better geographical coverage are available.

The global indicator pertaining to the indicator 2.1.1 on “prevalence of undernourishment”^[29] refers to lack of calorie intake, in terms of minimum dietary energy requirement norms. Since calorie consumption is collected by the NSS only at the household level and not at the individual level in India, the scope of analysis will be limited. District-level indicators using the NSS data may be possible only if pooling of state as well as central samples are conducted. For this reason, while using “prevalence of undernourishment” for global reporting, it may be prudent for India to use “percentage of underweight” (children) or “percentage of low BMI” (adults) for sub-national tracking. As undernutrition is a widely prevalent problem, with pockets showing highly worrying numbers, it will make policy sense to track these indicators disaggregated for socioeconomic categories at the district level. NFHS provides data for these indicators at the district level and disaggregated estimates can be generated for children, women and men, although not covering the overall population, and only children and the reproductive age group will remain a handicap. Moreover, given that undernourishment is not spread uniformly across all segments of the population at the sub-national level, any indicator expected to capture inequity in it should have the sufficient sample size at the sub-group level to provide comparable estimates.

Indicator 3.1.1 on “maternal mortality ratio” is an extremely important one, given the scale of the problem in the Indian context. Data on maternal deaths in the two years preceding the survey have been collected in NFHS rounds 1, 2 and 4, which can give comparable estimates.^[30] While the earlier sample design may have been inadequate for robust state-level estimates of MMR, the enhanced sample of the latest round is expected to give estimates for states with confidence intervals to specify the level of precision. However, the relative sampling errors may be huge for smaller states, and therefore, it is kept in category 2. The Sample Registration System, under the office of the Registrar General of India (RGI), also gives estimates of MMR from pooled data for states, but covers only bigger states.^[31]

The data for indicator 3.3.1 on new cases of HIV infections per 1,000 unaffected populations is generated by the National AIDS Control Organisation (NACO) on a regular basis through estimation based on Integrated Biological and Behavioural Surveys (IBBS) and HIV sentinel surveys (HSS). The IBBS and the HSS are regarded by the World Health Organization (WHO) as the best sources for generating statistics for HIV/AIDS, based on disease-characteristics using a network of health facilities delivering care to the populations exposed to it through a specific mode of transmission. The latest round of NFHS too offers to provide information on HIV prevalence for adult women and men at the national level and for 11 states/groups of states/UTs, including high HIV prevalence states. These estimates are expected to be used to triangulate the HIV estimates that are based on NACO’s surveillance data. ^[32]

Data for the indicator 3.3.2 on Tuberculosis (TB) incidence is often sourced from the countrywide programme aimed to identify and treat people with TB. TB being a notifiable disease, the programme generates statistics based on numbers from public and private sectors. However, due to the limited success with notification of TB by the private sector, there is serious underestimation.^[33] The NSS morbidity round as well as NFHS have questions on (self-reported) TB, and can potentially give state-level estimates of self-reported TB. However, data quality and representativeness need to be ascertained to determine robustness of estimates at national and state level.

The data pertaining to the indicator numbered 3.8.6 of the draft NIF on the percentage of PLHIV (People Living with HIV) currently receiving ART (Antiretroviral Therapy) among diagnosed adults and children living with HIV is available only through HSS. This data might be available even at the district level from the programme records, as ART medication is strongly monitored via diagnostics, and all the hospitals providing free drugs maintain records at the state headquarters.^[34] However, coverage of private sector will still be an issue.

Use of Insecticide Treated Nets (ITN) in the malaria-endemic regions is incorporated as indicator 3.8.7 under the draft NIF. The NFHS gives the use of bed-net (for the previous night) by members of a family. However, the use of bed-nets the previous night may not be a good indicator to look at the disease burden, primarily because of bias in reporting due to seasonal changes, and this indicator needs revision. A possible alternative can be “the proportion of households that have an insecticide treated net/bed net,” based on NFHS data. However, this indicator can only report mere presence and not usage. Thus, complementing the information on presence, usage and classification of nets based on treatment can be added to the existing survey instruments.

The indicator 3.8.9 in the NIF reads “Proportion of population (aged 18 years and above) who are currently taking antihypertensive medication among number of adults (18 years and older) who are taking medication for hypertension with systolic blood pressure ≥ 140 mmHg, or with diastolic blood pressure ≥ 90mmHg.” This is likely to be a typographical error for “proportion of population (aged 18 years and above) on hypertensive medications to the total people with systolic blood pressure ≥ 140, or with diastolic blood pressure ≥ 90mmHg.” The information on blood pressure is collected in NFHS-4 in the biomarker survey separately for men and women (aged 15–54) along with self-reporting on being prescribed drugs. The survey covers just the eligible population of reproductive age group, thus limiting the usefulness in estimating the real burden of hypertension, a disease that mostly impacts older people. However, a combination of NFHS and Longitudinal Aging Study in India (LASI) survey on aged can generate trend analysis on the mutually exclusive age categories. First wave of LASI is currently underway and will only cover people aged 45 years and above. Despite this weakness, it still can be a valuable addition to the NIF.

The question related to the target 3.8.11 on the proportion of women (aged 30–49) ever being examined for cervical cancer can be accessible from NFHS women questionnaire. Re-classifying eligible women (from 15–49) to the required age group will indicate proportion of women ever undergoing a cervix examination. However, it is unclear from the way the question is framed and the placement of the question (with screening for oral cavity and breasts) in the questionnaire and interviewer manuals whether this question on ever undergoing a cervix examination meant undergoing a preventive screening for cancer or not. If the indicator in NIF aims to assess the coverage of cancer-specific screening, the percentage of women who have ever undergone breast examination will capture the essence, since it is a common type of cancer among women and the information is available with NFHS.

The indicator 3.8.13 on number of Outpatient Department (OPD) visits per person per year and Inpatient Department (IPD) admission rate per 100 population per year is given directly by the NSSO surveys on health. The data is available from the NSSO morbidity survey. However, robust district-level numbers may not be available unless central and state samples are pooled.

3. Indicators Under Category 3:

Predictably, category 3 constitutes the largest number of indicators, reflecting the huge scope for improvement. In all, 19 indicators out of the total 43 fall under this category.

The indicator 2.1.2 pertaining to the prevalence of moderate or severe food insecurity based on the recommended FIES does not have any existing data source in India. However, FIES is being canvassed as part of an ongoing survey called “Comprehensive National Nutrition Survey” (CNNS) conducted by Unicef and MoHFW.^[35]

The incidence of malaria (indicator 3.3.3) is not available through any source except district-wise malaria incidence and blood examination reports provided by the National Vector Borne Disease Control Programme (NVBDCP). The NSS too gives some information through the healthcare utilisation in the surveys on “social consumption of health,” which will now give us data after every five years. NVBDCP data reports are deemed incomplete as they do not cover the private sector.^[36] The reported diagnosis and/or symptom for which IPD and OPD care was sought in NSS gets coded as “nature of ailments,” and can be used to ascertain incidence of malaria if a separate code other than “all other fevers” is given for malaria. Interestingly, in the previous morbidity round (2004), a separate code for malaria was given, which has been removed in the latest round for reasons that are unclear.^[37] The estimates, however, may suffer from seasonality bias, as malaria is mostly reported during monsoon and post-monsoon period. Ideally, data for this indicator should be provided by routine surveillance.

Data for the indicator 3.3.4 on incidence of viral hepatitis (including A, B, C, D, E) are not available through any surveys. Although the disease codes in NSS do not include hepatitis, the report clarifies that the working definition for jaundice includes hepatitis as well.^[38] While self-reported information may not be ideal for assessing the true incidence, NSS and NFHS might consider having a separate question on hepatitis. As with malaria, ideally, routine surveillance alone can fetch the information for this indicator, although only hepatitis A and B (not C, D and E) are notifiable, under the regulatory framework of the GoI. Since data is not currently available, immunisation coverage can be used instead as a temporary proxy for risk, even as only hepatitis B is part of universal immunisation.

The indicator 3.3.5 on the proportion of population requiring interventions for the treatment of the Neglected Tropical Diseases (NTDs)^[39] does not have data sources except through the corresponding programme Management Information System (MIS), which do not cover the private sector adequately. However, some diseases such as dengue, chikungunya and viral leishmaniasis are endemic in particular seasons in certain spots. Therefore, the government can think of conducting rapid assessments of the burden during outbreaks through household surveys conducted in the endemic areas or through sentinel surveys. In the NSS morbidity round, in the “nature of ailment” variable, apart from “all other fevers,” dengue must be captured separately with a separate code. As an emerging disease causing considerable morbidity/mortality burden, dengue can be tracked as a representative indicator for NTDs. Given that previous exposure to dengue is known to increase the potency of Zika infection, more data on dengue is necessary for system preparedness.  Since SDGs mandate monitoring of the disease burden for the NTDs, MoSPI can think of including them as a separate disease category in the health survey conducted by NSS.

The data pertaining to indicator 3.4.1 on the mortality estimates attributable to cardiovascular diseases, cancer, diabetes or chronic respiratory diseases are not collected in any regular surveys, although some morbidity estimation is available. Only Medical Certification on Causes of Death (MCCD) surveys conducted by Sample Registration System (SRS) can give data on causes of death through household surveys. However, if the goal is to reduce the burden of non-communicable diseases (NCDs), then taking the second-best available statistics on “morbidity due to NCDs” instead of mortality is a feasible option.

Data on causes of death may become redundant if seven years’ worth of data are pooled to get state estimates, as is being done currently.^[40] Rates for diabetes and hypertension can be used as proxy for risk, provided by NFHS. Additionally, self-reported cancer, asthma and heart disease numbers for men and women can be used from NFHS.

For indicator 3.4.2 on suicide mortality rate, data sources within surveys remain absent. National Crime Records Bureau (NCRB) is the only source of data on suicide mortality. The NCRB data is incomplete, given the substantial definitional issues on suicide as well as under-reporting.^[41] Thus, relying on SRS–MCCD based on verbal autopsy remains the only way. The estimates for 2007–13 give data for bigger states and UTs,^[42] which come with considerable delay. Alternatively, NFHS household questionnaire has a question—“Was the death in the family due to an accident, violence, poisoning, homicide or suicide?”—that can be used to estimate this indicator, if the data can be collected for these categories using separate codes rather than a composite one. Alternatively, NSSO morbidity surveys can be used to obtain information on disease burden under “mental illnesses” at the state level.

Reliable data for indicators 3.5.1 and 3.5.2 on coverage of treatment for substance abuse and harmful use of alcohol among people aged 15 years and above remain absent in India. A one-off survey titled “National Survey on Extent, Pattern and Trends of Drug Abuse in India” was done in 2004 by Ministry of Social Justice and Empowerment (GoI) and the UN Office of Drugs and Crime. There are routine data available from 122 centres offering interventions on Drug De-Addiction Programme (DDAP), MoHFW. Usefulness of routine data remains an issue as the coverage of private sector is a major challenge. In the long run, strengthening collection of routine data from the DDAP, including private sector, can be the way forward.

In the case of alcohol, there is a need to streamline the construction of data collected by the regular surveys (NFHS and NSS). NFHS asks a question on alcohol consumption to men and women eligible for regular investigation, and NSS records consumption in the consumer-expenditure surveys. NFHS asks men and women about their drinking habits including frequency and type of alcohol and NSS records quantities consumed by households in a month. NFHS records the frequency and type of alcohol but not quantity. NSS cannot give individual consumption, and is likely to underestimate because of underreporting as well as poor recording of consumption outside home.

Defining “harmful use” by restricting type and/or frequency (proxy for quantity) can be a way forward. In addition, NFHS and NSS can revise questions regarding alcohol consumption respectively. NFHS can include a question on quantity along with frequency and type of alcohol consumption. NSS can consider including individual consumption of alcohol and tobacco products in a separate module to record consumption of food, tobacco and liquor (in the house and outside) by individuals. For the short run, the NIF indicator can be the proportion of people usually drinking “hard liquor” and drinking “almost every day,” given the availability of this data with NFHS for the district level.

Data for the indicator related to mortality due to traffic injuries (3.6.1) have issues similar to suicide mortality data. NCRB remains unreliable, and no household survey data are available on death rates other than SRS. In “causes of death,” SRS has a category “unintentional injuries: motor vehicle accidents,” but its policy relevance is questionable if we pool seven years’ data, with considerable delay, to get state estimates.

The NSS morbidity survey question on the “nature of ailment” across IP and OP can work as a proxy. However, currently, “accidental injury, road traffic accidents and falls” are grouped together as a single variable. Separate data on road traffic accidents will have to be collected by NSS in the “nature of ailment” question. Alternatively, the NFHS question, “Was the death in the family due to an accident, violence, poisoning, homicide or suicide?” can be tweaked to have a separate answer on traffic accidents.

Data required for the indicator 3.8.3 on the percentage of children (aged 12–23 months) who received three doses of pentavalent vaccine before their first birthday is not available through any survey. The NFHS survey, however, collects information on immunisation for four of the five diseases covered under pentavalent vaccine barring HiB vaccine for haemophilus influenzae type b. Subsequent waves of NFHS can add this information. For now, the percentage of fully immunised children—percentage of children (aged 12–23 months) fully immunised for BCG, measles, and three doses each of polio and DPT—can be a viable alternative, for which data is available at the district level. Especially if full immunisation coverage is not part of the final NIF as an indicator for target 4.2, it should be part of target 3.8.

Data for the indicator 3.8.4 in the NIF, the percentage of children under five years of age with suspected pneumonia (cough and difficulty breathing, not due to a problem in the chest and a blocked nose) in the two weeks preceding the survey who sought care from appropriate healthcare facility or provider, does not exist in any national-level source in India. Nevertheless, all rounds of the NFHS collect data on the “percentage of children with acute respiratory infection (ARI) in the last two weeks preceding the survey who sought care”, which can be considered a proxy for pneumonia. This data is available for the district level.

NIF’s indicator 3.8.5 on the percentage of TB cases successfully treated (plus treatment completed) is available from RNTCP programme data at the state and district levels. However, coverage of the private sector remains a major gap and the programme data cannot give reliable estimates of cure rate; therefore, an alternative indicator needs to be explored. NFHS-4 gives information on morbidity as well as treatment, although the exact information on cure rates may not be available. NFHS-4 can give the “proportion of people self-reporting TB who are ever treated.” In the long run, the coverage and engagement of private sector should be scaled up to get reliable estimates of the burden and to ensure treatment adherence, thus reducing chances of developing resistance.

Indicator 3.8.8 on population using safely managed drinking water and sanitation services is a combination of two other NIF indicators (6.1.1 and 6.2.1). While NSS and NFHS collect data on the nature of source of water as well as sanitation facility, the phrase “safely managed” is difficult to operationalise, and currently, there is no data available in the Indian statistical system on this aspect, other than on some elements such as handwashing. Dropping these three indicators is suggested, as NIF already has other indicators on “proportion of population using an improved drinking water by source” as well as on “percentage of population using basic sanitation services,” which have been discussed earlier.

The indicator 3.8.10 in the NIF, reads “proportion of population (aged 18 years and above) who are currently taking medication for diabetes (insulin or glycaemic control pills) among the number of adults (18 years and older) who are taking medication for diabetes or with fasting plasma glucose ≥ 7.0 mmol/L,” which, in all probability, was meant to be the proportion of people aged 18 and above on medication for diabetes out of the total people with a fasting plasma glucose reading of more than or equal to 7.0 mmol/L (126mg/dL). However, NFHS has data on random (and not fasting) diabetes test, and information on medication is not available. That NFHS is limited to the reproductive age group somewhat limits the usefulness of information as well. However, as it allows for trend analysis every three years, a modified indicator covering treatment of diabetes, customised to align with available data with NFHS can still be a very valuable addition to the NIF.

Indicator 3.8.15 in the NIF, “percentage of health facilities with essential medicines and lifesaving commodities” does not have possible data sources in its current form. Centralised drug procurement and disbursal records of essential medicines by each state can give some information, but it cannot help estimate the shortfall of medicines. Arriving at an operational definition of “lifesaving commodity” may add an extra layer to the problem at hand. In this context, health-seeking behaviour by local population can be a good proxy of dependability and quality of public sector. NSS collects data on the percentage of people reporting ailments who didn’t choose government hospitals, because of “required specific services not available” or “available but quality not satisfactory,” which can be used in the revised NIF.

Indicators pertaining to mortality due to household/ambient air pollution (3.9.1), unsafe sanitation, lack of hygiene (3.9.2) and unintentional poisoning (3.9.3) are difficult to obtain from the household survey for the reason that attribution of death to these causes is difficult unless it is an extreme case. SRS data at the state level will have limited policy relevance because of the delay involved and the necessity for pooling many years of data. This set of indicators need to be revised, with a focus on morbidity burden instead. Self-reported asthma from NFHS will be a good proxy for 3.9.1 and self-reported diarrhoea from NFHS will be a good proxy for 3.9.2. Data for indicator 3.9.3, mortality rate attributed to unintentional poisoning, can be potentially collected by NFHS, which has the question: “Was the death in the family due to an accident, violence, poisoning, homicide or suicide?” However, as discussed earlier, data for these categories will have to be coded and collected separately to develop this indicator. Alternatively, NSS data on households reporting morbidity due to poisoning can be used to develop a new indicator.

 Indicator-wise scrutiny of availability of questions and/or biomarkers in each household survey instrument relating to the feasibility or tracking capacity yielded a comprehensive understanding of sources of data vis-à-vis the NIF. Based on the exercise, it can be concluded that respective health surveys need to furnish indicator data that are comparable across rounds and between surveys. The periodicity of surveys need to be driven by need, purpose, opportunity cost as well as by the potential for driving programmatic decisions on the ground. Preferably, year-long surveys capturing seasonal variability with standard recall periods, and internationally prescribed disease codes with sub-modules on focused diseases will allow for detailed analysis of disease burden at the sub-national levels. The sufficiency of sample representative of tenable administrative levels will not only enable target tracking at disaggregated levels but will also facilitate programme monitoring. Adding fresh questions on themes with limited availability of information through surveys will also be useful.

The current data-generation process can be smoothened to incorporate critical information on insurance enrolments, coverage and usage along with a wide-ranging information on consumption expenditure in general, and health expenditure in particular. In a nutshell, combining the strengths of various surveys and encompassing a holistic design that caters to the programmatic needs will be the way forward. The availability of data related to indicators will improve tremendously through a dialogue aimed at harmonising surveys undertaken by different agencies. Some surveys that cover different thematic areas run the risk of duplicating information sought from different purposively selected sub-populations. Internal consistency within each questionnaire can be improved by linking questions based on cause of hospitalisation/outpatient visits with the household-level health information where questions on specific illness were asked. Specific instances of changes in existing surveys through re-categorisation of responses as well as possibilities of streamlining survey questionnaires through improving upon minor inconsistencies of the schedules will be discussed in the concluding session.

Conclusion

In light of the discussion in the previous section, this paper suggests two sets of recommendations. The first will offer modifications to the existing collection tools so that data that are better aligned with the NIF can be regularly made available. The second set of recommendations will offer realistic modifications to the health and nutrition indicators of the NIF so that the framework itself is aligned with available data sources in the country.

Of all the indicators under category 1 for which relatively better data are available, barring three that deal with HRH and health spending, all others are sourced directly from NFHS. Apart from mortality and fertility indicators, NFHS gives district-level estimates for all indicators within this group. It is imperative for tracking India’s health and nutrition performance that NFHS surveys are conducted every three years, and delays similar to the one between NFHS-3 and 4 do not happen again.

Based on the analysis of the indicators and questions corresponding to each indicator, it is clear that even for many indicators under category 1, current surveys can only furnish reliable estimates at the state level, and sub-state estimates suffer from large sampling errors. If the decision by the MoHFW to enhance the periodicity of NFHS to every three years^[43] is indeed implemented, the sample size can also be enhanced further to enable production of robust district-level estimates for important indicators under this category. In a similar vein, an initiative driven by MoSPI at the centre is needed to pool either state and central samples or statistics furnished via both the samples. NITI Aayog’s Three-Year Action Agenda suggests an agency at the centre to collate and disseminate district-level data, comparable across states.^[44] Pooling of central and state NSS samples can be one of the first initiatives for such an agency to undertake, given the enormous amounts of money spent on collecting the state sample, which is seldom used. For example, the 66th Round of CES had all participating states, other than Gujarat, collecting information from at least an equal number of households as the central sample.^[45] However, to obtain reliable estimates for certain low prevalence diseases even at the state level, the use of confidence intervals arrived at using statistical methods can facilitate indicator tracking.

Suggestions for indicators of category 2 are mostly specific on streamlining existing surveys. The procedures through which the current surveys can address the data needs triggered by the draft NIF are diverse. A question on individual consumption of food, tobacco products, and alcohol (consumed inside and outside the house) along with itemised expenditure for a uniform recall period across rounds will be useful. Revising the broad international classification of diseases to incorporate diseases focused in SDG indicator framework (viz. malaria, TB, NTDs) as separate categories within the ICD framework will be useful in generating information on indicators. Planning a small module focusing on NCDs in the biomarker survey of NFHS furnishing information on medication will add tremendous value. This is also advisable given the changing profile of diseases in the context of the demographic and epidemiological transition. Consistent ailment coding will facilitate inter-temporal comparisons and target-tracking.

Current surveys do not capture the quantity of alcohol consumption. Adding a simple question on this will enable characterising alcohol consumption. However, defining “harmful use” of alcohol is left for countries to contextualise based on availability and quality apart from behavioural attributes. It needs to be borne in mind that routine surveys on health may not be able to collect information on substance abuse because of the stigma attached to their use. Particularly in cultures where reporting on alcohol/smoking tobacco use by women could be stigmatising, reliability of collected data may suffer. Nevertheless, emulating the Global Adult Tobacco Survey by WHO on tobacco consumption along with other substance abuse on regular basis can be considered with larger sample size to obtain disaggregation.

Deaths due to road accidents, unintentional poisoning and suicides are areas with scanty information in the surveys. The causes of death can be re-categorised within existing instruments, or more detailed information can be collected via modules on the dead within individual surveys. This does not undermine the importance of CRS, which is possibly the best source for producing the vital statistics such as this. Inclusion of better information regarding consumption expenditure in health surveys is also desirable and can facilitate estimation of impoverishment due to health payments/OOP expenditure.

Category 3 has indicators that need redefining based on the objective for which they need to be obtained. There are indicators that have greater annual variability, which can better be explored through rapid assessment surveys. Some indicators covering rarely occurring events could be explored through cohort studies or through employing the indirect estimation techniques. The former solution can be applied for vector-borne diseases including malaria and NTDs. The latter, however, pertains to mortality estimates (maternal deaths, deaths due to NCDs), and fertility estimates (adolescent birth rate). It is also worth considering that the indicators on mortality due to NCDs and on the population requiring interventions for the treatment of the NTDs be replaced with morbidity-related indicators, given the feasibility of generating these with minor tweaks within existing household surveys.

General Recommendations

Revising the classification of diseases in the national surveys to incorporate diseases focused on the SDG indicator framework is important. In addition, having NSS morbidity rounds on the ground for a year, rather than six months as done now, will curtail the seasonality bias. NSS may also add a separate question on the common NTDs that can then be linked to healthcare utilisation, so that morbidity as well as economic burden can be easily estimated. In addition, an NSS module on health, covering specific morbidities, can be canvassed with yearly rounds.

While the sampling errors are important and should be addressed, the non-sampling errors that creep into large-scale surveys due to lengthy schedules, large sample size to ensure geographic coverage, and non-response should be addressed. Further, quality of sensitive information such as tobacco use and alcohol consumption will improve if questions are asked to respective individuals rather than the head of the household in the household questionnaire.

Information from the morbidity modules cannot be used to estimate incidence rates, as it is based on lay reporting. Similarly, many respondents might not know the names of the disease for which symptoms were presented and treatment was sought. Some layer of conformity to the self-reported morbidity will make the information more robust. However, supplementing self-reported illness with information on diagnosis/medication does not reflect actual prevalence as diagnosis/medication is largely conditioned by access and programme coverage. Thus, cautious combinations of such complementary information can provide a better view of the disease situation.

Additionally, to make the statistics on OPD and IPD visits representative at the district level, it is worthwhile to consider pooling of central and state samples. If the poolability of the centre and state sample is not viable due to the structural issues in the data construct, then generating sub-state estimates from the matching sample of the state is recommended, provided synchronicity can be established. Further, indicators can be developed by judiciously combining statistics from NSS state samples and Census-based indirect estimates, especially for indicators that are not amenable to direct estimations.

Regarding NFHS, while working with the district-level survey design, experts must be engaged to explore options, within ethical limits, to allow codes and weights to districts, blocks and households for clubbing them to enable some level of parliamentary constituency-level analysis. Having constituency-level health and nutrition indicators can be an important step towards a re-prioritisation of public policy.

Given that the SDG indicators are meant to monitor inequity, it is important to recognise that the known axes of inequity such as gender, class and socioeconomic status need to be monitored. For that reason, the obtained estimates must be robust for time series comparisons, especially when concerning indicators that are declining, resulting in events getting rarer (which might result in existing sample yielding estimates with wide confidence intervals). In such cases, care should be exercised to revise the sample frame for routine sample surveys at adequate frequency to enable reliable comparisons.

Specific Recommendations towards Modifying Surveys:

• In the NSS morbidity round, adding a question on sector (Urban/Rural), to reflect the location of the healthcare facility where care is sought will yield valuable information for analysis. NFHS too should start collecting information on the location of the facility. The current urban/rural disaggregation is based on the location of the household.
• In the NSS employment round, it will be useful to have nurses and midwives as separate codes, rather than a combined code. This can be done for the Census as well.
• In the NSS employment round as well as the Census, adding a separate code for “community health workers,” given their prominent role in the delivery system.
• Within CES, adding a question on individual consumption of food, tobacco products, and alcohol (consumed inside and outside the house) along with itemised expenditure for a uniform recall period across rounds.
• Avoiding proxy reporting for behavioural attributes.
• Harmonising the codes of the question on health insurance between NSS and NFHS to make the information collected comparable.
• Adding questions in relevant national surveys to enable estimation of Prevalence of Undernourishment (PoU) and food insecurity at sub-national levels.
• In the NSS morbidity survey, asking a question on Malaria with a one-year recall, which can provide information that can be cross-checked with the IPD/OPD modules.
• In the NSS morbidity survey, asking a question on Dengue with a one-year recall, which can provide information that can be cross-checked with the IPD/OPD modules.
• In the NSS morbidity survey, asking a question on Hepatitis with a one-year recall, which can provide information that can be cross-checked with the IPD/OPD modules.
• Within the NFHS household questionnaire, adding questions on Malaria, Dengue and Hepatitis.
• Within the NSS morbidity schedule, modifying the “nature of ailments” codes to collect separate information on Malaria (currently fused with other fevers).
• Within the NSS morbidity schedule, modifying the “nature of ailments” codes to collect separate information on Dengue (currently not collected).
• Within the NSS morbidity schedule, modifying the “nature of ailments” codes to collect separate information on Hepatitis. Currently, the working definition of Jaundice within NSS morbidity round clarifies that it includes Hepatitis as well.
• Within the NSS morbidity schedule, modifying the “nature of ailments” codes to collect separate information on Chikungunya, to enable study of its economic burden.
• Modifying the answer codes of question no. 76 in the NFHS household questionnaire to collect information on accident, violence, poisoning, homicide and suicide separately. Category of poisoning can be changed to “unintentional poisoning” to avoid overlap with suicide. Category of “accidents” can be split to “road traffic accidents” and “other accidents” for collecting data aligned with the NIF.
• Adding a question in NFHS on the quantity of alcohol consumed.
• Modifying NSS morbidity round’s “nature of ailment” coding to separate accidental injury, road traffic accidents and falls.
• Adding a separate question in NFHS on the coverage of the pentavalent vaccine.
• The wealth-related questions that is an important marker of socioeconomic status in NFHS should be reframed in keeping with development context and comparability. For instance, availability of chair and table are perhaps now a weak marker to categorise households in different wealth quintiles.
• After consultation with experts, national surveys should have questions on physical activity, disability, and mental health.

The second set of recommendations is in the form of a modified NIF, based on the discussion in section III.

Table 1

A PROPOSED NATIONAL INDICATOR FRAMEWORK FOR HEALTH AND NUTRITION, COVERING SELECT INDICATORS
Targets	Draft National Indicator Framework (NIF)	Proposed NIF
2.1: By 2030, end hunger and ensure access by all people, in particular the poor and people in vulnerable situations, including infants, to safe, nutritious and sufficient food all year round.	Indicator 2.1.1: Prevalence of undernourishment	Low BMI for adults Underweight for children below five
	Indicator 2.1.2: Prevalence of moderate or severe food insecurity based on FIES	Indicator 2.1.2: Prevalence of moderate or severe food insecurity based on FIES
2.2: By 2030, end all forms of malnutrition, including achieving, by 2025, the internationally agreed targets on stunting and wasting in children under five years of age, and address the nutritional needs of adolescent girls, pregnant and lactating women and older persons.	Indicator 2.2.1: Prevalence of stunting (children under five years of age)	Indicator 2.2.1: Prevalence of stunting (children under five years of age)
	Indicator 2.2.2: Prevalence of wasting (children under five years of age)	Indicator 2.2.2: Prevalence of wasting (children under five years of age)
3.1: By 2030, reduce the global maternal mortality ratio to less than 70 per 100,000 live births.	Indicator 3.1.1: Maternal mortality ratio	Indicator 3.1.1: Maternal mortality ratio
	Indicator 3.1.2: Proportion of births attended by skilled health personnel	Indicator 3.1.2: Proportion of births attended by skilled health personnel
3.2: By 2030, end preventable deaths of newborns and children under five years of age, with all countries aiming to reduce neonatal mortality to at least as low as 12 per 1,000 live births and under-five mortality to at least as low as 25 per 1,000 live births.	Indicator 3.2.1: Under-five mortality rate	Indicator 3.2.1: Under-five mortality rate
	Indicator 3.2.2: Neonatal mortality rate	Indicator 3.2.2: Neonatal mortality rate
3.3: By 2030, end the epidemics of AIDS, TB, malaria and neglected tropical diseases, as well as combat hepatitis, water-borne diseases and other communicable diseases.	Indicator 3.3.1: Number of new HIV infections per 1,000 uninfected population	Indicator 3.3.1: Number of new HIV infections per 1,000 uninfected population
	Indicator 3.3.2: TB incidence per 100,000 population	Indicator 3.3.2: TB incidence per 100,000 population
	Indicator 3.3.3: Malaria incidence per 1,000 population	Indicator 3.3.3: Malaria incidence per 1,000 population
	Indicator 3.3.4: Viral hepatitis (including A, B, C, D, E) incidence per 100,000 population	Children age 12–23 months who have received three doses of hepatitis B vaccine (%)
	Indicator 3.3.5: Number of people requiring interventions against neglected tropical diseases (dengue, chikungunya, kala-azar, leprosy, lymphatic filariasis, soil transmitted helminths, v. leishmaniasis)	Indicator 3.3.5: Number of people who sought IPD/OPD care in the last one year/15 days for dengue
3.4: By 2030, reduce by one-third premature mortality from NCDs through prevention and treatment, and promote mental health and well-being.	Indicator 3.4.1: Mortality rate attributed to cardiovascular disease, cancer, diabetes or chronic respiratory disease	Indicator 3.4.1: Morbidity rates for: a) Blood sugar level: high (>140 mg/dl) (%) b) Hypertension Slightly above normal (Systolic 140–159 mm of Hg and/or Diastolic 90–99 mm of Hg) (%) c) Self-Reported Cancer d) Self-Reported heart disease
	Indicator 3.4.2: Suicide mortality rate	Indicator 3.4.2: Morbidity attributable to mental disorders
3.5: Strengthen the prevention and treatment of substance abuse, including narcotic drug abuse and harmful use of alcohol.	Indicator 3.5.1: Coverage of treatment interventions (pharmacological, psychosocial, and rehabilitation and aftercare services) for substance use disorders	Indicator 3.5.1: Proportion of households reporting ganja and “other drugs” (NSS)
	Indicator 3.5.2: Harmful use of alcohol (aged 15 years and above) within a calendar year in litres of pure alcohol	Indicator 3.5.2: Proportion of people usually drinking hard liquor and drinking almost every day (NFHS)
3.6: By 2020, halve the number of global deaths and injuries from road traffic accidents.	Indicator 3.6.1: Death rate due to road traffic injuries	Indicator 3.6.1: Morbidity attributable to road traffic accidents.
3.7: By 2030, ensure universal access to sexual and reproductive healthcare services, including for family planning, information and education, and the integration of reproductive health into national strategies and programmes.	Indicator 3.7.1: Proportion of women (aged 15–49 years) who have their need for family planning satisfied with modern methods	Indicator 3.7.1: Proportion of women (aged 15–49 years) who have their need for family planning satisfied with modern methods
	Indicator 3.7.2: Adolescent birth rate (aged 15–19 years)	Indicator 3.7.2: Women (aged 15–19 years) who were already mothers or pregnant at the time of the survey (%)
	Indicator 3.7.3: Proportion of deliveries attended by skilled health personnel	Repeat Indicator. Dropped.
	Indicator 3.7.4: Proportion of institutional deliveries	Indicator 3.7.4: Proportion of institutional deliveries
3.8: Achieve universal health coverage, including financial risk protection, access to quality essential healthcare services and access to safe, effective, quality and affordable essential medicines and vaccines for all.	Indicator 3.8.1: Proportion of women (aged 15–49 years) who have their need for family planning satisfied with modern methods	Repeat Indicator. Dropped.
	Indicator 3.8.2: Percentage of women (aged 15–49) with a live birth in a given reference period who received ANCs four times or more	Indicator 3.8.2: Mothers who had full antenatal care (at least four antenatal visits, at least one tetanus toxoid injection, and took iron folic acid tablets or syrup for 100 or more days)
	Indicator 3.8.3: Percentage of children (aged 12–23 months) who received three doses of pentavalent vaccine before their first birthday	Indicator 3.8.3: Children (aged 12–23 months) fully immunised (BCG, measles and three doses each of polio and DPT) (%)
	Indicator 3.8.4: Percentage of children with suspected Pneumonia (cough and difficulty breathing, not due to problem in chest or blocked nose) in the two weeks preceding the survey who sought care from appropriate health facility or provider	Indicator 3.8.4: Percentage of children with ARI in the last two weeks preceding the survey (%) who sought care
	Indicator 3.8.5: Percent TB cases treated (cured plus treatment completed) among those notified to the health authorities during a specified period	Indicator 3.8.5: Proportion of people suffering from TB, medically treated.
	Indicator 3.8.6: Percentage of PLHIV currently receiving ART among detected number of children and adults living with HIV	Indicator 3.8.6: Percentage of PLHIV currently receiving ART among detected number of children and adults living with HIV
	Indicator 3.8.7: Percentage of population in malaria-endemic areas who slept under an ITN the previous night and/or percentage of population at risk protected by IRS during a specified time period	Indicator 3.8.7: the proportion of households that have an insecticide treated net
	Indicator 3.8.8: Percentage of population using safely managed drinking water and safely managed sanitation services	Dropped.
	Indicator 3.8.9: Proportion of population (18 years and older) who are currently taking antihypertensive medication among the number of adults (18 years and older) who are taking medication for hypertension with systolic blood pressure >= 140 mmHg, or with diastolic blood pressure >=90mmHg	Indicator 3.8.9: Proportion aged 18 years and above) taking antihypertensive medication among total population (aged 18 years and above) with systolic BP >= 140 mmHg or with diastolic BP >=90mmHg
	Indicator 3.8.10: Proportion of population (18 years and older) who are currently taking medication for diabetes (insulin and glycaemic control pills) among numbers of adults (18 years and older) who are taking medication for diabetes or with fasting plasma glucose >=7.0 mmol/L	Indicator 3.8.10: Of those reporting diabetes, the proportion who seek treatment
	Indicator 3.8.11: Proportion of women (aged 30–49) who reported ever screened for cervical cancer to the total women (aged 30–49) screened for cervical cancer in the last five years	Indicator 3.8.11: Women (aged 30–49 years) who have ever undergone examinations of breast
	Indicator 3.8.12: Age-standardised prevalence of tobacco use among persons aged 15 and above	Repeat Indicator. Dropped.
	Indicator 3.8.13: Number of OPD visits per person per year and IPD admission rate per 100 population per year	Indicator 3.8.13: Number of OPD visits per person per year and IPD admission rate per 100 population per year
	Indicator 3.8.14: Total physician, nurses and midwives per 1,000 population	Indicator 3.8.14: Total physician, nurses and midwives per 1,000 population
	Indicator 3.8.15: Percentage of health facilities with essential medicines and lifesaving commodities	Indicator 3.8.15: Percentage of people reporting ailments who didn’t choose government hospitals, because “required specific services not available” or “available but quality not satisfactory”
	Indicator 3.8.17: Poverty headcount due to OOP expenditure on health	Indicator 3.8.17: Poverty headcount due to OOP expenditure on health
	Indicator 3.8.18: OOPs on healthcare	Indicator 3.8.18: OOPs on healthcare
3.9: By 2030, substantially reduce the number of deaths and illnesses from hazardous chemicals and air, water and soil pollution and contamination.	Indicator 3.9.1: Mortality rate due to household/ambient air pollution	Indicator 3.9.1: Proportion of respondents reporting Asthma
	Indicator 3.9.2: Mortality rate attributed to unsafe water, unsafe sanitation and lack of hygiene	Indicator 3.9.2: Prevalence of diarrhoea (reported) in the last two weeks preceding the survey (%)
	Indicator 3.9.3: Mortality rate attributed to unintentional poisoning	Indicator 3.9.3: Percentage of households reporting morbidity due to poisoning (NSS)
3.a: Strengthen the implementation of the WHO Framework Convention on Tobacco Control in all countries, as appropriate.	Indicator 3.a.1: Age-standardised prevalence of current tobacco use among persons aged 15 years and above	Indicator 3.a.1: Age-standardised prevalence of current tobacco use among persons aged 15 years and above
3.b: Support the research and development of vaccines and medicines for the communicable and non-communicable diseases that primarily affect developing countries, provide access to affordable essential medicines and vaccines, in accordance with the Doha Declaration on the TRIPS Agreement and Public Health, which affirms the right of developing countries to use to the full the provisions in the Agreement on Trade-Related Aspects of Intellectual Property Rights regarding flexibilities to protect public health, and, in particular, provide access to medicines for all.	Indicator 3.b.1: OOP expenditure on health	Repeat Indicator. Dropped.
	No indicator in the NIF	Indicator 3.b.2: Percent of households incurring large expenditure on drugs
3.c: Substantially increase health financing and the recruitment, development, training and retention of the health workforce in developing countries, especially in least developed countries and small island developing states.	Indicator 3.c.1: Total physicians, nurses and midwives per 10,000 population	Repeat Indicator. Dropped.
6.1: By 2030, achieve universal and equitable access to safe and affordable drinking water for all.	Indicator 6.1.1: Proportion of population using safely managed drinking water services	Dropped.
	Indicator 6.1.2: Proportion of population using an improved drinking water by source	Indicator 6.1.2: Households with an improved drinking-water source
6.2: By 2030, achieve access to adequate and equitable sanitation and hygiene for all and end open defecation, paying special attention to the needs of women and girls and those in vulnerable situations	Indicator 6.2.1: Proportion of population using safely managed sanitation services, including a handwashing facility with soap and water	Dropped.
	Indicator 6.2.2: Percentage of population using basic sanitation services	Indicator 6.2.2: Households using improved sanitation facility (%)

In the past, there have been suggestions to integrate existing sample surveys in India coming from the former chairperson of the National Statistical Commission. Making sure that the raw data from surveys are available for analysis is important and the government, if necessary, can leverage the Collection of Statistics Act of 2008.^[46] With universalisation of birth and death registration and better surveillance, many fundamental challenges based on the dependence on sample surveys will disappear. However, in the interim, optimising India’s capacity to track its health and nutrition targets will depend on how the country’s national surveys are streamlined, so they can have the potential to provide robust estimates of population health and nutrition at the district level. The recommendations made in this paper should help harness that potential, and enable regular tracking of district-level health and nutrition indicators.

Endnotes

• Healthcare
• Sustainable Development
• India
• health
• healthcare
• India
• nutrition
• SDGs
• Sustainable development

The views expressed above belong to the author(s). ORF research and analyses now available on Telegram! Click here to access our curated content — blogs, longforms and interviews.