The full report behind every NFHS-5 figure in the editor’s note. It sets out the sample, the questionnaires and the biomarker protocol, including the finger-prick HemoCue Hb 201+ test used for haemoglobin. Read it to see how much the fact sheets have always left to the full report, and what an NFHS-6 national report would need to contain to restore the missing series.
Deep Dive · Health & Data
What NFHS-6 stopped counting
The sixth National Family Health Survey reports 101 indicators for India where the fifth reported 131. Anaemia, child mortality, the sex ratio, literacy and sanitation are among those missing. A reading list on why, on what the government now cites in their place, and on how far those numbers can stand in for the old ones.
India
Health & nutrition
Official statistics
18 readings
IM
ImpactMojo Editorial
Curated by the ImpactMojo team
Written after updating forty-one ImpactMojo pages to NFHS-6, which meant matching the two rounds' fact sheets line by line and finding, indicator by indicator, what the government cites now. Every figure below was checked against the fact sheets, the answers given in the Lok Sabha, or the reports named beside it.
House Pick
Editor's note
The fact sheets of NFHS-6, fielded between May 2023 and December 2024 and released on 29 May 2026, report 101 indicators for India. NFHS-5 reported 131. Matched line by line, 88 of the old indicators continue and 3 continue under a new definition. Pre-school attendance now covers children aged 2 to 4, the hepatitis B measure is now the birth dose, and land ownership, once counted for each woman, is now counted for each household. The remaining 40 have no counterpart. They include all seven anaemia measures, the neonatal, infant and under-five mortality rates, the sex ratio of the population and at birth, the registration of births and deaths, improved sanitation, clean cooking fuel, the literacy of women and men, cancer screening and knowledge of HIV. Thirteen indicators are new. The fact sheets are marked provisional and cover every state and union territory except Manipur.
The government has given its reason in Parliament. Answering Lok Sabha Unstarred Question 2098 on 31 July 2026, the Minister of State for Health said the list was settled on the advice of a Technical Advisory Committee of several ministries, “in line with the Data Harmonization initiatives of the Government of India, so as to avoid inconsistencies arising from fragmented and unharmonized data available across different official data sources.” Anaemia is to come from a survey by the Indian Council of Medical Research that draws venous blood. The infant mortality rate and the sex ratio at birth are to come from the Sample Registration System. At a briefing on 7 June, as reported by DD News and The Tribune, officials said the missing indicators had not been dropped and would appear in the full national report. No date has been given for that report, and a second Lok Sabha question asking whether the indicators were collected at all received no answer on that point.
On anaemia the method argument favours the government. NFHS-5 measured haemoglobin on a drop of capillary blood with a portable HemoCue analyser. The studies in Section 02 find that capillary readings run higher than venous ones, by enough to move a prevalence figure by as much as 20 to 30 percentage points. The decision was reported in May 2023, before NFHS-6 fieldwork began. What followed is harder to defend. The ICMR survey, now named SAMPADA, has finished fieldwork, and an RTI reply of 1 September 2026, reported by The South First, says it has no date for a report. Until it publishes, the most recent measured anaemia figures for India are NFHS-5's, from 2019-21. Two answers given in the Lok Sabha in July 2026 still quoted them: 67.1 per cent of children aged 6 to 59 months and 57 per cent of women aged 15 to 49.
The substitutes measure different things. The Sample Registration System puts the sex ratio at birth at 918 girls per 1,000 boys for 2022-24. The Civil Registration System puts it at 933 for 2024. NFHS-5 gave 929. The NSS Comprehensive Annual Modular Survey of 2022-23 reports improved latrines for 97.8 per cent of households, counting only households with access to a latrine and including shared facilities. NFHS-5's 70.2 per cent counted every household and required a facility not shared with another. The Periodic Labour Force Survey reports literacy for everyone aged seven and above from what the household says. NFHS asked women and men aged 15 to 49 to read a sentence. Each of these is a sound number in its own right. None continues the NFHS series, and a chart that joins one to the other will show a change of method as if it were a change in the country.
One thing in the NFHS-6 fact sheet needs an explanation from IIPS. Its column of NFHS-5 values does not always match the NFHS-5 fact sheet. The second dose of measles-containing vaccine among children aged 24 to 35 months is 31.9 per cent in the NFHS-5 sheet and 58.6 per cent in the NFHS-6 sheet’s comparison column, and no footnote gives a reason. The 58.6 has since been repeated in Parliament. Men who have ever used the internet go from 57.1 to 51.2, and about ten smaller figures differ by a few tenths of a point. Until IIPS explains the differences, take NFHS-5 figures from NFHS-5’s own report.
How to read this list. Four sections. The surveys themselves, with the official reason in the government’s own words. The anaemia question, where the method evidence is strong and points both ways at once. The systems that now carry the other numbers, and how complete they are. Then two explorers on this site that put NFHS-5 and NFHS-6 side by side.
On the citations. Where an official host refused the connection, the same document was read from a named mirror. Press reports are attributed to the outlet and used only for what they report the government or ICMR to have said. We could not reach the Rajya Sabha answer of 28 July 2026 on haemoglobin, the 2024 national infant and neonatal mortality rates in the SRS Statistical Report, or any official reason for dropping literacy, cancer screening, HIV knowledge or the contraceptive method mix, and nothing here depends on them.
Section 01
The surveys and the stated reason
What NFHS-5 measured and how, the external standard for checking whether a renamed indicator is still the same one, and the government’s explanation in its own words.
The numerator, denominator and caveats for each indicator used in DHS-family surveys, of which NFHS has been one. NFHS-6 is the first round IIPS ran without external technical support, so this guide is the outside benchmark for asking whether an indicator with a new name, or an old name and a new comparison value, is still the same measure.
The government’s explanation in full, on one page: the Technical Advisory Committee, the appeal to data harmonisation, the move of anaemia to an ICMR survey using venous blood, and the Sample Registration System for infant mortality and the sex ratio at birth. Read it with Lok Sabha Unstarred Question 1141 of 24 July 2026, which asked whether the missing indicators were collected in NFHS-6 and when they would be released, and whose answer repeats the committee paragraph without addressing either.
The quality checks the IIPS team applies across NFHS rounds, written by the people who run the survey. It is the nearest thing in print to a data-quality appendix for the series, and the place to start before deciding whether a difference between two rounds is a change in the country or in the survey.
Section 02
The anaemia question
Capillary blood, venous blood and the threshold for calling someone anaemic. The evidence that the old method overstated anaemia is strong. None of it argues for a round with no estimate at all.
The review to read first. It finds that “significant differences in Hb concentration between capillary and venous blood samples are common”, with capillary readings usually higher, and works through why, using a survey from Uttar Pradesh among its examples. Kurpad, one of its authors, is also an author of the Indian perspective at the end of this section.
Puts a size on the problem. Capillary HemoCue readings exceeded venous readings by 5.7 g/L in preschool children and 8.8 g/L in non-pregnant women, and correcting for the difference could shift prevalence by up to 20.7 and 28.2 percentage points. That is the scale of change a reader should expect when a venous estimate is set beside NFHS-5.
Matched DHS surveys using capillary blood against venous-blood surveys from the same countries and years and found anaemia estimates 9 to 31 points apart, lower with venous blood each time. The gap is systematic, and it is not peculiar to India.
The current international position: “the best practice is to collect and use venous blood” with an automated haematology analyser. It supports the government’s choice of method. It says nothing in favour of leaving a survey round without an estimate while the venous survey is analysed.
India’s only national venous-blood survey of anaemia in children and adolescents so far, with iron, folate, vitamin B12 and inflammation markers alongside haemoglobin. It found anaemia in 41 per cent of children aged 1 to 4, 24 per cent of those aged 5 to 9 and 28 per cent of adolescents. It covers ages 1 to 19 only, so for women and men there is still no venous estimate at all.
The second lever on the headline number, separate from the method. Using the healthy children in CNNS, it derives Indian cut-offs 1 to 2 g/dL below WHO’s, which would take prevalence in that sample from 30.0 to 10.8 per cent. Read it so that a change of threshold and a change of blood sample are never mistaken for each other, or for a change in how many people are anaemic.
The policy argument that came before the decision. NFHS-5’s 67.1 per cent of children “might be an overestimate”, the authors write, against CNNS’s 30.7 per cent, of which a third was due to iron deficiency, and iron-centred programmes may be aimed at the wrong target. Judge it against the four papers above, and note that its argument is for measuring better, which is a different thing from measuring later.
Section 03
Where the other numbers come from now
Mortality, registration and sanitation now come from registration systems and other surveys. These readings show how those systems work and where their coverage is thin.
The clearest published account of how the Sample Registration System works: its frame of about 1.1 million households in 1998-2003 and 1.3 million in 2004-14, and the verbal autopsies built on top of it. SRS now carries the infant mortality rate and the sex ratio at birth that NFHS-6 no longer reports. Its estimates are annual and cover a sample frame. NFHS’s were for the five years before each survey and came from birth histories, so the two series should be quoted side by side and never joined.
The source the government points to for birth and death registration. It reports registration levels of 99.1 per cent for births and 99.4 per cent for deaths in 2024. Read the glossary before quoting either. The level is registered events divided by events the SRS estimates should have happened, so it measures completeness against a model, where NFHS asked whether each young child’s birth had been registered (89.1 per cent in NFHS-5). The same report finds females to be 39.6 per cent of registered deaths.
The baseline for the 2024 claim. National completeness of death registration was 76.6 per cent in 2015, Uttar Pradesh’s was 44.2 per cent, and age was missing from about 37 per cent of records. A rise to 99.4 per cent in nine years is possible. It is also the kind of change to check before relying on it.
Compares SRS and CRS life tables for 2017. Child deaths are “uniformly under-reported” in CRS, as are female deaths in many states, and SRS probabilities are of limited precision for tracking trends. Neither system replaces the other, and neither replaces a household survey that asks mothers about every child they have borne.
Asking about each household member recorded 20 to 21 points more open defecation than asking one question about the household, and 28 to 29 points more in households that had received help to build a latrine. Read it before treating any household-level sanitation figure, the Swachh Bharat Mission’s declarations of open-defecation-free villages included, as a stand-in for the NFHS measure that has gone.
Section 04
Practice
Two explorers on this site that set NFHS-5 and NFHS-6 side by side, for the indicators both rounds report.
State figures with rural and urban areas shown separately. Use it to see which of the continuing indicators moved between rounds, and how much of a state’s change sits in its villages or its towns.
District maps across rounds. The indicators NFHS-6 no longer reports stop at NFHS-5 here, and the gap on the map is the subject of this list.