Last week’s release of early results from the sixth National Family Health Survey (NFHS) made headlines, not just for its main findings, but also for information that the NFHS no longer reports. The NFHS factsheet carried 101 indicators where the previous round had 131. Gone was the prevalence of anaemia — the condition that affects roughly two-thirds of India’s children and more than half its pregnant women, by the last count — as well as infant and child mortality rates, the sex ratio at birth, and the share of households cooking with clean fuel.
Some of these changes were defensible. Anaemia in past rounds was measured by pricking a finger and reading a drop of capillary blood, a method that is convenient in field surveys but likely overstates prevalence compared to the venous-blood standard recommended by the World Health Organization. Anaemia estimation, according to the ministry of health and family welfare, will now be part of the Indian Council of Medical Research’s Diet and Biomarkers Survey. The Ministry also noted that mortality rates are already measured as part of the Sample Registration System (SRS). However, the timing of these changes, which includes removing the anaemia figures that called the effectiveness of the Anaemia Mukt Bharat campaign into question, and the clean-fuel data that undercut the Ujjwala scheme, is why critics smell a coverup.
The critics and defenders of the changes in reporting both implicitly assume that the way India uncovers the most basic facts about its people’s health should be through large, periodic, sample-based surveys, and that the question under discussion is simply about which indicators should be included. That may the wrong question. Perhaps we should be asking why we need the NFHS or the SRS at all.
As with many things about modern India’s public health, the reliance on surveys to measure routine health indicators originates in colonial India. The imperial administration invested heavily in the decennial census, systematised from 1881, because population counts served revenue, representation, and the classification of subjects. It invested almost nothing in civil registration — the continuous recording of individual births and deaths — because registering them had no value to the British. Although the Births, Deaths and Marriages Registration Act of 1886 was passed, it was largely flouted, whereas census surveys were carried out promptly and in great detail. My own familiarity with these comes from reading the reports of my great-grandfather, who served as the census commissioner of the state of Travancore in the early 1900s. Those reports were more than simple population counts — they covered the cultural, linguistic, and anthropological story of those times.
As a result, independent India inherited a formidable machinery for counting the living once a decade, but with a near-absent record of how they were born, sickened, and died. Sadly, that asymmetry has never been corrected. Although the proportion of deaths registered is now greater than ninety percent, even now barely a fifth of registered deaths carry a medically certified cause.
Rather than complete the continuous record, India has relied on sample surveys to cover the gaps. The Sample Registration System, initiated in 1969, exists to estimate the birth rates, death rates, and infant and child mortality that a functioning registration system would generate automatically. The NFHS, on the other hand, measures nutrition, anaemia, contraceptive use, immunisation coverage, antenatal care, and sanitation — indicators that cannot come from any register, and can be captured only by visiting a household. For that information, a survey is the right way to collect data. But India also leans on the NFHS for vital events — infant and child mortality, the sex ratio at birth — which should instead come from routine registration.
It is not that rich countries don’t use surveys. The United Kingdom, which has excellent birth and death registration — the kind it never built for India — runs the Health Survey for England every year. The United States has long used its National Health and Nutrition Examination Survey, a model on which the NFHS was based. Population anaemia, in fact, is measured in these countries through dedicated biomarker surveys, much as India now proposes. However, these surveys supplement a routine system rather than trying to replace it.
The United Kingdom has no population register at all, yet recording a death is a legal duty — the record fed directly to the Office for National Statistics, which publishes mortality counts weekly, eleven days after the period. Causes are coded to international standards and linked to hospital records, so morbidity and death can be read together in near real-time. Sweden uses a personal identity number linked to parish registers centuries older, against which births, deaths, diagnoses, and cancers are recorded as they occur. Neither country would use a household survey to discover its infant mortality rate or its sex ratio at birth, because those facts are collected routinely, as they happen.
A more relevant comparison is China, which started out poor and populous like India and created the Disease Surveillance Points system, a sample-based interim, in 1980. That system was plagued by the same under-reporting that afflicts India’s. But, since 2013 China has merged its overlapping sample systems into an integrated national mortality surveillance system, explicitly to build toward comprehensive vital registration and provincially representative cause-of-death data. China has treated its survey as a bridge it intended to cross, whereas India has treated the NFHS and SRS as permanent.
It is unfortunate that a nation of 1.4 billion has to depend on periodic surveys to know how many of its infants die and how many of its girls are born. The colonial state did not build the continuous record because the people it recorded were subjects, not citizens. Independent India must correct that error. We should run health surveys, as every developed country does, to measure the nutrition, behaviors, and exposures no register can capture. But the unfinished business of Viksit Bharat is to make the registration of every birth, death, and cause so complete and routine that the most basic facts of a citizen’s life and death are never again something we have to go looking for every five years.
Ramanan Laxminarayan is president, One Health Trust. The views expressed are personal
