Data Sources for Mortality and Morbidity Measures: SRS, CRS, and More

Whenever a government talks about infant mortality dropping, life expectancy rising, or a state having high disease burden, there is a quiet army of data systems doing the heavy lifting behind those numbers. Mortality and morbidity measures don’t appear out of thin air. They come from a network of registrations, surveys, and censuses, each with its own purpose, strengths, and blind spots. Understanding where these numbers come from is the first step to using them responsibly, whether you’re writing a research paper, planning a health intervention, or simply trying to read a news headline critically.

Why we need multiple data sources
Vital registration systems
The Civil Registration System (CRS)
The Sample Registration System (SRS)
Annual Health Surveys (AHS)
Why district-level data matters
Census and National Sample Survey
The Census of India
The National Sample Survey (NSSO)
How these sources fit together

Why we need multiple data sources

Mortality data tells us how many people die, when, and from what. Morbidity data tells us who is sick, with which conditions, and how often. Both are essential for public health planning, but no single system can capture all of this perfectly. A village in interior Odisha and a metropolitan ward in Mumbai have very different realities, and a one-size-fits-all approach would miss crucial details.

That’s why a layered system has evolved. Some sources record events continuously, like a flowing river. Others take periodic snapshots through large surveys. A few zoom in on specific districts or population groups. Together, they form a fairly complete, though imperfect, picture of population health.

Vital registration systems

Vital registration is the bedrock of demographic data. It refers to the continuous recording of vital events, mainly births and deaths, as they happen. Two systems work in parallel here, and they are often confused even by students of population studies.

The Civil Registration System (CRS)

The Civil Registration System is the official, legally mandated mechanism for recording every birth and death that occurs in the country. It was formalised under the Registration of Births and Deaths Act, 1969, which made registration compulsory and aimed to standardise the process across all states and union territories.

The CRS works through a three-tier structure. The Registrar General at the central level coordinates the system, Chief Registrars run things at the state level, and local registrars at the district and panchayat or municipal level handle day-to-day registration. When a baby is born in a hospital or a death occurs at home, family members or institution heads are required to notify the local registrar within 21 days. The certificate that is issued isn’t just a piece of paper. It is the foundation for legal identity, school admissions, inheritance claims, insurance settlements, and access to government schemes.

The strength of the CRS is its individual-level, legally valid coverage. The weakness, historically, has been completeness. Estimated birth registration stood around 88 percent and death registration at roughly 77 percent in recent assessments, with sharper gaps in rural and remote areas. States like Kerala and Tamil Nadu achieve near-complete registration, while several northeastern and central states lag behind.

Recent modernisation efforts, including online registration portals, mobile applications, and integration with the Medical Certification of Cause of Death scheme, are pushing the CRS toward becoming a real-time, reliable source of mortality data. Some researchers argue that the CRS is now more reliable than the SRS for measuring adult mortality in several states, which signals a slow but important shift.

The Sample Registration System (SRS)

While the CRS was being built up over decades, the country needed reliable vital statistics immediately. The Sample Registration System was set up to fill that gap. The SRS was piloted in 1964-65 and became fully operational in 1969-70, designed as an interim measure but eventually becoming one of the most trusted sources of demographic data in the country.

The SRS works on a clever dual-recording principle. In a representative sample of locations, a local part-time enumerator, often an Anganwadi worker or school teacher, continuously records every birth and death as it occurs. Independently, every six months, an SRS supervisor visits the same households and collects the same information afresh. The two sets of records are then matched and reconciled. Any discrepancies trigger a re-verification visit. By 2023, the SRS sample had grown to 8,839 units covering more than 8.8 million people, which makes it one of the largest demographic surveys in the world.

The indicators that come out of the SRS are the ones you see quoted everywhere: crude birth rate, crude death rate, infant mortality rate, neonatal mortality rate, under-five mortality rate, maternal mortality ratio, and total fertility rate. Since 1999, the SRS has also integrated a verbal autopsy component, where trained surveyors interview family members of the deceased to determine the probable cause of death. This Million Death Study collaboration has produced landmark estimates on causes of death, especially highlighting the growing burden of non-communicable diseases.

The big advantage of the SRS is its dual methodology, which reduces the omission errors common to single-source data. The limitation is that estimates are reliable mainly at the state level. You cannot, for instance, pull out a credible infant mortality rate for a single district from SRS data alone. That gap is where another survey steps in.

Annual Health Surveys (AHS)

The Annual Health Survey was born from a recognition that national and state averages can hide enormous internal variation. A state may look reasonable on paper, but specific districts within it might have child mortality rates comparable to sub-Saharan Africa. To plan targeted interventions, you need district-level data.

Conducted by the Office of the Registrar General of India during the 11th Five-Year Plan period, the AHS covered all 284 districts of the eight Empowered Action Group (EAG) states and Assam. The EAG states, Bihar, Chhattisgarh, Jharkhand, Madhya Pradesh, Odisha, Rajasthan, Uttarakhand, and Uttar Pradesh, were identified as having particularly poor maternal and child health indicators and high fertility rates. Together with Assam, they account for nearly half of the country’s population.

The scale of the AHS was unprecedented. The baseline sampled about 18.2 million people across 3.6 million households, with district-level sample sizes calibrated to each district’s expected infant mortality rate. For the first time, India had annual, district-level estimates of indicators like infant mortality rate, under-five mortality rate, neonatal mortality rate, and post-neonatal mortality rate in the very regions where child survival was most precarious.

Why district-level data matters

The AHS revealed sharp inequalities that were previously invisible. Districts in Madhya Pradesh, Chhattisgarh, Jharkhand, Odisha, and Rajasthan showed extremely high neonatal mortality rates, with some districts reporting more than 2,000 neonatal deaths per 100,000 live births. These granular numbers allowed planners to direct resources, train more frontline workers, and design district-specific interventions instead of relying on broad-brush state programmes.

The AHS also covered other dimensions of health, including disability, fertility behaviour, and access to maternal care, which made it a valuable cross-sectional resource. Limitations include that the survey was confined to nine high-focus states and was discontinued after a few rounds, with subsequent district-level information being drawn from the Health Management Information System and the National Family Health Survey.

Census and National Sample Survey

Beyond births and deaths, we need to understand how sick a population is, what kinds of conditions people live with, and how often they use health services. This is where the Census and the National Sample Survey come in.

The Census of India

The Census is the largest data collection exercise in the country, conducted once every ten years. While it is best known for population counts, it also captures demographic, social, and economic data including disability. The 2011 Census collected information on eight categories of disability: seeing, hearing, speech, movement, mental retardation, mental illness, any other disability, and multiple disabilities. Categories like mental illness were added for the first time, and earlier categories were redefined to improve reliability.

The strength of the Census is its complete coverage. Because it surveys everyone, disability prevalence can be calculated even for small districts and sub-populations. The weakness is the medical and somewhat subjective definition used. Enumerators don’t conduct medical examinations, so reporting depends on respondents’ perception. As a result, Census figures on disability tend to be lower than those generated by detailed surveys, which ask more screening questions. The 2001 Census estimated about 11.8 million people with disabilities, while the corresponding NSS estimate was 26.5 million, a stark gap that illustrates how methodology shapes results.

The National Sample Survey (NSSO)

The National Sample Survey Office, now part of the National Statistical Office, has been a workhorse for socio-economic data since 1950. Its health surveys, conducted periodically since 1953-54, are one of the primary sources of quantitative data on the health sector in the country.

The 52nd round in 1995-96, the 60th round in 2004, the 71st round in 2014, and the 75th round in 2017-18 are particularly important reference points. The 75th round covered 1,13,823 households and over 5.55 lakh persons, generating detailed information on morbidity, hospitalisation, maternity and childbirth, and the condition of the elderly. The 76th round in 2018 focused specifically on persons with disabilities and provided rich data on disability prevalence, support services, and rehabilitation access.

What does the NSSO measure that other sources miss? Several things. Self-reported ailments in a 15-day reference period give us morbidity prevalence rates. Hospitalisation patterns over a one-year recall period reveal what proportion of people sought inpatient care, in which type of facility, and at what cost. Out-of-pocket expenditure on health, broken down by medicines, consultation, investigation, and transport, helps quantify the financial burden of illness on households. The 75th round, for instance, showed that private hospitals accounted for 55 percent of inpatient hospitalisation, with public hospitals handling 42 percent, a finding with direct implications for health financing policy.

The NSSO has also enabled long-term analysis of India’s epidemiological transition. By comparing morbidity patterns across rounds, researchers have shown how the disease burden is shifting from infectious diseases toward non-communicable diseases, a transition that varies dramatically across states and socio-economic groups.

How these sources fit together

Each source answers a different question. The CRS gives legally valid individual records and is becoming increasingly reliable. The SRS provides robust, comparable estimates of fertility and mortality at the national and state level. The AHS filled a critical gap by offering district-level data in the most vulnerable regions. The Census provides the broadest possible coverage of disability and population characteristics. The NSSO digs deep into morbidity, hospitalisation, healthcare utilisation, and out-of-pocket spending.

A good population studies researcher learns to triangulate. If SRS estimates of infant mortality differ from AHS district averages, that’s not a problem to be hidden; it’s a clue that points to genuine variation or methodological differences. If Census disability figures don’t match NSSO estimates, the gap itself tells a story about definitions, screening questions, and reporting culture.

The trend in recent years has been toward digital integration, real-time reporting, and linking these systems with administrative data like the Health Management Information System and Ayushman Bharat Digital Mission records. As these come together, we may eventually have a far more responsive and detailed picture of population health than ever before.

What do you think? If you had to design a single new data system that would combine the strengths of the SRS, CRS, AHS, Census, and NSSO without inheriting their weaknesses, what would it look like? And which of these sources do you think gives the most honest picture of health inequalities across regions?

References

PDF 📄

Data Sources for Mortality and Morbidity Measures: SRS, CRS, and More

Table of Contents

Why we need multiple data sources 🔗

Vital registration systems 🔗

The Civil Registration System (CRS) 🔗

The Sample Registration System (SRS) 🔗

Annual Health Surveys (AHS) 🔗

Why district-level data matters 🔗

Census and National Sample Survey 🔗

The Census of India 🔗

The National Sample Survey (NSSO) 🔗

How these sources fit together 🔗