Health data standards in India: ICD-10, LOINC, FHIR, ABDM and OMOP explained

“Is it standards-based?” is a question every hospital asks of its software, and the answer is often a list of acronyms. Each standard does a different job, and knowing which job is the difference between records that can be shared, studied and used by AI, and records that cannot.

This article explains the main ones in plain terms, for hospital IT heads, medical records teams and anyone building or buying health software in India.

The short version

Standard	What it does	Example
ICD-10	Codes diagnoses	Type 2 diabetes is E11 wherever it is recorded
LOINC	Codes lab tests and observations	A haemoglobin from any lab is the same test
FHIR	Formats records for exchange between systems	A lab report sent from the lab to the clinic
ABDM	India’s network for sharing records with consent	A patient’s records reaching a new doctor through their ABHA
OMOP	A common model for analysing data from many sites	One study run across many hospitals

Codes: what a record means

ICD-10 is the World Health Organization’s classification of diseases. A diagnosis coded in ICD-10 means the same thing in every hospital, however the doctor typed it. ICD-10 is still widely used in Indian hospital records. WHO’s successor, ICD-11, came into effect internationally in January 2022, and countries are adopting it gradually.

LOINC codes laboratory tests and clinical observations. It is maintained by the Regenstrief Institute. Without it, “Hb”, “Haemoglobin” and “HGB” from three labs are three different things to a computer; with it, they are one test that can be charted over time.

Clinical terminologies such as SNOMED CT go further, coding findings, procedures and more detail than ICD allows. They matter most where a record needs fine-grained meaning, for example for decision support.

Formats: how a record moves

FHIR (Fast Healthcare Interoperability Resources), from the standards body HL7, defines how health data is packaged and exchanged: a patient, an observation, a diagnostic report, each as a structured “resource” that another system can read. FHIR R4 is the version most widely used.

Older hospital systems often still exchange data using HL7 version 2 messages, especially between laboratory and hospital systems.

India’s network: ABDM

The Ayushman Bharat Digital Mission links a patient’s records across providers using their ABHA number, with the patient’s consent for each exchange. Providers that hold records act as health information providers; those that request them act as health information users; consent is managed separately from both. Records are exchanged in FHIR-based formats. By August 2026 the government reported 96.43 crore ABHA numbers and more than 110 crore health records linked.

Research: OMOP

OMOP is a common data model maintained by the OHDSI community, which describes it as an open standard “designed to standardize the structure and content of observational data.” Data from different hospitals is converted into the same tables and vocabularies, so one analysis can run across all of them, often without patient-level data leaving each site.

It is how large research networks work. Europe’s DARWIN EU, set up by the European Medicines Agency in 2022, has brought in about 40 data partners who convert their data to OMOP. OHDSI also has an India chapter.

FHIR and OMOP are not alternatives. FHIR moves records one patient at a time, for care. OMOP organises many patients’ records, for study. A system that captures coded, structured data can feed both.

What this means for a hospital

• Code at the point of care. Diagnoses chosen from a list that stores ICD-10, lab tests mapped to LOINC. Coding afterwards is slower and less accurate.
• Ask how data leaves. Can the system export and exchange in FHIR? Is it connected to ABDM?
• Validate what you send. A FHIR resource that fails validation is a record another system cannot read.
• Keep the codes with the words. Store the code and the clinician’s text together, so meaning is never lost in either direction.
• Think about research early. If you may one day contribute to research, data captured as coded fields now can be mapped to a model such as OMOP later; free text cannot, at least not reliably.

Where 92health fits

92health records care as structured, coded data: diagnoses chosen from searchable lists and stored with their ICD-10 codes, lab results validated as FHIR R4, hospital feeds over HL7 version 2 where older systems need them, and records linkable to a patient’s ABHA with consent. That is the foundation for sharing records, for data quality and for real-world evidence.

Sources

• OHDSI. Data standardization: the OMOP Common Data Model, and regional chapters.
• European Medicines Agency. DARWIN EU.
• PIB. ABDM factsheet, August 2026.
• Draft Health Data Management Policy, version 2, April 2022 (National Health Authority).