LOINC (Logical Observation Identifiers Names and Codes)

One-sentence definition: LOINC is a universal, freely available coding standard developed by the Regenstrief Institute for identifying medical laboratory observations, clinical measurements, survey instruments, and diagnostic study results to enable consistent data exchange across healthcare systems.

Full Definition

LOINC is a vocabulary of more than 100,000 codes that identify observations — lab results, vital signs, clinical assessments, questionnaires, and diagnostic study types. Each LOINC code names a specific type of measurement or finding in a way that any system can recognize, regardless of the local codes a lab or EHR uses internally.

Developed and maintained by the Regenstrief Institute in Indianapolis, LOINC is free to download and use in commercial products without royalties. It has been in continuous development since 1994 and is now a regulatory requirement in the United States for lab result exchange.

LOINC standardizes the what was measured — the identity of the observation itself. It is distinct from SNOMED CT, which captures clinical findings and diagnoses (the what was found), and from ICD-10, which encodes diagnoses for billing and administrative reporting. In practice, these terminologies work together: LOINC identifies what test was run, SNOMED CT can describe the coded result value, and ICD-10 captures the resulting diagnosis.

LOINC was created in 1994 to solve a specific problem: labs were transmitting results in proprietary, institution-specific codes that receiving systems couldn’t interpret without manual lookup tables. A CBC panel from one hospital used different codes than the same panel from a neighboring hospital. Regenstrief’s solution was to assign each distinct observation a universal, stable identifier — one that would mean the same thing everywhere.

Context and Usage

Where This Term Appears

LOINC appears throughout the healthcare data exchange stack:

• FHIR resources: Observation.code, DiagnosticReport.code, Questionnaire.item.code
• Lab order and result interfaces: LIS-to-EHR integrations, lab order entry systems
• Public health reporting: electronic case reporting (eCR), syndromic surveillance
• Bulk Data exports: FHIR Bulk Data and analytics pipelines for population health
• Quality measures: CMS eCQMs, HEDIS measures, and clinical decision support rules
• US Core and FHIR implementation guides: required for vital signs, lab results, and clinical assessments

Common Usage Examples

In conversation: “We map all vitals to LOINC codes before sending to the HIE — otherwise the receiving system has no idea what VIT-HR-001 means.”

In documentation: “The Observation resource must include a LOINC code in Observation.code for the result to pass US Core validation.”

In technical contexts:

"code": {
  "coding": [
    {
      "system": "http://loinc.org",
      "code": "8867-4",
      "display": "Heart rate"
    }
  ]
}

Why LOINC Exists

Before LOINC, labs transmitted results using whatever codes their local systems happened to use. A serum glucose result from one hospital arrived as GLU-S; the same test from another arrived as 0001-GLU. The receiving EHR had no reliable way to know these were the same observation without a custom mapping table for each sender.

Regenstrief Institute created LOINC in 1994 to give every distinct type of observation a single, universally recognized identifier. Rather than a simple flat list, LOINC encodes the precise nature of each measurement through a structured six-part naming convention — so the code itself carries information about what was measured, how, in what specimen, and at what scale.

HL7 adopted LOINC as a preferred terminology for observations early on. US federal programs (Meaningful Use, then Promoting Interoperability) mandated it for lab result exchange. FHIR’s core specification references it throughout, and US Core profiles require LOINC for vital signs, lab results, and clinical assessments. ONC certification criteria and CMS quality reporting programs are all anchored to it.

Key Characteristics

Six-Part Structure

Every LOINC long name encodes six axes of information in a colon-separated format. This structure is what makes LOINC codes precise and unambiguous:

1. Component — the analyte or what is being measured (e.g., Glucose)
2. Property — the characteristic of the component being quantified (e.g., MCnc = mass concentration)
3. Time aspect — point in time vs. a duration (e.g., Pt = point in time, 24H = 24-hour collection)
4. System — the specimen or body system (e.g., Ser/Plas = serum or plasma, Urine, Blood)
5. Scale type — how the result is expressed (e.g., Qn = quantitative, Ord = ordinal, Nom = nominal, Nar = narrative)
6. Method — how it was measured, when it matters (e.g., Immunoassay, Calculated); optional

The LOINC long name for serum glucose (2345-7) is Glucose:MCnc:Pt:Ser/Plas:Qn — glucose measured as a mass concentration at a point in time in serum or plasma, expressed as a quantitative result. The method is unspecified, so it applies regardless of analyzer.

Code Types (Lab, Clinical, Survey)

LOINC covers far more than laboratory tests:

• Laboratory — the largest category: clinical chemistry, hematology, coagulation, urinalysis, microbiology, serology, toxicology, and more
• Clinical — vital signs (heart rate 8867-4, systolic blood pressure 8480-6, body weight 29463-7), physical exam findings, clinical assessments, imaging study types
• Survey and questionnaire instruments — standardized tools like the PHQ-9 depression screen, PROMIS scales, CAGE alcohol assessment, Morse Fall Scale, and hundreds of others; each instrument and each individual item within it has a LOINC code

Roughly 30% of LOINC codes are outside the laboratory category. Teams building vital signs pipelines, questionnaire-based workflows, or clinical assessment tools all need LOINC, not just teams working on lab interfaces.

Naming Convention

The LOINC long name encodes all six axes joined by colons. LOINC also provides:

• Short name — a condensed label suitable for column headers and constrained UI displays
• Display name — a more readable label intended for end-user interfaces

LOINC codes are numeric with a check digit (e.g., 2345-7). Once a code is assigned it is never changed. If a concept is retired or split, the original code is marked inactive but remains in the database and is never reused. This stability is essential for longitudinal data — a code recorded in a patient chart five years ago means the same thing today.

Technical Considerations

Free and Open Access

LOINC is free to download and use. A free registration at loinc.org is required to download the full table, but there are no licensing fees — use in commercial products is permitted without royalties. This stands in contrast to some other clinical terminologies that require paid licenses for commercial use.

The full LOINC table is distributed as CSV and Excel. LOINC is also accessible programmatically via any FHIR terminology server that supports the $lookup operation against http://loinc.org, making it straightforward to look up codes and display names at runtime.

Version Updates

LOINC releases two updates per year, typically in February and August. Each release adds new codes and may mark existing codes as deprecated (inactive). Deprecated codes are never deleted or reassigned — a code that was valid in 2010 remains in the table, marked inactive, so historical data remains interpretable.

Most clinical systems are tolerant of LOINC version differences for common codes — the meaning of 8867-4 (heart rate) does not change between releases. For more specialized use cases, such as regulatory reporting or quality measure validation, it is good practice to document which LOINC version a mapping set was validated against.

Implementation details: For how LOINC is used in specific FHIR resources (Observation coding, DiagnosticReport integration, Questionnaire instruments, and US Core conformance requirements), see the canonical reference → FHIR Terminology.

Relationship to Other Terms

Related Terms

• FHIR — LOINC is the required observation coding system throughout FHIR and its US implementation guides
• SNOMED CT — clinical findings, diagnoses, and coded result values; often paired with LOINC
• RxNorm — medication coding; the LOINC counterpart for drug identification
• ICD-10 — diagnosis coding for billing and administrative reporting

Contrasting Terms

• LOINC vs SNOMED CT: LOINC identifies what was measured (the observation type); SNOMED CT identifies what was found (clinical findings, result concepts, diagnoses). They are complementary. In a FHIR Observation, Observation.code uses LOINC; Observation.valueCodeableConcept may use SNOMED CT. For example, LOINC 2339-0 (glucose) with a SNOMED result of 166928006 (elevated).

• LOINC vs CPT: CPT codes are used for billing — they capture the performance of a test (procedure codes for insurance reimbursement). LOINC codes identify the result of the test for clinical data exchange. A single CPT code for a metabolic panel maps to many LOINC codes for the individual results within it.

• LOINC vs ICD-10: ICD-10 encodes diagnoses that appear on claims and problem lists. LOINC encodes the observations — lab results, vital signs, assessments — that clinicians use to arrive at those diagnoses. They operate at different layers of the clinical workflow.

Common Misconceptions

Misconception 1: LOINC is Only for Lab Tests

• Incorrect belief: LOINC only matters for laboratory interfaces and can be ignored when building vital signs monitoring, questionnaire tools, or clinical assessment workflows.
• Reality: Approximately 30% of LOINC codes cover clinical observations, vital signs, survey instruments, clinical documents, and radiology study types — not lab tests.
• Why it matters: US Core vital signs profiles require LOINC codes. PHQ-9, PROMIS, and other standardized instruments have LOINC codes for the panel and each item. Teams building anything beyond pure lab interfaces still need to handle LOINC correctly to pass validation and achieve interoperability.

Misconception 2: LOINC Codes Contain Units

• Incorrect belief: Different units for the same measurement require different LOINC codes — for example, glucose in mg/dL vs mmol/L must be different codes.
• Reality: LOINC codes identify the type of measurement (the property axis encodes “mass concentration”), not the specific unit. Units are carried separately in Observation.valueQuantity.unit.
• Why it matters: 2345-7 (serum glucose, quantitative) is the correct code whether the reporting lab uses mg/dL or mmol/L. Using separate codes for units is a mapping error that breaks aggregation and decision support rules. The unit is data in the observation payload, not part of the observation identity.

Why LOINC Matters

Without LOINC, lab results arrive in provider-specific codes that downstream systems cannot interpret without custom per-sender translation tables. Every new data feed requires a new mapping effort. With LOINC, the meaning of an observation travels with the data.

LOINC enables aggregation at scale. Population health analytics, quality measures, clinical decision support rules, and disease registries all depend on being able to query for a specific observation type across patients and organizations. That is only possible when the same observation type consistently uses the same code.

Regulatory requirements reinforce this: USCDI v3+ requires LOINC for laboratory results. ONC certification criteria reference it. CMS quality reporting programs depend on it. FHIR implementation guides — US Core, QI-Core, mCODE — mandate LOINC for specific resources and profiles.

For teams building FHIR-based integrations, LOINC is not optional. It is the shared vocabulary that makes the data meaningful outside the system that produced it.

Cross-References

Related Glossary Terms

• FHIR — The data exchange standard that uses LOINC as its primary observation coding system
• SNOMED CT — Clinical terminology paired with LOINC for coded result values

---

Last reviewed: February 15, 2026 Definition authority: Regenstrief Institute Content status: Canonical reference