In clinical trials, laboratory tests and observations need to be identified consistently across systems to support reliable analysis and regulatory submissions. LOINC Codes, short for Logical Observation Identifiers Names and Codes, provide a standardised way to represent these observations, particularly in laboratory data.
LOINC is a universal coding system used to assign unique identifiers to health measurements, tests, and clinical observations. This helps create a more consistent reference point when data moves between different systems and sources.
What are LOINC codes used for?
LOINC codes are used to provide a consistent and standardised way to identify laboratory tests and clinical observations across different settings. Their core purpose is to ensure that each test result is unambiguously understood, regardless of how it is labelled within a local laboratory or system.
In laboratory data, a test name alone is often not sufficient to fully describe what has been measured. LOINC codes address this by capturing the full context of a result, including the analyte, specimen type, and measurement characteristics. This allows a single, well-defined concept to represent a test in a way that remains stable across different environments.
This approach supports consistency across:
Labs
Different laboratories frequently use local naming conventions for the same test. LOINC codes align these variations under a single standard identifier, reducing confusion and the need for manual interpretation.
Systems
As data moves between laboratory information systems, electronic health records, and clinical trial databases, LOINC ensures that the meaning of each observation is preserved without relying on local labels.
Studies
In multi-centre clinical trials, lab data is often collected from multiple sources. LOINC enables this data to be harmonised, supporting cleaner aggregation, comparison, and analysis.
In practice, LOINC codes are widely used to support interoperability, allowing systems to exchange and interpret laboratory data reliably. Within clinical workflows, this reduces the risk of misinterpretation and limits the need for repeated data mapping. In research and clinical development, it also helps ensure that laboratory results can be integrated and analysed consistently, particularly when preparing standardised datasets for downstream use.
LOINC codes in SDTM: how does one code relate to multiple variables?
In the Study Data Tabulation Model (SDTM), laboratory data is structured across several variables, each capturing a different aspect of a test and its result. A single LOINC code, however, represents a fully defined observation concept. This creates a one-to-many relationship, where one LOINC code aligns with multiple SDTM variables that together describe that same test.
For example, a laboratory result in the LB domain may include variables such as:
- LBTEST / LBTESTCD to describe the test name
- LBORRES for the original result
- LBORRESU for units
- LBSPEC for specimen type
While these are stored separately in SDTM, a single LOINC code encapsulates the combined meaning of these elements, including the analyte, property, timing, system (specimen), scale, and sometimes method.
This reflects the concept of pre-coordination. In LOINC, multiple attributes of a laboratory observation are predefined and bundled into one code, rather than being constructed dynamically from separate fields. In contrast, SDTM represents these attributes across multiple variables in a more decomposed format.
Understanding this relationship is important in practice. LOINC codes provide a standardised reference that aligns with the combined meaning of several SDTM variables taken together. In many implementations, LOINC is used to support mapping alongside LB variables, and may be included as a supplemental or mapped standard reference rather than a primary SDTM variable.
This has practical implications for:
Standardisation
Using LOINC alongside SDTM helps ensure that laboratory tests are consistently defined across datasets, even when source data varies between sites or labs.
Submission readiness
For regulatory submissions, aligning SDTM laboratory data with recognised standards such as LOINC can support clearer interpretation and improved traceability, particularly when integrating data from multiple studies or sources.
From a data management perspective, this highlights the importance of careful mapping. Assigning a LOINC code requires consideration of multiple SDTM fields together, rather than relying on a single variable. Where source data is incomplete or ambiguous, forcing a mapping can introduce inconsistencies, so a balanced and well-documented approach is preferred.
How do LOINC codes work in practice?
A practical way to understand LOINC codes is to look at a simple laboratory test and how it is represented across systems. Consider a routine blood glucose test. Different laboratories might label this test in slightly different ways, such as ‘Glucose’, ‘Blood sugar’, or ‘Serum glucose level’. While the wording varies, the underlying measurement is the same.
A LOINC code provides a single, standard identifier for that specific observation. For example, a fasting blood glucose measured in serum might be mapped to a LOINC code such as 1558-6, which reflects not just the analyte (glucose), but also the specimen type (serum), the property being measured (substance concentration), and the scale of the result.
In practice, this creates a clear mapping:
-
A local lab test name (e.g. 'Serum glucose')
- A corresponding LOINC code representing the fully defined observation
- The result value and unit, which are interpreted consistently using that standard definition
This mapping supports interpretation by removing ambiguity. Even if two labs use different names or formats, the shared LOINC code ensures that both results refer to the same type of measurement. This becomes particularly important when data is pooled across sites or studies, where small differences in naming could otherwise lead to misclassification or additional reconciliation work.
It also supports standardisation in downstream processes. Once a test is linked to a LOINC code, it becomes easier to integrate that data into clinical datasets, compare results across populations, and maintain consistency during analysis and reporting.
Importantly, the value of LOINC in this context is providing a stable reference point behind it. This allows organisations to preserve their existing workflows while still enabling consistent interpretation and data exchange across systems.
What does a LOINC code look like?
At a surface level, a LOINC code appears as a simple numeric identifier followed by a hyphen and a check digit, for example: 1234-5. This format is designed to be concise and system-friendly, allowing the code to be easily stored, transmitted, and referenced across different platforms.
However, the visible code itself is only a label. The identifier itself does not encode the test meaning. The real meaning of a LOINC code comes from the set of underlying attributes that define the observation it represents. Each code is defined by a combination of standardised components, including:
-
What is being measured (e.g. glucose)
- The property being assessed (e.g. substance concentration)
- The timing or aspect of the observation, where relevant
- The specimen or sample type (e.g. serum or urine)
- The scale of the result (e.g. quantitative or ordinal)
- The method used, where that matters to the definition
These attributes are not always visible in the code itself but are defined in the LOINC database. Together, they form a precise and structured description of a laboratory test or clinical observation
This distinction is important in practice. While the code (e.g., 1234-5) is what appears in datasets and systems, it is the underlying definition that ensures consistent interpretation. Two tests may have similar names locally, but if their underlying attributes differ, they will be assigned different LOINC codes.
In this way, the LOINC format separates the simplicity of a compact identifier from the complexity of a fully defined observation, allowing both efficient data exchange and precise standardisation.
LOINC code implementation: source and maintenance
In practice, LOINC codes are sourced from upstream systems, most commonly from the performing laboratory or laboratory information system (LIS). Many central labs already assign LOINC codes to their tests as part of their standard processes, meaning these codes can often be carried through into clinical datasets rather than derived retrospectively.
This source-driven approach helps preserve the original intent of the test definition. Because LOINC codes are based on specific attributes such as analyte, specimen, and measurement type, assigning them at the point of test definition or result generation tends to be more reliable than attempting to map them later using limited or inconsistent metadata.
Where LOINC codes are not provided by the source, they may need to be mapped during data processing. This requires careful review of available test information to ensure that the selected code accurately reflects the observation. Incomplete or ambiguous source data can make this process challenging, and in such cases, it is generally preferable to avoid forcing a mapping that may introduce inaccuracies.
LOINC itself is a maintained external standard. It is developed and updated regularly by the Regenstrief Institute, with new codes added and existing definitions refined over time to reflect evolving clinical practice. As a result, implementation involves not just selecting codes, but also managing the below factors:
Version control
Ensuring that the version of LOINC used is documented and applied consistently within a study or programme.
Updates and changes
Understanding that new tests may require newer codes, and that definitions can be clarified or expanded across versions.
Traceability
Maintaining clear documentation of how and where codes were sourced or assigned.
From a clinical data management perspective, this means LOINC should be treated as a dynamic reference standard. Organisations typically establish internal guidance on how to handle versioning, when to accept source-provided codes, and how to approach mapping where needed.
Overall, effective implementation relies on using source-provided LOINC codes where available, applying careful and well-documented mapping when necessary, and maintaining awareness of the standard’s ongoing updates. This helps support consistency, traceability, and reliability in laboratory data across systems and studies.
Where do LOINC codes fit with other standards?
LOINC codes are one part of a broader ecosystem of clinical data standards, each designed for a different purpose. Understanding how LOINC fits alongside other standards helps clarify its role in both clinical practice and clinical research.
At a high level, LOINC focuses on identifying laboratory tests and clinical observations. It answers the question: what was measured? This makes it particularly relevant for laboratory data, vital signs, and other structured observations.
In contrast, CDISC terminology is used within clinical trial standards such as SDTM to ensure consistency in how data is structured and labelled. For example, CDISC defines controlled terms for variables like LBTESTCD or units, helping standardise how data is organised within a dataset. While LOINC defines the observation itself, CDISC terminology supports how that observation is represented within a regulatory data model.
ICD (International Classification of Diseases) serves a different purpose again. It is used to classify diagnoses and medical conditions, rather than laboratory measurements. Where LOINC answers what test was performed, ICD answers what condition was identified or treated. This distinction is important, as the two systems are not interchangeable but can be complementary within the same dataset or workflow.
SNOMED CT is closer to a clinical terminology for concepts such as findings, disorders, procedures, and some observable entities, while FHIR is a data exchange standard used to structure and transport healthcare information between systems. In practice, LOINC may identify the observation, SNOMED CT may describe related clinical concepts, and FHIR may carry those coded data between systems.
At a glance, these standards often work together as such:
-
LOINC provides a standard identifier for the laboratory test or observation
- CDISC terminology ensures the data is structured consistently within clinical trial datasets
- ICD classifies diseases or clinical outcomes associated with the patient
By using each standard for its intended purpose, organisations can achieve clearer data definition, better interoperability, and more reliable integration across systems and studies.
LOINC code limitations: what can go wrong?
While LOINC codes provide a strong foundation for standardising laboratory data, their use in practice is not without challenges. Understanding these limitations helps ensure that codes are applied appropriately and that data quality is not compromised. These include:
Absence of suitable codes
Not every laboratory test, particularly newer or highly specialised assays, may have an exact match in the LOINC database. In such cases, selecting a near match can introduce ambiguity if the underlying attributes do not fully align with the actual test performed.
Variation in lab support
Not all laboratories consistently assign or maintain LOINC codes within their systems. This can lead to gaps in source data, requiring additional mapping effort during data processing, which may depend on limited metadata.
Over-specification
Since LOINC codes are defined using detailed attributes, selecting a code that is too specific (for example, including a method that is not clearly confirmed) can misrepresent the observation. This is particularly relevant when source data does not capture all defining characteristics of the test.
Forced mappings based on assumptions
When key details such as specimen type or measurement method are missing, there may be pressure to assign a LOINC code to maintain completeness. However, assigning a code without sufficient evidence can introduce inconsistencies and reduce the reliability of the dataset.
Therefore, a balanced approach is needed. It is often preferable to leave a LOINC code unassigned, or clearly document uncertainty, rather than apply a potentially incorrect mapping. Careful review, appropriate use of available metadata, and transparent documentation all play an important role in maintaining data integrity when working with LOINC.
LOINC codes and FDA submissions
In regulatory submissions, laboratory data is typically provided in standardised formats such as SDTM. Within this context, the use of consistent and well-defined coding systems plays an important role in supporting data clarity, traceability, and efficient review. LOINC codes contribute to this by providing a recognised way to identify laboratory tests and observations across datasets, helping to support traceability and reviewer clarity.
LOINC is best treated as an established standard for identifying many laboratory tests in relevant submission contexts, while still recognising that not every test will have a suitable code. When applied appropriately, LOINC codes can help reviewers interpret laboratory data more consistently, particularly in studies involving multiple sites or external laboratory providers.
In submission-ready datasets, LOINC codes are typically included alongside SDTM variables. They act as an additional layer of standardisation, helping to align the meaning of laboratory tests beyond local naming conventions. This can be particularly useful when integrating data across studies or when supporting cross-study analysis.
A practical consideration is how to handle situations where LOINC codes are unavailable or uncertain. In such cases, it is generally acceptable to leave the field unpopulated rather than assign a code based on incomplete information. Clear documentation of coding decisions, including any gaps or limitations, is important to maintain transparency.
Overall, the role of LOINC in FDA submissions is to support clearer, more consistent laboratory data without overstating its necessity. When used thoughtfully, it enhances the interpretability of datasets and aligns with the broader regulatory expectation for well-structured and standardised clinical data.
Conclusion
LOINC codes play an important role in improving how laboratory data is defined, shared, and interpreted across clinical systems and studies. By providing a consistent reference for laboratory tests and observations, they help reduce ambiguity and support more reliable data integration.
Their value is most evident in structured workflows, particularly where laboratory data must be standardised within datasets such as SDTM and prepared for analysis or regulatory submission. In these settings, LOINC complements existing data structures by adding a clear and consistent layer of meaning.
At the same time, effective use of LOINC requires careful judgement. Decisions around mapping, handling incomplete source data, and selecting the appropriate level of specificity directly affect data quality. Applying codes without sufficient context can introduce inconsistencies, making a balanced and well-documented approach essential.
FAQs
What is a LOINC code?
A LOINC code is a standardised identifier used to represent a specific laboratory test or clinical observation. It provides a consistent way to define what has been measured, independent of local naming conventions.
What are LOINC codes used for?
LOINC codes are used to ensure that laboratory tests and observations are identified consistently across labs, systems, and studies. This supports clearer data exchange, integration, and interpretation.
What is a LOINC code in a lab?
In a laboratory setting, a LOINC code acts as a standard reference for a specific test. It helps ensure that a result generated by one lab can be understood accurately by another system or organisation without relying on local test names.
What is the difference between ICD and LOINC codes?
LOINC codes identify laboratory tests and observations, while ICD codes classify diseases and diagnoses. In simple terms, LOINC describes what was measured, whereas ICD describes what condition a patient has.
What does a LOINC code look like?
A LOINC code typically appears as a number followed by a hyphen and a check digit, for example 2345-7. The code itself is a compact identifier, while its full meaning comes from the detailed definition linked to it.
Are LOINC codes mandatory in clinical trials?
LOINC codes are not always mandatory, but their use is widely encouraged to support standardisation and clearer interpretation of laboratory data. In many cases, they are included where available, particularly for submission-ready datasets.
How are LOINC codes assigned to lab tests?
LOINC codes are typically assigned by central laboratories or source systems. If not provided, they may be mapped during data processing using available test details such as analyte, specimen, and measurement type.
What happens if a LOINC code is not available?
If no suitable LOINC code exists or if the available information is insufficient, it is generally preferable to leave the field unassigned rather than apply an uncertain or incorrect mapping.
Do LOINC codes replace CDISC variables in SDTM?
No, LOINC codes do not replace SDTM variables. They complement them by providing a standardised reference for the laboratory test, while SDTM variables continue to structure and describe the data.
Quanticate’s statistical programming team can support with laboratory data standardisation, LOINC mapping, traceability, and submission-ready dataset preparation. Request a consultation and a member of our team will be in touch.
