LOINC Integration Guide (for IT / developers)

This guide is written for hospital information system (HIS) developers and data engineers, explaining how to map in-house lab codes to standard LOINC codes.

Why LOINC?

Hospitals commonly use their own codes (such as GLU-AC), but exchanging data between institutions or uploading to a health passbook requires a common language. LOINC is that standard.

Integration strategy

1. Build a mapping table

Use the search_loinc(mode="code", ...) tool to find the standard code that best matches each in-house item.

Comparing the key attributes: When choosing a LOINC code, confirm that these six axes agree: 1. Component: the analyte (such as glucose) 2. Property: the measured property (such as mass concentration) 3. Time aspect: the timing (such as Pt = point in time, or 24H) 4. System: the specimen type (such as serum/plasma or urine) 5. Scale: the measurement scale (such as quantitative) 6. Method: the measurement method (a method-less code is acceptable when it does not affect interpretation)

Using the tool:

# Search for "urine protein"
search_loinc(mode="code", keyword="Urine Protein")
# The results distinguish qualitative from quantitative

2. Verify the reference ranges

Once mapped, use query_loinc(mode="reference_range", ...) to confirm that the standard reference values for that LOINC code are close to the in-house laboratory's. A large discrepancy may mean the wrong code was chosen (for example a code with different units or a different measurement method).

3. Automate the conversion

In the exporter module: 1. Read the in-house lab result. 2. Convert it to LOINC via the mapping table. 3. Call or consult query_loinc(mode="reference_range", ...) to fill in standard reference range information (where in-house data is missing). 4. Package it as a FHIR Observation resource.

Common issues

• Unit conversion: LOINC usually has a default unit (such as mg/dL). If the hospital uses mmol/L, convert the numeric value during the transformation.