What problem does it solve?
This Skill converts laboratory values and reference ranges from clinical text into structured low, normal, high, critical, or unknown results without silently guessing when inputs are incomplete or ambiguous.
Core Features & Use Cases
- Reference Range Parsing: Handles closed ranges such as 135–145 and one-sided bounds such as <5, <=5, >10, and >=10.
- Abnormality Classification: Compares numeric results with parsed ranges while respecting inclusive and exclusive boundaries.
- Explicit Flag Handling: Honors originating laboratory flags such as H, L, C, HH, and LL before applying derived comparisons.
- Clinical Data Workflows: Turns extracted lab entities into structured signals suitable for flagged laboratory tables, FHIR Observation interpretation, and downstream clinical NLP workflows.
- Safe Interpretation: Returns unknown for non-numeric values, unparseable ranges, or unrecognized flags, and keeps unit conversion outside the classification logic.
Quick Start
Use the parsing lab values skill to classify each laboratory result in the provided clinical text using its value, reference range, and any originating laboratory flag.
