FHIRkit

Handle FHIR resources in a more efficient, and pythonic way

Why FHIRKit

FHIRKit is a wrapper around Pydantic that will help you parse FHIR JSON data. It comes with some utility functions to display resources 👀 and validate codes in those resources ✅.

After implementing calls to terminology servers over and over again 😥, we realized that Python developpers and data scientists want code like this:

target_code = SCTCoding(code="298364001")
vs = target_code.descendants() # create a valueset by retrieving all descendants of a SNOMED-CT code
obs = Observation.parse_file("my_observation.json")
assert obs.code in vs, f"Observation has an invalid code={obs.code}, please provide a {target_code}"
#
# AssertionError
# Observation has an invalid code="55406008 | Hyperalgesia |", please provide a "298364001 | Finding of head region |"

Calls to terminology servers are handled for you and boilerplate code is reduced to simple Python operations 🎉.

The goals of this package in a nutshell:

• Parse FHIR/JSON(https://build.fhir.org/json.html) resources using predefined Pydantic models. These models automatically handle all inconveniences like choice data typess and datetime parsing for you.

• Perform validations and lookups for code, coding, CodeableConcept elements and expand ValueSets resources in a pythonic way.

• Make FHIR Resources and FHIR Elements readable when printing them in the command-line or displaying them in a Jupyter Notebook. FHIRkit has some handy defaults for __str__, __repr__ and __repr_html__ that take care of your development and debug experience. No huge nested dicts in your terminal or notebook.

• Turn a bunch of FHIR resources with the same type in a FHIR DataFrame (still under development)