Hazard¶

DO NOT EDIT BELOW THIS LINE UNLESS YOU KNOW WHAT YOU ARE DOING

Hazard name¶

Invalid FHIR gender value

General utility label¶

[2]

Likelihood scoring¶

TBC

Severity scoring¶

TBC

Description¶

Gender input converted to lowercase without proper validation against FHIR R4 required values (male|female|other|unknown), potentially allowing invalid values to be stored.

Causes¶

Input converted to .lower() but typos like "femail" pass lowercase conversion
No validation against FHIR R4 enum before submission
FHIR server may reject or silently store invalid value

Effect¶

Invalid gender value submitted to FHIR server, either rejected causing error or stored incorrectly.

Hazard¶

Gender-specific clinical rules not applied correctly (pregnancy checks for "female" patients, prostate screening for "male" patients).

Hazard type¶

WrongPatientContext

Harm¶

Inappropriate treatment or missed screening. Teratogenic drug given to patient who should be screened for pregnancy. Prostate cancer screening missed for male patient.

Existing controls¶

None identified during initial analysis.

Assignment¶

Clinical Safety Officer

Labelling¶

TBC (awaiting scoring)

Project¶

Clinical Risk Management

Hazard controls¶

Design controls (manufacturer)¶

Add Pydantic validator in backend schemas enforcing FHIR R4 gender enum before API submission. Use Literal["male", "female", "other", "unknown"] type annotation. Reject any value not in enum with 400 error: "Invalid gender value. Must be one of: male, female, other, unknown."
Replace free-text gender input with dropdown/radio button selection in frontend forms providing only valid options. Prevents typos entirely by not allowing text entry.
Add frontend TypeScript enum matching FHIR gender values: enum FHIRGender { MALE = "male", FEMALE = "female", OTHER = "other", UNKNOWN = "unknown" }. Enforce at compile time.
Implement backend validation middleware that checks all FHIR resources before submission, verifying gender field matches FHIR R4 specification. Log validation failures for debugging.
Add data migration script to scan existing FHIR patients, identify any with invalid gender values, update to "unknown" with audit log entry documenting correction.

Testing controls (manufacturer)¶

Unit test: Attempt to create patient with gender="femail" via API. Assert 400 Bad Request returned with error message specifying valid values. Verify no FHIR resource created.
Unit test: Test each valid gender value (male, female, other, unknown) successfully creates patient. Test capitalized inputs ("Male", "FEMALE") are normalized to lowercase correctly.
Integration test: Mock FHIR server to reject invalid gender, verify backend catches rejection and returns meaningful error to frontend instead of generic 500 error.
Unit test: Test edge cases: null gender, empty string gender, numeric gender value. Assert all invalid inputs rejected at Pydantic validation layer.

Training controls (deployment)¶

Train registration staff on importance of accurate gender documentation for clinical decision support (pregnancy checks, cancer screening).
Document workflow: "If patient's gender not easily categorized as male/female, select 'other' and document details in clinical notes. Use 'unknown' only when patient unable/unwilling to provide information."
Include in diversity training: Explain difference between sex assigned at birth (biological), gender identity (self-identification), and clinical sex (for medical purposes). Document sensitivity when asking questions.

Business process controls (deployment)¶

NHS Trust policy: Gender must be documented for all patients at registration. Registration incomplete without gender field populated.
Clinical governance: Gender-specific clinical decision support rules (pregnancy screening for females 12-55) must account for 'other' and 'unknown' values by prompting clinician for manual assessment.
Data quality audit: Monthly scan for patients with 'unknown' gender, review cases to determine if information can be obtained and updated through clinical contact.

Residual hazard risk assessment¶

TBC — awaiting initial controls implementation.

Hazard status¶

Draft from LLM

Code associated with hazard¶

backend/app/fhir_client.py:120-125
backend/app/fhir_client.py:217