Protocol Deviation Classifier#

Clinical trial protocol deviation classification tool, based on GCP and ICH E6 guidelines, automatically determines whether deviations belong to "major deviations" or "minor deviations".

When to Use#

• Use this skill when the task needs Determine whether an incident in a clinical trial is a "major deviation.
• Use this skill for data analysis tasks that require explicit assumptions, bounded scope, and a reproducible output format.
• Use this skill when you need a documented fallback path for missing inputs, execution errors, or partial evidence.

Key Features#

See ## Features above for related details.

• Scope-focused workflow aligned to: Determine whether an incident in a clinical trial is a "major deviation.
• Packaged executable path(s): scripts/main.py.
• Reference material available in references/ for task-specific guidance.
• Structured execution path designed to keep outputs consistent and reviewable.

Dependencies#

• Python 3.8+
• No third-party dependencies (pure Python standard library implementation)

Example Usage#

See ## Usage above for related details.

bash

cd "20260318/scientific-skills/Data Analytics/protocol-deviation-classifier"
python -m py_compile scripts/main.py
python scripts/main.py --help

Example run plan:

1. Confirm the user input, output path, and any required config values.
2. Edit the in-file CONFIG block or documented parameters if the script uses fixed settings.
3. Run python scripts/main.py with the validated inputs.
4. Review the generated output and return the final artifact with any assumptions called out.

Implementation Details#

See ## Workflow above for related details.

• Execution model: validate the request, choose the packaged workflow, and produce a bounded deliverable.
• Input controls: confirm the source files, scope limits, output format, and acceptance criteria before running any script.
• Primary implementation surface: scripts/main.py.
• Reference guidance: references/ contains supporting rules, prompts, or checklists.
• Parameters to clarify first: input path, output path, scope filters, thresholds, and any domain-specific constraints.
• Output discipline: keep results reproducible, identify assumptions explicitly, and avoid undocumented side effects.

Quick Check#

Use this command to verify that the packaged script entry point can be parsed before deeper execution.

bash

python -m py_compile scripts/main.py

Audit-Ready Commands#

Use these concrete commands for validation. They are intentionally self-contained and avoid placeholder paths.

bash

python -m py_compile scripts/main.py
python scripts/main.py --help
python scripts/main.py --input "Audit validation sample with explicit symptoms, history, assessment, and next-step plan." --format json

Workflow#

1. Confirm the user objective, required inputs, and non-negotiable constraints before doing detailed work.
2. Validate that the request matches the documented scope and stop early if the task would require unsupported assumptions.
3. Use the packaged script path or the documented reasoning path with only the inputs that are actually available.
4. Return a structured result that separates assumptions, deliverables, risks, and unresolved items.
5. If execution fails or inputs are incomplete, switch to the fallback path and state exactly what blocked full completion.

Features#

• Automatic Classification: Automatically determines severity based on deviation description
• Risk Assessment: Assesses impact on subject safety, data integrity, and scientific validity
• Regulatory Basis: Classification basis complies with GCP, ICH E6, and FDA/EMA guidelines
• Report Generation: Generates deviation classification reports that meet regulatory requirements
• Chinese Support: Full support for Chinese clinical trial scenarios

Deviation Classification Standards#

Major/Critical Deviation#

Deviations that may affect trial data integrity, subject safety, or trial scientific validity:

CategoryExamplesInformed ConsentPerforming research procedures without informed consent, using expired/incorrect informed consent formsInclusion/Exclusion CriteriaEnrolling subjects who don't meet inclusion criteria, enrolling subjects who meet exclusion criteriaInvestigational ProductOverdose administration, contraindicated concomitant medication, incorrect route of administration, randomization errorSafetyNot performing safety monitoring as required by protocol, missing SAE/SUSAR reports, delayed reportingBlindingUnblinding by unauthorized personnel, unrecorded emergency unblinding proceduresData IntegrityFalsifying/fabricating data, systematic missing of critical dataProhibited OperationsViolating key operational procedures of trial protocol, not performing key efficacy assessments

Minor Deviation#

Deviations unlikely to affect trial data integrity, subject safety, or trial scientific validity:

CategoryExamplesVisit WindowSlightly exceeding visit time window (e.g., within a few days), delay of non-critical visitsSample CollectionMinor timing deviations in non-critical sample collection, slight delays in sample processingQuestionnaire CompletionQuality of life questionnaires/diary cards submitted a few days lateData RecordingDelays in non-critical data recording, spelling/formatting errorsProcedure ExecutionAdjustment of secondary procedure execution order, omission of non-critical assessments (e.g., height measurement)DocumentationDelays in source document signatures, missing secondary documents (e.g., non-critical examination reports)

Usage#

Python API#

python

from scripts.main import DeviationClassifier

# Initialize classifier
classifier = DeviationClassifier()

# Classify single deviation
result = classifier.classify(
    description="Subject visit delayed by 2 days",
    deviation_type="Visit Window"
)
print(result.classification)  # "Minor Deviation"
print(result.confidence)      # 0.92
print(result.rationale)       # Classification rationale explanation

# Batch classification
deviations = [
    {"description": "Blood sample collected without informed consent", "type": "Informed Consent"},
    {"description": "Quality of life questionnaire submitted 3 days late", "type": "Data Collection"}
]
batch_results = classifier.classify_batch(deviations)

# Generate report
report = classifier.generate_report(batch_results)

CLI Usage#

text

# Classify single deviation
python scripts/main.py classify --description "Subject visit delayed by 2 days" --type "Visit Window"

# Batch classification from file
python scripts/main.py batch --input deviations.json --output report.json

# Interactive classification
python scripts/main.py interactive

# Assess deviation impact
python scripts/main.py assess \
  --description "Subject accidentally took double dose of investigational drug" \
  --safety-impact high \
  --data-impact medium \
  --scientific-impact medium

Input Format#

JSON Input File Format:

json

[
  {
    "id": "DEV-001",
    "description": "Subject visit delayed by 2 days",
    "type": "Visit Window",
    "occurrence_date": "2024-01-15",
    "severity_factors": {
      "safety_impact": "none",
      "data_impact": "low",
      "scientific_impact": "low"
    }
  },
  {
    "id": "DEV-002",
    "description": "Blood collection performed without informed consent",
    "type": "Informed Consent",
    "severity_factors": {
      "safety_impact": "high",
      "data_impact": "high",
      "scientific_impact": "high"
    }
  }
]

Output Format#

Classification Result:

json

{
  "id": "DEV-001",
  "classification": "Minor Deviation",
  "classification_en": "Minor Deviation",
  "confidence": 0.92,
  "rationale": "Visit time window slightly delayed (2 days), does not affect subject safety, data integrity, or trial scientific validity.",
  "risk_factors": {
    "safety_risk": "none",
    "data_integrity_risk": "low",
    "scientific_validity_risk": "none"
  },
  "regulatory_basis": [
    "ICH E6(R2) Section 4.5",
    "GCP Section 6.4.4"
  ],
  "recommended_actions": [
    "Document in file",
    "Track trends"
  ]
}

Classification Algorithm#

Classification based on the following assessment dimensions:

1. Subject Safety Impact (Safety Impact)
   • None: No impact
   • Low: Minor impact
   • Medium: Moderate impact
   • High: Serious impact
2. Data Integrity Impact (Data Integrity Impact)
   • None: No impact
   • Low: Minor impact on non-critical data
   • Medium: Partial impact on critical data
   • High: Serious damage to critical data
3. Trial Scientific Validity Impact (Scientific Validity Impact)
   • None: No impact
   • Low: Minor impact on statistical power
   • Medium: May affect primary endpoint
   • High: Seriously affects trial conclusion

Classification Rules:

• Any dimension is High → Major Deviation
• Safety dimension is Medium and Data/Science either is Medium+ → Major Deviation
• Other cases → Minor Deviation

Regulatory Basis#

• ICH E6(R2) Good Clinical Practice Guideline
• ICH E6(R3) Good Clinical Practice Guideline (Draft)
• FDA 21 CFR Part 312 (IND Regulations)
• FDA Guidance for Industry: Oversight of Clinical Investigations
• EMA Reflection Paper on Risk Based Quality Management
• NMPA Good Clinical Practice for Drug Clinical Trials

Notes#

1. This tool provides classification recommendations, final determination must be confirmed by clinical quality assurance personnel
2. Serious/critical deviations must be reported to sponsor and ethics committee immediately
3. It is recommended to regularly review deviation trends and implement CAPA (Corrective and Preventive Actions)
4. Classification standards may vary by regulatory agency, trial type, and protocol requirements

Risk Assessment#

Risk IndicatorAssessmentLevelCode ExecutionPython/R scripts executed locallyMediumNetwork AccessNo external API callsLowFile System AccessRead input files, write output filesMediumInstruction TamperingStandard prompt guidelinesLowData ExposureOutput files saved to workspaceLow

Security Checklist#

• No hardcoded credentials or API keys
• No unauthorized file system access (../)
• Output does not expose sensitive information
• Prompt injection protections in place
• Input file paths validated (no ../ traversal)
• Output directory restricted to workspace
• Script execution in sandboxed environment
• Error messages sanitized (no stack traces exposed)
• Dependencies audited

Prerequisites#

text

# Python dependencies
pip install -r requirements.txt

Evaluation Criteria#

Success Metrics#

• Successfully executes main functionality
• Output meets quality standards
• Handles edge cases gracefully
• Performance is acceptable

Test Cases#

1. Basic Functionality: Standard input → Expected output
2. Edge Case: Invalid input → Graceful error handling
3. Performance: Large dataset → Acceptable processing time

Lifecycle Status#

• Current Stage: Draft

• Next Review Date: 2026-03-06

• Known Issues: None

• Planned Improvements:

  • Performance optimization
  • Additional feature support

Output Requirements#

Every final response should make these items explicit when they are relevant:

• Objective or requested deliverable
• Inputs used and assumptions introduced
• Workflow or decision path
• Core result, recommendation, or artifact
• Constraints, risks, caveats, or validation needs
• Unresolved items and next-step checks

Error Handling#

• If required inputs are missing, state exactly which fields are missing and request only the minimum additional information.
• If the task goes outside the documented scope, stop instead of guessing or silently widening the assignment.
• If scripts/main.py fails, report the failure point, summarize what still can be completed safely, and provide a manual fallback.
• Do not fabricate files, citations, data, search results, or execution outcomes.

Input Validation#

This skill accepts requests that match the documented purpose of protocol-deviation-classifier and include enough context to complete the workflow safely.

Do not continue the workflow when the request is out of scope, missing a critical input, or would require unsupported assumptions. Instead respond:

protocol-deviation-classifier only handles its documented workflow. Please provide the missing required inputs or switch to a more suitable skill.

Response Template#

Use the following fixed structure for non-trivial requests:

1. Objective
2. Inputs Received
3. Assumptions
4. Workflow
5. Deliverable
6. Risks and Limits
7. Next Checks

If the request is simple, you may compress the structure, but still keep assumptions and limits explicit when they affect correctness.

Inputs to Collect#

• Required inputs: the user goal, the primary data or source file, and the requested output format.
• Optional inputs: output directory, formatting preferences, and validation constraints.
• If a required input is unavailable, return a short clarification request before continuing.

Output Contract#

• Return a short summary, the main deliverables, and any assumptions that materially affect interpretation.
• If execution is partial, label what succeeded, what failed, and the next safe recovery step.
• Keep the final answer within the documented scope of the skill.

Validation and Safety Rules#

• Validate identifiers, file paths, and user-provided parameters before execution.
• Do not fabricate results, metrics, citations, or downstream conclusions.
• Use safe fallback behavior when dependencies, credentials, or required inputs are missing.
• Surface any execution failure with a concise diagnosis and recovery path.