empathy-framework 4.6.6__py3-none-any.whl → 4.7.1__py3-none-any.whl

wizards/sbar_report.py

@@ -1,324 +0,0 @@
-"""
-A wizard for generating SBAR (Situation, Background, Assessment, Recommendation) reports.
-
-This multi-step wizard guides users through creating a structured SBAR report,
-which is a standard communication tool in healthcare settings.
-"""
-
-from datetime import datetime
-
-from fastapi import APIRouter, HTTPException, status
-from pydantic import BaseModel, Field
-from utils.api_responses import create_error_response, create_success_response
-from utils.logging import get_logger
-
-from services.openai_client import get_client
-
-router = APIRouter(prefix="/wizards/sbar-report", tags=["wizards"])
-logger = get_logger(__name__)
-
-# In-memory storage for wizard state. In production, use Redis or a database.
-wizard_sessions: dict[str, dict[str, str]] = {}
-
-# --- Pydantic Models for Wizard Steps ---
-
-
-class StartSbarResponse(BaseModel):
-    wizard_id: str
-    message: str
-    next_step: str
-
-
-class SbarStepInput(BaseModel):
-    wizard_id: str = Field(..., description="The unique ID for this wizard session.")
-    text: str = Field(..., description="The text content for the current step.")
-
-
-class SbarStepResponse(BaseModel):
-    wizard_id: str
-    next_step: str
-    message: str
-
-
-class GenerateSbarInput(BaseModel):
-    wizard_id: str = Field(..., description="The unique ID for this wizard session.")
-    recommendation: str = Field(..., description="The 'Recommendation' text.")
-
-
-class GenerateSbarResponse(BaseModel):
-    wizard_id: str
-    sbar_report: str
-
-
-class DirectSbarInput(BaseModel):
-    patient_id: str = Field(..., description="Patient identifier")
-    situation: str = Field(..., description="The 'Situation' text")
-    background: str = Field(..., description="The 'Background' text")
-    assessment: str = Field(..., description="The 'Assessment' text")
-    recommendation: str = Field(..., description="The 'Recommendation' text")
-    care_setting: str = Field(default="med-surg", description="Care setting context")
-
-
-class DirectSbarResponse(BaseModel):
-    sbar_report: str
-    care_setting: str
-
-
-# --- Wizard Endpoints ---
-
-
-@router.post("/start", response_model=StartSbarResponse, summary="Step 1: Start the SBAR wizard")
-async def start_sbar_wizard():
-    """
-    Initializes a new SBAR report wizard session and returns a unique `wizard_id`.
-    """
-    import uuid
-
-    wizard_id = str(uuid.uuid4())
-    wizard_sessions[wizard_id] = {}
-    logger.info(f"Started SBAR wizard session: {wizard_id}")
-
-    return create_success_response(
-        {
-            "wizard_id": wizard_id,
-            "message": "SBAR wizard started. Please proceed to add the 'Situation'.",
-            "next_step": "situation",
-        }
-    )
-
-
-def _update_step(wizard_id: str, step_name: str, text: str, next_step_name: str):
-    """Helper function to update a step in the wizard session."""
-    if wizard_id not in wizard_sessions:
-        return create_error_response(
-            "Wizard session not found.", status.HTTP_404_NOT_FOUND, "wizard_not_found"
-        )
-
-    wizard_sessions[wizard_id][step_name] = text
-    logger.info(f"Updated SBAR session {wizard_id} with step: {step_name}")
-
-    return create_success_response(
-        {
-            "wizard_id": wizard_id,
-            "message": f"'{step_name.capitalize()}' received. Please proceed to add the '{next_step_name.capitalize()}'.",
-            "next_step": next_step_name,
-        }
-    )
-
-
-@router.post("/situation", response_model=SbarStepResponse, summary="Step 2: Add Situation")
-async def add_situation(step_input: SbarStepInput):
-    """Adds the **Situation** component to the SBAR report."""
-    return _update_step(step_input.wizard_id, "situation", step_input.text, "background")
-
-
-@router.post("/background", response_model=SbarStepResponse, summary="Step 3: Add Background")
-async def add_background(step_input: SbarStepInput):
-    """Adds the **Background** component to the SBAR report."""
-    return _update_step(step_input.wizard_id, "background", step_input.text, "assessment")
-
-
-@router.post("/assessment", response_model=SbarStepResponse, summary="Step 4: Add Assessment")
-async def add_assessment(step_input: SbarStepInput):
-    """Adds the **Assessment** component to the SBAR report."""
-    return _update_step(step_input.wizard_id, "assessment", step_input.text, "recommendation")
-
-
-@router.post(
-    "/generate",
-    response_model=DirectSbarResponse,
-    summary="Generate Complete SBAR Report (Direct)",
-)
-async def generate_sbar_report_direct(input_data: DirectSbarInput):
-    """
-    Generate a complete SBAR report directly from all provided components.
-    This is a simplified endpoint that doesn't require wizard session management.
-    """
-    # Validate that all required fields are present
-    if not all(
-        [
-            input_data.situation,
-            input_data.background,
-            input_data.assessment,
-            input_data.recommendation,
-        ]
-    ):
-        return create_error_response(
-            "All SBAR components (situation, background, assessment, recommendation) are required.",
-            status.HTTP_400_BAD_REQUEST,
-            "missing_required_fields",
-        )
-
-    # Build context-aware prompt based on care setting
-    care_setting_context = {
-        "icu": "critical care setting with focus on hemodynamic stability and ventilator management",
-        "emergency": "emergency department with focus on rapid assessment and triage",
-        "home-health": "home health setting with focus on patient/caregiver education and safety",
-        "med-surg": "medical-surgical unit with focus on routine monitoring and care coordination",
-    }.get(input_data.care_setting, "general clinical setting")
-
-    prompt = f"""
-    Format the following clinical information into a clear, professional SBAR report for a {care_setting_context}.
-
-    Use these exact headings:
-    - SITUATION
-    - BACKGROUND
-    - ASSESSMENT
-    - RECOMMENDATION
-
-    Format the report professionally with proper spacing and organization suitable for clinical handoff.
-    Patient ID: {input_data.patient_id}
-
-    Situation: {input_data.situation}
-
-    Background: {input_data.background}
-
-    Assessment: {input_data.assessment}
-
-    Recommendation: {input_data.recommendation}
-    """
-
-    try:
-        client = get_client()
-        if not client:
-            # Fallback: Generate basic SBAR report without AI enhancement
-            logger.warning(
-                "OpenAI client not available - generating basic SBAR report without AI formatting"
-            )
-            sbar_report = f"""SBAR Report - Patient ID: {input_data.patient_id}
-Care Setting: {care_setting_context.upper()}
-
-SITUATION
-{input_data.situation}
-
-BACKGROUND
-{input_data.background}
-
-ASSESSMENT
-{input_data.assessment}
-
-RECOMMENDATION
-{input_data.recommendation}
-
-Generated: {datetime.now().strftime('%Y-%m-%d %H:%M:%S')}
-Note: This report was generated without AI enhancement. Configure OPENAI_API_KEY for enhanced formatting.
-"""
-            return {"sbar_report": sbar_report, "care_setting": input_data.care_setting}
-
-        response = client.chat.completions.create(
-            model="gpt-4o-mini",
-            messages=[
-                {
-                    "role": "system",
-                    "content": f"You are an expert nursing assistant specializing in clinical communication in a {care_setting_context}. Format SBAR reports clearly and professionally.",
-                },
-                {"role": "user", "content": prompt},
-            ],
-            temperature=0.3,  # Lower temperature for more consistent formatting
-        )
-        sbar_report = response.choices[0].message.content
-        logger.info(
-            f"Successfully generated SBAR report for patient {input_data.patient_id} in {input_data.care_setting} setting"
-        )
-
-        return {"sbar_report": sbar_report, "care_setting": input_data.care_setting}
-    except Exception as e:
-        logger.error(f"SBAR report generation failed: {e}", exc_info=True)
-        raise HTTPException(
-            status_code=status.HTTP_500_INTERNAL_SERVER_ERROR,
-            detail=f"Failed to generate SBAR report: {str(e)}",
-        )
-
-
-@router.post(
-    "/generate-wizard",
-    response_model=GenerateSbarResponse,
-    summary="Step 5: Add Recommendation and Generate Report (Wizard)",
-)
-async def generate_sbar_report_wizard(step_input: GenerateSbarInput):
-    """
-    Adds the final **Recommendation** component and generates the complete,
-    formatted SBAR report using an AI model. (Wizard session-based flow)
-    """
-    wizard_id = step_input.wizard_id
-    if wizard_id not in wizard_sessions:
-        return create_error_response(
-            "Wizard session not found.", status.HTTP_404_NOT_FOUND, "wizard_not_found"
-        )
-
-    session_data = wizard_sessions[wizard_id]
-    session_data["recommendation"] = step_input.recommendation
-
-    # Verify all parts are present
-    required_parts = ["situation", "background", "assessment", "recommendation"]
-    if not all(part in session_data for part in required_parts):
-        missing = [part for part in required_parts if part not in session_data]
-        return create_error_response(
-            f"Cannot generate report. Missing parts: {', '.join(missing)}",
-            status.HTTP_400_BAD_REQUEST,
-            "missing_sbar_parts",
-        )
-
-    prompt = f"""
-    Format the following clinical information into a clear and concise SBAR report.
-    Use clear headings for each section (Situation, Background, Assessment, Recommendation).
-    Ensure the language is professional and suitable for a clinical handoff.
-
-    Situation: {session_data['situation']}
-    Background: {session_data['background']}
-    Assessment: {session_data['assessment']}
-    Recommendation: {session_data['recommendation']}
-    """
-
-    try:
-        client = get_client()
-        if not client:
-            # Fallback: Generate basic SBAR report without AI enhancement
-            logger.warning(
-                f"OpenAI client not available for session {wizard_id} - generating basic SBAR report"
-            )
-            sbar_report = f"""SBAR Report
-
-SITUATION
-{session_data['situation']}
-
-BACKGROUND
-{session_data['background']}
-
-ASSESSMENT
-{session_data['assessment']}
-
-RECOMMENDATION
-{session_data['recommendation']}
-
-Generated: {datetime.now().strftime('%Y-%m-%d %H:%M:%S')}
-Note: This report was generated without AI enhancement. Configure OPENAI_API_KEY for enhanced formatting.
-"""
-            # Clean up the session
-            del wizard_sessions[wizard_id]
-
-            return {"wizard_id": wizard_id, "sbar_report": sbar_report}
-
-        response = client.chat.completions.create(
-            model="gpt-4o-mini",
-            messages=[
-                {
-                    "role": "system",
-                    "content": "You are an expert nursing assistant specializing in clinical communication.",
-                },
-                {"role": "user", "content": prompt},
-            ],
-        )
-        sbar_report = response.choices[0].message.content
-        logger.info(f"Successfully generated SBAR report for session {wizard_id}")
-
-        # Clean up the session
-        del wizard_sessions[wizard_id]
-
-        return {"wizard_id": wizard_id, "sbar_report": sbar_report}
-    except Exception as e:
-        logger.error(f"SBAR report generation failed for session {wizard_id}: {e}", exc_info=True)
-        raise HTTPException(
-            status_code=status.HTTP_500_INTERNAL_SERVER_ERROR,
-            detail=f"Failed to generate SBAR report: {str(e)}",
-        )