Upload 4 files

logic/care_gap_engine.py

@@ -0,0 +1,39 @@
+# File: logic/care_gap_engine.py
+
+import pandas as pd
+from datetime import datetime
+
+def evaluate_care_gaps(df: pd.DataFrame, config):
+    today = datetime.today()
+    rules = config["care_gap_rules"]
+    results = []
+
+    for _, row in df.iterrows():
+        gaps = []
+
+        if row['gender'] == 'F' and rules['Breast Cancer Screening']['min_age'] <= row['age'] <= rules['Breast Cancer Screening']['max_age']:
+            if not row['last_mammogram'] or (today - pd.to_datetime(row['last_mammogram'], errors='coerce')).days > rules['Breast Cancer Screening']['interval_days']:
+                gaps.append("Breast Cancer Screening")
+
+        if rules['Colorectal Cancer Screening']['min_age'] <= row['age'] <= rules['Colorectal Cancer Screening']['max_age']:
+            if not row['last_colonoscopy'] or (today - pd.to_datetime(row['last_colonoscopy'], errors='coerce')).days > rules['Colorectal Cancer Screening']['interval_days']:
+                gaps.append("Colorectal Cancer Screening")
+
+        if row.get('systolic_bp', '') != "" and pd.to_numeric(row['systolic_bp'], errors='coerce') > rules['Blood Pressure Control']['bp_threshold']:
+            gaps.append("Blood Pressure Control")
+
+        if row.get('hba1c_value', '') != "" and pd.to_numeric(row['hba1c_value'], errors='coerce') >= 9:
+            gaps.append("Diabetes: Poor HbA1c Control")
+
+        if row.get('FollowUp_Scheduled', '').strip().lower() == 'no' or row.get('Primary_Care_Established', '').strip().lower() == 'no':
+            gaps.append("Follow-Up Care")
+
+        if row.get('Previous_Readmissions', '').isdigit() and int(row['Previous_Readmissions']) >= 3:
+            gaps.append("Readmission Risk")
+
+        results.append({
+            'patient_id': row['patient_id'],
+            'care_gaps': gaps
+        })
+
+    return pd.DataFrame(results)

logic/exporter.py

@@ -0,0 +1,29 @@
+# File: logic/exporter.py
+
+from docx import Document
+import os
+import io
+import zipfile
+
+def export_to_docx_and_zip(report_data, summaries, output_dir="assets/sample_reports"):
+    zip_buffer = io.BytesIO()
+    with zipfile.ZipFile(zip_buffer, "w") as zipf:
+        for _, row in report_data.iterrows():
+            pid = row['patient_id']
+            doc = Document()
+            doc.add_heading(f'Patient Summary: {pid}', 0)
+            doc.add_paragraph(f"Risk Score: {row['risk_score']}")
+            doc.add_paragraph(f"Projected Revenue Gain: ${row['projected_gain']}")
+            doc.add_paragraph("Care Gaps:")
+            for gap in row['care_gaps']:
+                doc.add_paragraph(f"• {gap}", style='List Bullet')
+            doc.add_paragraph("\nSummary:")
+            doc.add_paragraph(summaries[pid])
+
+            doc_path = f"Patient_{pid}_Report.docx"
+            temp_stream = io.BytesIO()
+            doc.save(temp_stream)
+            zipf.writestr(doc_path, temp_stream.getvalue())
+
+    zip_buffer.seek(0)
+    return zip_buffer, report_data['patient_id'].tolist()

logic/financial_model.py

@@ -0,0 +1,26 @@
+# File: logic/financial_model.py
+
+def estimate_financial_recovery(care_gaps_df, patient_data, base_rate, sdoh_keys):
+    results = []
+    hcc_weights = {
+        'HCC18': 0.4, 'HCC85': 0.5, 'HCC19': 0.3
+    }
+
+    for _, row in care_gaps_df.iterrows():
+        patient = patient_data[patient_data['patient_id'] == row['patient_id']].iloc[0]
+        hccs = [h.strip() for h in str(patient.get('hcc_codes', '')).split(';') if h.strip() in hcc_weights]
+        risk_score = sum(hcc_weights[h] for h in hccs)
+
+        sdoh_count = sum(1 for key in sdoh_keys if str(patient.get(key, '')).strip().lower() == 'yes')
+        sdoh_modifier = 1 + 0.05 * sdoh_count  # Each SDOH adds 5% to the opportunity
+
+        care_gap_count = len(row['care_gaps'])
+        expected_gain = base_rate * (risk_score * 0.02 * care_gap_count) * sdoh_modifier
+
+        results.append({
+            'patient_id': row['patient_id'],
+            'risk_score': round(risk_score, 2),
+            'projected_gain': round(expected_gain, 2)
+        })
+
+    return pd.DataFrame(results)

logic/nlp_report.py

@@ -0,0 +1,68 @@
+# File: logic/nlp_report.py
+
+from huggingface_hub import hf_hub_download
+from llama_cpp import Llama
+import logging
+
+MODEL_REPO = "bartowski/deepcogito_cogito-v1-preview-llama-8B-GGUF"
+MODEL_FILENAME = "deepcogito_cogito-v1-preview-llama-8B-Q8_0.gguf"
+
+model_path = hf_hub_download(
+    repo_id=MODEL_REPO,
+    filename=MODEL_FILENAME,
+    cache_dir="models"
+)
+
+llm = Llama(
+    model_path=model_path,
+    n_ctx=2048,
+    chat_format="chatml",
+    verbose=False
+)
+
+def generate_patient_summary(patient_row, tone="executive", override_prompt=None):
+    try:
+        gaps = patient_row['care_gaps']
+        gain = patient_row['projected_gain']
+        score = patient_row['risk_score']
+
+        social_flags = []
+        if str(patient_row.get("Lives_Alone", "")).strip().lower() == "yes":
+            social_flags.append("lives alone")
+        if str(patient_row.get("Housing_Instability", "")).strip().lower() == "yes":
+            social_flags.append("experiencing housing instability")
+        if str(patient_row.get("Transportation_Access", "")).strip().lower() == "yes":
+            social_flags.append("has limited transportation access")
+        if str(patient_row.get("Food_Insecurity", "")).strip().lower() == "yes":
+            social_flags.append("faces food insecurity")
+        if str(patient_row.get("Primary_Care_Established", "")).strip().lower() == "no":
+            social_flags.append("does not have a primary care provider")
+        if str(patient_row.get("FollowUp_Scheduled", "")).strip().lower() == "no":
+            social_flags.append("has no follow-up scheduled")
+
+        social_context = "; ".join(social_flags) if social_flags else "no major social risk factors identified"
+
+        base_prompt = override_prompt or f"""
+You are a healthcare analyst creating summaries for value-based care programs.
+Tone: {tone}
+
+Patient ID: {patient_row['patient_id']}
+Risk Score: {score}
+Projected Revenue Gain: ${gain}
+Care Gaps Identified: {', '.join(gaps)}
+Social Risk Factors: {social_context}
+
+Write a concise and insightful clinical summary, including key gaps and social considerations. Offer targeted recommendations.
+"""
+
+        response = llm.create_chat_completion(
+            messages=[{"role": "user", "content": base_prompt}]
+        )
+
+        if "choices" in response and response["choices"]:
+            return response["choices"][0]["message"]["content"]
+        else:
+            return "[Error] No summary returned."
+    except Exception as e:
+        logging.error(f"LLM error: {e}")
+        return "[Error generating summary]"