lab_report_01.md

Gaps: Apollo Report vs structured_data (Supabase)

Coverage

• Structured lab metrics: ~98%
• Overall report completeness: ~95–96%

---

Missing Data

1. Lab Summary (Page 3)

• Organ-system grouping (Blood, Endocrine, Cardiac, Renal, etc.)
• Counts by status: Normal / Needs Attention / Requires Action

2. “Parameters Needing Attention” (Page 4)

• Grouped abnormal parameters
• Associated clinical descriptions for each flagged metric

3. Urine Panel (Page 6)

Missing fields:

• Transparency
• Nitrite
• Leukocyte Esterase
• Pus Cells
• Epithelial Cells

4. CBC Extension

• Corrected TLC

5. Panel Metadata

• “Sample Collected On” (per panel)
• Panel grouping (CBC, LFT, KFT, Lipid, Thyroid)

---

Missing Clinical Context

1. Parameter Descriptions

• HbA1c, Urea, Triglycerides, Vitamin D, LFT explanations

2. Interpretation Bands

• HbA1c classification (Normal / Prediabetic / Diabetic)
• Lipid risk categories
• BMI classification

---

Missing Risk Model Structure

1. AI Risk Scores (Page 11)

Incomplete structure:

• Model descriptions (AICVD, Prediabetes, COPD)
• Basis (inputs used)
• Accuracy statements
• Time horizon (e.g., COPD 3-month risk)

2. Risk Interpretation

• “Your score vs acceptable score” comparison
• Narrative explanations

---

Schema Gaps

1. Flat Structure

• No separation of:
  • labs
  • vitals
  • risk_scores
  • clinical_notes

2. Mixed Data Types

• Metrics, risk scores, and clinical notes stored under a single metrics object

---

Summary

• Missing primarily:

  • Summary layers
  • Clinical explanations
  • Risk model context
  • Panel-level structure

• structured_data is complete for raw values but lacks:

  • interpretability
  • grouping
  • clinical intelligence