🏥 Lab Turnaround Time Optimization

📌 Overview

Laboratory turnaround time (TAT) is one of the most critical performance metrics for clinical labs.
This project analyzes synthetic lab workflow data to identify bottlenecks across pre-analytical, analytical, and post-analytical stages.
The goal is to reduce delays, improve SLA compliance, and optimize resource allocation.

📂 Dataset

• File: data/lab_events.csv
• Size: ~20,000 synthetic lab events
• Columns include:
  • Patient/test info (order_id, patient_id, test_code, priority)
  • Workflow stages (collection_time, receipt_time, start_analysis_time, verification_time, report_time)
  • Operational context (bench, instrument_id, shift, weekday)
  • QA indicators (canceled, recollect_flag)

🔒 Note: This dataset is synthetic and safe to share publicly. In practice, the same analysis can be applied to real, de-identified lab data.

📊 Key Metrics

• Total TAT: collection → report
• Stage TATs: pre-analytical, analytical, post-analytical
• SLA hit rate: % of tests meeting defined TAT thresholds
• 95th percentile TAT: robust measure of outliers
• Shift/bench comparisons: where bottlenecks occur

🚀 Usage

Run Locally

git clone https://github.com/gortegam/Lab-Turnaround-Time-TAT-Optimization
cd lab-tat-optimization
pip install -r requirements.txt
streamlit run app.py

The dashboard will launch at http://localhost:8501

Live Demo

👉 Try the Streamlit Dashboard

(If inactive, fork this repo and deploy on your own Streamlit Cloud account.)

---

📈 Dashboard Features

• Filters: by test type, shift, and priority
• KPIs: Median TAT, 95th percentile TAT, SLA hit rate
• Breakdowns:
  • TAT by test (median vs 95th percentile)
  • TAT by shift
  • TAT by priority
• Trends:
  • Daily median TAT over time
  • Daily SLA compliance rate

---

📑 Repo Structure

lab-tat-optimization/
├─ data/
│  └─ lab_events.csv            # synthetic dataset
├─ notebooks/
│  └─ 01_eda_tat_baseline.ipynb # baseline EDA & bottleneck analysis
├─ app.py                       # Streamlit dashboard
├─ requirements.txt
└─ README.md

---

🔍 Preliminary Insights (from synthetic dataset)

• CBC and CMP tests show shorter TATs (median <2h) vs Pathology Review (median ~2 days).
• Evening shift experiences the highest delays (median TAT ↑ 20% vs Day).
• STAT requests cut pre-analytical time in half, but increase routine backlog.
• Instrument InstC has a higher cancellation rate, contributing to TAT outliers.

(Replace with real findings as analysis progresses.)

---

🛠️ Next Steps

• Add root cause analysis (bench & instrument level).
• Simulate staffing changes (queueing model).
• Generate optimization recommendations (e.g., shift coverage adjustments).

---

📜 License

This project is released under the MIT License.