Project Overview
This project analyses live NHS England A&E Quality Indicators data covering 140 active NHS Trusts over 25 months (December 2023 to December 2025). The analysis tracks four core operational KPIs — total wait time, treatment start time, ambulance assessment time, and 7-day reattendance rate — to identify which Trusts are under the most operational pressure and whether performance is improving system-wide.
Business Question
Does treatment delay in NHS A&E departments correlate with unplanned patient reattendance? And which Trusts require immediate operational attention versus which reflect a broader, improving national trend?
Key Insights & Findings
Blackpool Teaching Hospitals — A Severe Outlier
Blackpool Teaching Hospitals recorded the longest average A&E wait in the dataset, at 352 minutes — 83 minutes above the next-worst Trust and more than double the national median of 169 minutes. This scale of divergence points to a Trust-specific operational issue rather than a regional or seasonal pattern.
National Wait Times Are Stabilising
England's national median A&E wait fell from 186 minutes in December 2023 to 157 minutes by August 2024, and has held steady in a narrow 159–167 minute range since spring 2025 — evidence that system-wide performance has genuinely improved and stabilised, not just dipped temporarily.
Treatment Speed Does Not Predict Reattendance
Across 135 Trusts, treatment speed showed no meaningful correlation with reattendance rate. Northumbria Healthcare and Maidstone and Tunbridge Wells have among the fastest treatment times in the dataset, yet the two highest reattendance rates — 15.2% and 14.7%. This suggests reattendance is driven by factors beyond how quickly patients are treated, such as discharge process or case-mix.
Methodology & Approach
The analysis followed a full data pipeline built in Python across four consolidated notebooks:
- Data Preparation: Loaded and cleaned three raw NHS source files, removed the national ENG aggregate row and closed Trusts, handled suppressed values (5-or-fewer patient counts) without dropping them, and joined into a master analytical dataset
- Analysis: Derived performance flags, excluded denominator KPIs from breach-rate calculations, and produced regional and Trust-level aggregations
- Visualisation: Built four publication-grade charts using Matplotlib and Seaborn
- Reporting: Generated a 7-sheet Excel report via openpyxl for stakeholder handover
Every stage was independently verified against the raw data — including catching and correcting a mislabeled KPI (a negative-value anomaly initially attributed to the wrong measure) and an inflated breach-rate calculation caused by unfiltered denominator KPIs.
Tools & Technologies
Business Recommendations
Strategic Actions Based on Findings
- Commission a focused operational review of Blackpool Teaching Hospitals — 352-minute mean wait, 83 minutes above the next-worst Trust. Case-mix and demand profile should be reviewed alongside operational factors.
- Investigate what factors — operational or seasonal — contributed to the national wait time stabilisation from spring 2025 onward, and confirm the trend holds at regional level, not just nationally.
- Investigate the 27 Trusts with reattendance rates above 10% for shared clinical or discharge-process causes, since treatment speed is confirmed not to be the driver.
Impact & Value Delivered
This analysis provides NHS operational leadership with:
- Clear identification of the single highest-risk Trust requiring immediate review
- Confirmation that national-level improvement is real and sustained, not a temporary dip
- Evidence that reattendance is not explained by treatment speed alone — redirecting investigation toward discharge and clinical process factors
- A fully reproducible, documented analytical pipeline suitable for handover to another analyst