Skip to content

Latest commit

 

History

History
55 lines (33 loc) · 3.15 KB

README.md

File metadata and controls

55 lines (33 loc) · 3.15 KB

NHS Patient, Service, and Twitter Data Analysis

🔦 Disclaimer: I completed the following project as part of the LSE Data Analytics Career Accelearator Course (April 2024 - November 2024)

🔖 Grade 82%

NHS

❓ Can NHS GP services achieve their patient-focused and financial goals by better utilising existing resources?

📌 Overview

For this project, I worked with real-world NHS data to explore how GP services could optimise resources while balancing patient care and financial sustainability. This was a challenging but rewarding analysis, as I dealt with both structured healthcare data and unstructured social media data.

I analysed three datasets on GP appointments and service performance, along with NHS-related tweets scraped from X (formerly Twitter). My goal was to identify seasonal trends in patient appointments, consultation methods, and wait times while also assessing public sentiment about NHS services.

🛠️ Approach & Tools

1. Data Cleaning & Wrangling

  • Processed structured NHS GP appointment data (2020–2022) using Python (Pandas, NumPy).
  • Removed 21,604 duplicate records and aligned NHS regional codes for accurate geographic analysis.

2. Exploratory & Descriptive Analysis

  • Conducted time-series analysis to identify seasonal patterns, wait times, and consultation trends.
  • Used Matplotlib & Seaborn for visualising NHS appointment trends.

3. Consultation & Resource Analysis

  • Examined face-to-face vs. telephone consultations to assess COVID-19’s impact on patient behaviour.
  • Analysed the underutilisation of Primary Care Networks (PCNs) as a potential solution for reducing GP workloads.

4. Social Media Sentiment Analysis

  • Scraped and analysed NHS-related tweets by engagement metrics (hashtags, retweets, and favourites) using Pandas to assess public sentiment and engagement.
  • Found that NHS-related hashtags were underutilised, highlighting an opportunity to improve social media outreach.

📊 Business Impact

  • Identified seasonal appointment fluctuations, with higher demand in autumn and lower volumes in winter due to holiday closures.
  • Found that Monday appointments were the highest, likely due to weekend backlog.
  • Determined that same-day appointments remained high despite COVID-19, but next-day availability was very low, indicating resource strain.
  • Discovered a shift from face-to-face to telephone consultations during COVID-19, with in-person visits slowly recovering by 2022.
  • Twitter analysis revealed a lack of NHS-related hashtags in public discourse, suggesting a gap in patient engagement via social media.

🎯 Key Takeaways

  • Gained hands-on experience working with large-scale healthcare data, ensuring data integrity and accuracy.
  • Developed time-series analysis skills, identifying patterns in appointment volumes and consultation methods.
  • Strengthened my ability to analyse public sentiment, using real-world social media data to complement structured datasets.
  • Learned how to translate healthcare analytics into recommendations, focusing on improving patient access and resource distribution.

🔙 Return to Portfolio