Marie Camara | BI Analyst

Reducing No-Shows in Healthcare Platforms

appointment

Industry: Healthcare

Role: Healthcare Data Analyst

Business problem

Digital health platforms have valuable appointment data, yet appointment no-shows continue to drain revenue, waste clinical time, and hinder access to care for other patients.

Objectives

  • Identify key factors driving no-shows across specialties and patient segments
  • Analyze appointment and patient behavior through six analytical dimensions
  • Provide actionable, quantified recommendations to reduce no-show rates.

Methodology and Tools Used

A structured five-stage pipeline — from synthetic dataset construction through interactive Power BI delivery.

Data Collection: Synthetic dataset built to simulate real appointment behaviours on digital health platforms — 100 records, 10 variables covering demographics, appointment details, and behavioral indicators.

Data Cleaning: Verified data types, checked for missing values, and ensured consistency across categorical fields, including Gender, Specialty, SMS Reminder, and No-Show status.

Data Transformation: Created calculated columns in Power BI: age groups, booking delay bins (0–3 / 4–7 / 8–14 / 15+ days), and patient risk segmentation based on prior no-show history.

Data Analysis: Built DAX measures (No-Show Rate, Revenue Loss, Total Appointments) and performed segmentation analysis across 6 dimensions: specialty, time slot, delay range, SMS, age group, and prior history.

Data Visualization: Designed a 3-page Power BI dashboard — Executive Overview, Drivers Analysis, and Insights & Recommendations — with slicers, a risk heatmap, and a decomposition tree.

Tools Used: Microsoft Excel, Power BI, Canva.

Key Deliverables

Every deliverable was designed to support immediate decision-making.

graph

Key Insights

  • Prior no-shows predict a 66% miss rate.
  • SMS reminders reduce no-shows by 18 points, but 49% of appointments lack the
  • Booking over 14 days increases risk nearly twofold.
  • 14:00 and 11:00 are the highest-risk slots; 10:00 outperforms every other time
  • Dermatology is a structural outlier at 71% — specific slots reach 100%
  • The 25–59 age group drives no-shows by rate and volume, not just under-25s.

Recommendations

  • Automated Risk-Based SMS Triggering
  • Active Reconfirmation for Long-Lead Bookings
  • Dermatology Overbooking Pilot
  • Diagnose the SMS Gap Before Buying New Tools.