#7 - Predicting No-Shows: The Surprising Science Behind Missed Appointments

Code & Cure

Why do so many doctor’s appointments end in empty waiting rooms? Nearly one in four scheduled visits turn into no-shows, disrupting care, wasting resources, and straining already overburdened systems. But a new study shows we might be able to see these gaps coming—and stop them.

By analyzing over a million healthcare visits, researchers used machine learning to uncover surprising predictors of missed appointments. The top signal? How far in advance the appointment was booked. Appointments scheduled more than 60 days out had the highest odds of being missed—more telling than age, income, or insurance status. Other key factors included continuity with the same provider, a patient’s past attendance, distance to the clinic, and even the weather.

This episode unpacks how models like random forests and gradient boosting sift through massive datasets to identify no-show risks—not just for populations, but for individual patients. These insights open the door to smarter, more personalized interventions: tighter scheduling windows, transportation support, or ensuring patients see familiar faces.

Tune in to explore how AI could help healthcare systems run smoother, deliver more timely care, and keep more patients from vanishing in the first place.

References:

Predicting Missed Appointments in Primary Care: A Personalized Machine Learning Approach
Wen-Jan Tuan, Yifang Yan, Bilal Abou Al Ardat, Todd Felix and Qiushi Chen
Annals of Family Medicine, July/August 2025

Credits:

Theme music: Nowhere Land, Kevin MacLeod (incompetech.com)
Licensed under Creative Commons: By Attribution 4.0
https://creativecommons.org/licenses/by/4.0/

บท

1. Introducing Missed Appointments Prediction (00:00:00)

2. Key Factors in Appointment No-Shows (00:01:24)

3. Understanding Machine Learning Models (00:04:15)

4. How Decision Trees and Forests Work (00:10:52)

5. Evaluating Model Accuracy (00:18:04)

6. Most Influential Predictive Factors (00:23:09)

7. Turning Predictions Into Interventions (00:28:42)