Predict No-Shows in a Network of Dental Clinics
A chain of 80 clinics loses 14% of appointments to patients who do not show up: empty chairs, idle hygienists and €3.1M a year of lost revenue. Management wants to decide for each appointment whether to send an extra reminder, ask for confirmation or do controlled overbooking. Clinic managers must understand and accept the criterion.
Where the money comes from
What to do now
- 1Measure the cost of an empty chair hour by specialty.
- 2No-show model with 2 years of data.
- 3Pilot of reminders and overbooking in 10 clinics versus a control.
For your committee
One page with the decision, the impact, the risks, the steps and the KPI to track.
The print dialog opens: choose "Save as PDF".
Watch out: more volume is not always more profit
Aggressive overbooking fills the schedule but creates waits and negative reviews that cost future patients. Only overbook where the no-show probability justifies it.
KPI for the committee
Chair occupancy and no-show rate.
Detail for your data team
Proposed technique: Logistic Regression
Why: Simple tabular data, a binary decision and the need for managers to trust the model: Logistic Regression gives well-calibrated probabilities (essential for deciding overbooking by adding up probabilities) and coefficients that can be explained in a meeting.
Cost of getting it wrong: An undetected no-show is a lost chair hour (~€90); a false positive with overbooking is a patient waiting and a bad review.
Data needed: History of 1.2M appointments with outcome (attended / did not attend).; Appointment data: day, time, booking lead time, type of treatment, practitioner.; Patient history: previous no-shows, tenure, whether they have a financed treatment in progress.; Context: distance to the clinic, forecast rain.