How do you calculate the no-show rate?
No-shows divided by the appointments that should have happened in the period. If 300 appointments were booked in the month and 45 patients didn't come, the rate is 15%.
An appointment canceled ahead of time counts on neither side: it wasn't going to happen, and the slot could be filled.
What is the most common mistake?
Counting as attended the appointment that passed and nobody recorded. No-shows are usually marked by exception, and the rest stays booked forever. If that rest counts as attended, the rate drops without anyone showing up.
The honest math uses only appointments with a recorded outcome (attended or no-show) and shows separately how many were left without one.
How many appointments before the rate means something?
With few, one no-show moves the rate by several points. Below about ten appointments with an outcome, look at the count (2 of 8), not the percentage.
How do you compare before and after a change?
Two windows of the same length, with the same rule: for example, the first 30 days of the schedule against the last 30. Comparing a holiday month with a busy month measures the season, not the change.
And the comparison shows what changed, not why. If the rate dropped, several things may have changed at the same time.