What counts as a no-show?
A no-show, or did not attend (DNA), is a booked appointment that was not cancelled and where the patient did not arrive. A late cancellation is a cancellation made inside your notice window, such as the period stated in your cancellation policy. Write down how you treat edge cases before you count: a patient who arrives too late to be seen, a telehealth patient who never joins, a patient who rings at the start time to rebook, and a group session where some participants do not come. A figure that changes definition from month to month cannot show a trend, so the written rule matters more than which option you pick.
- Did not attend: booked, not cancelled, patient absent
- Late cancellation: cancelled inside your notice window
- Cancelled with notice: left out of the no-show rate
- Arrived too late to be seen: choose DNA or attended, and record the rule
Which appointments belong in the denominator?
Use appointments that were still booked when their start time arrived: those attended plus those marked DNA. Leave out appointments cancelled with notice, blocked time, meetings, administrative bookings and test entries. Some reports divide DNAs by every appointment ever booked, including those cancelled weeks earlier. That gives a smaller percentage from the same events, which is why figures from different sources rarely compare. Either method can work if you use it every month. If you need to compare with a published figure, find out which denominator it used first. The table below shows how much the choice moves the result.
Where do the numbers come from in your practice software?
Each system stores non-attendance differently, so read its help pages before you trust a report. In Cliniko, the Missed appointments report lists appointments marked Did not arrive alongside cancellations, but filters DNAs by appointment date and cancellations by cancellation date. Cliniko's appointments schedule report leaves out appointments marked Did not arrive, so a schedule count gives attended appointments only and you add DNAs back to get the denominator. In Halaxy, the Non-attendance option sets the status to Did Not Attend and keeps the invoice, while deleting an appointment removes it from the calendar along with its invoice and linked notes.
For Best Practice, MedicalDirector, Nookal, Splose or any other system, check the status names your practice uses and run a test report on a week you can verify by hand against the appointment book.
Which counting traps distort the rate?
The arithmetic is simple; the errors come from how events are recorded. If reception deletes a no-show booking to free the slot, the DNA disappears from every report. If a no-show is recorded as a cancellation, it lands in the wrong rate. Date filters that differ between DNAs and cancellations can place related events in different months. Before you report a figure, take a sample of appointments from the month and check each status against the appointment book and any notes. Fix the recording habit as well as the number, or the same error returns next month.
- Bookings deleted instead of marked DNA
- No-shows recorded as cancellations
- Different date filters for DNAs and cancellations
- Group sessions counted once instead of per participant
- Telehealth no-shows recorded inconsistently
How should you break the rate down?
A single practice-wide figure tells you whether there is a problem; the breakdown tells you where. Split the same month's data by practitioner, appointment type, day and time, and lead time, meaning the number of days between booking and appointment. Add new versus returning patients and, if you send reminders, the reminder outcome: confirmed, no reply or failed delivery. Each split points to a different fix. A high rate on long-lead bookings suggests a reminder or confirmation gap, while a high rate on one appointment type may mean patients are unclear about what the appointment is for.
What should you do with the figure each month?
Record the monthly rate in one place with the definition beside it, and review it with the same people each month. Change one thing at a time, such as reminder timing, a confirmation request or a waitlist process for short-notice gaps, and compare the following two or three months with the earlier trend. Use your own average fee per appointment if you want to put a dollar value on the gap; the no-show cost calculator on this site does that arithmetic. Keep the counting rules fixed while you test, or you will not know whether the change or the counting moved the number.