Research
Studying Appointment Waitlist Offer Windows and Booking Errors
A cohort design for assessing whether explicit waitlist offer windows fill cancellations without increasing overlap, confusion, or corrections.
Aim and cohort definition
This study asks how a defined offer window affects cancellation-fill rate and booking errors. Include appointment openings offered through the documented waitlist workflow. Exclude slots filled through ordinary scheduling before a waitlist offer was created, and report exclusions by reason.
The unit of analysis is one open slot, with linked offers ordered by timestamp. A pilot should verify that the scheduling system can distinguish an offer from a confirmed reservation.
Measures and protocol
The primary outcome is the proportion of eligible openings filled from the waitlist before the slot begins. Guardrails include overlapping promises, duplicate bookings, customer corrections, and offers sent after a slot was filled. Secondary measures include response time, expired offers, and number of contacts per filled slot.
Choose offer-window rules before the test and keep them stable within each comparison period. Record the slot type, lead time, contact channel, and offer expiration. Avoid collecting unrelated patient or customer details in the analysis file.
Analysis approach
Compare windows within similar appointment types and lead-time bands. Present counts and denominators alongside rates. Use survival or time-to-response summaries only when sample size and data quality justify them; a clear distribution may be more useful than a complex model.
Inspect every overlap or double-booking event. Average fill rate cannot compensate for a serious scheduling error. Repeat the analysis with unknown responses treated as declines and then as accepts to show sensitivity to missing records.
Operational interpretation
Adopt the tested window when fill rate improves by a preset practical amount and guardrails remain stable. Shorten or lengthen it only through a new controlled period. Results depend on appointment type, lead time, customer channel, and local scheduling rules, so they should not be generalized beyond the observed cohort.