Field notes
PMS
Field notes
HMS
Why OPD queues still bunch up at 11 a.m., and a fix that worked
Every OPD we have measured has the same shape: a calm 9 a.m., a brutal 11 a.m. peak, a thin afternoon, and a small evening surge. The peak is not random, it is the predictable result of how appointments are booked.
We tried four fixes at a 60-bed multi-speciality hospital running ElyHMS. Only one of them actually flattened the curve.
The thing that did not work: “just add more slots”
Adding more 11 a.m. slots is the obvious move and the wrong one. Patients book the slot they want, not the slot you have. More slots at the peak time just attract more peak-time bookings.
The thing that did not work: SMS reminders
Reminders reduce no-shows by about 8-12% in our data, which is real money, but they do not change when people arrive. They arrive at the same hour, just more reliably.
The thing that did not work: doctor-side overbooking
Overbooking the 11 a.m. slot to compensate for no-shows works on the day no one cancels and explodes on the day everyone shows up. We saw a near-mutiny from one consultant on day three.
The thing that worked: queue-position SMS at T-30 minutes
Thirty minutes before each appointment, we sent the patient their live queue position and an estimated time-to-doctor. Patients with a number greater than three started arriving twenty to thirty minutes later, voluntarily. The 11 a.m. peak dropped by 41% in three weeks, and the afternoon hours filled in. No one was forced; they just got better information.
Related reading
More from the field notes stack.
AI agents
Playbook