Key takeaways
- Stream scheduling gives each patient their own slot; it is predictable for patients but sensitive to late arrivals and long visits.
- Wave scheduling books several patients at the start of a period and sees them in turn; it absorbs no-shows but makes some patients wait.
- Modified wave, two patients at the start of the hour and singles after, is a common middle ground.
- Choose by visit length variability, no-show rate, rooming capacity and how much waiting your patients will tolerate.
The three models in plain terms
Stream scheduling is what most people picture: one patient every 15 or 20 minutes, each with their own time. Wave scheduling books a group of patients at the start of a period, say three at 9:00, and the team sees them in order of arrival or readiness during that hour. Modified wave sits in between: two patients at the top of the hour and single patients at set intervals after, leaving a gap later in the hour to catch up.
Each model is a bet about variability. If every visit took exactly the scheduled time and every patient arrived on time, stream would be perfect. In reality, visit lengths vary widely. The mean time spent with the physician at a US office visit was 23.5 minutes in 2019, but the median was 19.4 minutes[1]. A mean well above the median means a minority of long visits pull the average up, and those long visits are what make stream schedules run late.
Source: CDC National Center for Health Statistics, 2019[1]
Show the data
| Item | Value |
|---|---|
| Mean | 23.5 min |
| Median | 19.4 min |
Example day sheets
| Time | Stream | Modified wave |
|---|---|---|
| 9:00 | Patient A (20 min) | Patients A and B |
| 9:20 | Patient B (20 min) | Patient C |
| 9:40 | Patient C (20 min) | Open for catch-up or same-day |
| 10:00 | Patient D (20 min) | Patients D and E |
| 10:20 | Patient E (20 min) | Patient F |
| 10:40 | Patient F (20 min) | Open for catch-up or same-day |
Example: both sheets see six patients in two hours. Under stream, if Patient B's visit runs 15 minutes long, every later patient waits 15 minutes unless someone else is quick. Under modified wave, the team starts with two patients roomed at once, so the provider moves between them while the medical assistant prepares the next, and the open slot at 9:40 absorbs most overruns. If someone does not show, the wave still has a patient ready.
Side-by-side comparison
| Factor | Stream | Wave | Modified wave |
|---|---|---|---|
| Patient waiting time | Lowest when the day runs on time | Highest for the last patient in each wave | Moderate |
| Effect of a no-show | Provider idle for the slot | Absorbed by the next patient in the wave | Partly absorbed |
| Effect of a long visit | Delay passes to every later patient | Absorbed within the wave | Absorbed by the catch-up slot |
| Rooming and staff needs | One room and one MA per provider can work | Needs extra rooms and quick rooming | Needs two rooms ready at the start of each hour |
| Patient perception | Clear appointment time | Can feel like being double-booked | Usually acceptable if waits stay short |
| Best fit | Predictable visits, low no-shows | High no-shows, very variable visits | Most mixed primary care and specialty clinics |
When stream scheduling fits
- Visit lengths are predictable, such as procedure clinics with standard times or highly protocolized follow-ups.
- No-show rates are low, so empty slots are rare.
- Patients expect and value an exact time, such as employed patients fitting visits into a work day.
- The practice has only one room and one assistant per provider, so rooming two patients at once is not possible.
- The team is disciplined about start times and keeps visits within their slots.
Stream also makes waiting time easy to measure, because each patient has a clear scheduled time. If you choose stream, protect it with small buffers and a firm late-arrival policy, or one late patient early in the session will cascade through the day.
When wave or modified wave fits
- No-shows are frequent enough that empty slots regularly leave providers idle.
- Visit lengths vary widely, with some short visits and some long ones in the same session.
- The clinic has extra exam rooms and a medical assistant who can room patients quickly.
- Patients are often early or late, so strict slot times are hard to keep anyway.
- The practice wants to keep a catch-up slot without leaving it empty when the day runs on time.
No-shows are a live concern for many practices. In MGMA's 2025 poll, 27% of practices said no-show rates had increased versus the prior year, 60% said they stayed the same and only 13% said they fell[2]. And no-shows were the most common top patient access focus for 2026, chosen by 27% of practice leaders[3]. A model that tolerates no-shows is worth considering when reminders and policies have already been tightened. The pressure continued into 2026, when 32% of medical groups reported higher no-show rates year to date than in 2025[4].
A fourth option: clustering similar visits
Some practices get the benefits of a wave without changing appointment times by clustering similar visits. All annual physicals go in one session, all diabetes follow-ups in another, all procedures on one afternoon. The team sets up once for that visit type, the medical assistant runs the same rooming routine back to back, and supplies and forms are ready.
Clustering works best for visit types with predictable steps and enough volume to fill a session. It can sit inside a stream or a wave model. The risk is reduced flexibility for patients, since a patient who can only come on Tuesday may not find their visit type available, so most practices cluster only part of each week.
Measuring the effect of a change
Pick measures before switching models so the comparison is fair. Use the same four for the pilot provider and a comparison provider who stays on the old model.
- Waiting time from check-in to rooming, from EHR or check-in timestamps.
- Day-end delay: how late the last patient of each session is roomed compared with the scheduled time.
- Visits completed per session, to check that capacity did not fall.
- Patient comments about waiting, from a short checkout question or survey.
Run the pilot long enough to include busy and quiet weeks, usually four to six weeks, and review it with the provider, the medical assistant and the scheduler together.
Decision checklist
Measure visit length spread
From your EHR timestamps, compare scheduled and actual visit times by visit type. A wide spread favors wave or modified wave.
Measure no-shows by session
If several sessions a week lose two or more patients, a wave model may recover that time.
Count rooms and assistants
Wave models need at least two rooms ready at the start of each period and fast rooming. Without them, wave becomes a crowded waiting room.
Ask patients
A short survey or comments at checkout tell you how much waiting your patients will accept and whether they would prefer an exact time.
Pilot with one provider
Run modified wave for one provider for four to six weeks and compare waiting time, day-end delay and visits completed with a provider on stream.
Telling patients about wave appointments
Patients booked into a wave may arrive with another patient booked at the same time. If they notice, they may assume the practice is overbooking. Be transparent without making it complicated.
Script: “Your appointment is at 9:00. Please arrive by 8:50 so we can check you in. We see patients in the order they are ready, and most people are with the doctor within 20 minutes of their appointment time.”
Only promise a wait you can keep. If waits in a wave regularly exceed what you tell patients, the wave is too large or rooming is too slow.
Front desk staff should also know how to answer the patient who says, “Someone else was booked at nine too.” A short, honest reply works: “We book a couple of patients at the start of each hour so the doctor is never waiting, and we see people in the order they are ready.” Most patients accept that if the wait that follows is short.
Worked Example: switching one provider to modified wave
Example: a fictional orthopedic practice has one surgeon whose clinic runs about 40 minutes late by noon on most days, with one or two no-shows per session. New patient visits vary from 15 to 50 minutes. The practice tries modified wave: two patients at the top of each hour, one at 20 past, and a catch-up slot at 40 past that can be used for same-day post-operative checks.
Example results after six weeks: average day-end delay falls from about 40 minutes to about 15, the surgeon sees the same number of patients, and post-operative patients get seen faster because the catch-up slot is used for them when the day is on time. Average waiting time from check-in to rooming rises slightly for patients in the first position of each hour. The practice keeps the model for this surgeon and leaves two other providers on stream.
Modified wave is not the same as routine double-booking, which is covered in when double-booking works and when it fails. For the template that sits under any model, see how to build provider scheduling templates, and for same-day demand, open access scheduling. New patients, who made up 16.8% of US office visits in 2019[1], usually belong in the first position of a wave so their longer visit starts on time.
Frequently asked questions
What is wave scheduling in a medical office?
Wave scheduling books several patients at the start of a time period, such as the top of each hour, and sees them in order of arrival or readiness during that period.
What is modified wave scheduling?
It books two patients at the start of each hour, then single patients at set intervals, with a gap later in the hour for catch-up or same-day needs.
Is wave scheduling the same as double-booking?
Not exactly. Wave scheduling is a planned model with the template, rooms and staffing built around it. Double-booking usually means adding a second patient to a slot designed for one.
Which model reduces patient waiting time most?
Stream scheduling gives the lowest waits when the day runs on time. When visits vary and no-shows are common, modified wave often reduces day-end delays enough that overall waits fall.
References
CDC National Center for Health Statistics (2019). National Ambulatory Medical Care Survey: 2019 National Summary Tables (opens in a new tab) ↑
Medical Group Management Association (MGMA) (2025). Patient no-shows in 2025: What's changing and what to do about it (opens in a new tab) ↑
Medical Group Management Association (MGMA) (2025). Patient access priorities for 2026: Tackling wait times, phones, no-shows and more (opens in a new tab) ↑
Medical Group Management Association (MGMA) (2026). About 1 in 3 medical groups see higher no-shows in 2026 as patients face higher costs (opens in a new tab) ↑