Key takeaways
- Split the visit into intervals: arrival to check-in, check-in to rooming, rooming to clinician, and checkout. Each has a different owner and fix.
- Measure from the later of arrival time and appointment time, so early arrivals do not make waits look worse than they are.
- Report the median and the 90th percentile, not the average. A few very long waits distort the mean.
- Tell waiting patients the truth early. An honest update at 15 minutes prevents most complaints.
Why wait time needs measuring, not guessing
Most practices know they have a waiting-room problem only when complaints arrive. By then, staff have a story about the cause, and the story is often wrong. Measuring the intervals inside a visit shows where time actually goes: at the desk, in the lobby waiting to be roomed, or in the exam room waiting for the clinician.
Wait times are on practice leaders’ minds. When MGMA asked about the top patient access focus for 2026, 21% named wait times[1]. And the clinical part of the visit is short by comparison: in 2019, the mean time spent with the physician at a US office visit was 23.5 minutes and the median 19.4 minutes[2]. A 30-minute wait before a 20-minute visit is what patients remember.
The intervals and who owns them
| Interval | Start and end | Captured by | Usual bottleneck | Owner |
|---|---|---|---|---|
| Desk | Arrival to check-in complete | Kiosk, arrival log or desk timestamp | Insurance problems, forms not done, one person at the desk | Front desk lead |
| Lobby | Check-in complete to roomed | EHR status change when the patient is called back | Rooms not turned over, medical assistant covering two providers, running behind | Clinical lead |
| Exam room | Roomed to clinician in | EHR status or room flag | Clinician running late, documentation from the previous visit | Provider and clinical lead |
| Visit | Clinician in to clinician out | EHR status | Visit type booked too short for the need | Provider |
| Checkout | Clinician out to departure | Checkout timestamp | Scheduling follow-ups, collecting payment, queue at checkout | Front desk lead |
The patient experiences the total, but each interval has its own cause and owner. A practice that tries to fix the total without the split usually ends up blaming the front desk for a delay that happens in the exam room, or the reverse.
Getting the timestamps
Most EHR and practice management systems record status changes, such as arrived, checked in, roomed and checked out, if staff click them at the right moment. The data is only as good as the clicks. Before relying on it, watch a morning and compare the clicks with what actually happened.
If the system cannot capture an interval, use a paper time sheet for two weeks. A simple card that travels with the patient, with boxes for arrival, check-in complete, roomed, clinician in, clinician out and departure, works well. Staff write the time from a single clock. Two weeks of cards is enough to find the main bottleneck.
One rule matters more than any other: measure from the later of the arrival time and the appointment time. A patient who arrives 20 minutes early and is roomed on time did not wait 20 minutes for the practice. Record early arrival separately; it tells you whether patients expect delays.
Use the median and the 90th percentile
Wait times are skewed. Most patients wait a little, and a few wait a long time. Those few pull the average up, so the average describes almost nobody. Emergency department data shows the effect clearly: in 2022, the median wait to see a physician, nurse practitioner or physician assistant was 16.0 minutes, but the mean was 38.1 minutes[3]. Only 40.6% of emergency department visits had the patient seen in fewer than 15 minutes[4].
Source: CDC National Center for Health Statistics, 2022[3]
Show the data
| Item | Value |
|---|---|
| Median wait | 16 min |
| Mean wait | 38.1 min |
For a practice, report two numbers for each interval: the median, which describes the typical patient, and the 90th percentile, which describes the bad days. A median that is fine with a 90th percentile that is very long means the practice is mostly on time but has specific sessions or providers that run badly. That is a different problem from everyone waiting a bit too long.
A two-week wait-time study
Agree the intervals and the clock
Use the five intervals above and a single clock. Tell staff the study is about the process, not individuals.
Collect for two normal weeks
Use EHR timestamps if they are reliable; otherwise use time cards for every visit. Avoid holiday weeks.
Calculate by interval
For each interval, calculate the median and the 90th percentile, overall and by provider, weekday and session.
Find the biggest gap
Pick the interval and session with the longest 90th percentile. That is where to start.
Test one change
Make one change, such as starting room turnover earlier or moving forms to before arrival, and repeat the study for one week.
Common causes and fixes at the front
- Forms completed in the lobby. Send them with the confirmation and ask patients to finish them before arriving; offer help for those who cannot.
- Insurance problems discovered at check-in. Run eligibility two to three days ahead and resolve failures before the day.
- Everyone arriving at once. Stagger appointment start times so the desk is not checking in four patients at 8:00.
- One person at the desk during peaks. Move phone work away from the check-in window at the busiest times.
- Checkout queues. Schedule follow-ups in the exam room where possible, and collect payment at check-in.
Text-when-ready check-in can move some of the lobby wait to the parking lot or a nearby coffee shop. About 91% of US adults own a smartphone[5], so most patients can receive a message, but keep a simple alternative for those who prefer to wait inside.
Tell patients what is happening
At 15 minutes past the appointment time: “Ms. Garcia, I wanted to let you know Dr. Okafor is running about 20 minutes behind this morning. You’re next after the patient she’s with. Would you like to wait, or would you prefer to reschedule? If you step out, we can text you when it’s nearly time.”
An update before the patient has to ask changes how the wait feels. Set a rule, such as any patient waiting 15 minutes past their time gets an update and a choice, and have the desk check the lobby against the schedule every quarter hour during clinic sessions.
Early arrivals, late patients and running behind
Three patterns complicate the numbers, and each deserves its own line in the study.
- Early arrivals. Record how early patients arrive. If many come well ahead of their time, they may expect to wait, or the reminder may tell them to arrive early for paperwork that could be done at home.
- Late patients. A patient who arrives after their appointment time can push every later visit back. Record late arrivals and how the practice handled them, and agree a policy for how late is too late to be seen as booked.
- Running behind from the first visit. If the first visit of a session starts late, every interval after it grows. Track the start time of the first visit in each session separately; it is often the simplest fix.
Separating these patterns keeps the study fair. A long lobby wait caused by a late first start is a different problem from one caused by slow room turnover, and needs a different owner.
Sharing results with the team
Share the study results with everyone involved in the visit: front desk, medical assistants and providers. Show the intervals as a simple bar for each, with the median and 90th percentile, and point to the session or provider where the longest waits cluster. Keep names out of the group discussion and talk about sessions and steps instead. Ask the people who work that session what they think is causing it before proposing a change; they usually know, and a change they suggest is more likely to stick.
Example: what the study showed
Example: a fictional internal medicine practice runs a two-week study. The desk interval has a median of four minutes and a 90th percentile of eleven. The lobby interval has a median of nine minutes but a 90th percentile of 34, concentrated in one provider’s afternoon sessions. The exam room interval is short except on Mondays.
Example, continued: the afternoon problem traces to a single medical assistant supporting two providers after lunch, because of a schedule change months earlier. The practice adds a second assistant for those sessions three days a week and repeats the study. The lobby 90th percentile for those sessions falls substantially. The front desk, which staff had blamed, was not the bottleneck. All figures are illustrative.
Keep measuring lightly
After the first study, put the median and 90th percentile for the lobby interval on the weekly practice scorecard if your system timestamps it reliably, and repeat the full study twice a year. For related measures, see reading your phone system’s call reports and setting front office goals staff can hit.
Frequently asked questions
How do you measure patient wait time in a medical office?
Record timestamps for arrival, check-in complete, roomed, clinician in, clinician out and departure, then calculate each interval from the later of arrival time and appointment time.
Should we report average wait time?
Report the median and the 90th percentile instead. Wait times are skewed, so a few long waits distort the average.
What is the most common cause of long waits?
It varies by practice, which is why the intervals matter. Measure arrival to check-in, check-in to rooming and rooming to clinician separately to find yours.
How should staff handle a provider running late?
Tell waiting patients early, at about 15 minutes past their time, give an honest estimate and offer a choice to wait, step out with a text alert or reschedule.
References
Medical Group Management Association (MGMA) (2025). Patient access priorities for 2026: Tackling wait times, phones, no-shows and more (opens in a new tab) ↑
CDC National Center for Health Statistics (2019). National Ambulatory Medical Care Survey: 2019 National Summary Tables (opens in a new tab) ↑
CDC National Center for Health Statistics (2022). National Hospital Ambulatory Medical Care Survey: 2022 Emergency Department Summary Tables (Table 3) (opens in a new tab) ↑
CDC National Center for Health Statistics (2022). FastStats - Emergency Department Visits (opens in a new tab) ↑
Pew Research Center (2026). Americans' Use of Mobile Devices in 2026 (Mobile Fact Sheet) (opens in a new tab) ↑