Back to all writing

The patient phone experience: how to measure it, and how to improve it

Most practices track one phone number, abandoned calls. The evidence says patients feel something else more. Which metrics matter, how to measure them in a week, and what to change first.

, November 20, 2026, 8 min read

The patient phone experience is what it is like to call your practice: how long it rings, what the caller hears first, how many times they repeat their date of birth, and whether they hang up with the thing they called for. Most practices measure one number, abandoned calls. The best evidence we have suggests that is not the number patients feel most.

This post is about measurement first and fixes second, because most of the phone problems I see are invisible from inside the practice. Nobody on staff ever calls the main line at 8:05 on a Monday.

What is the patient phone experience?

The patient phone experience is everything that happens to a patient between dialing a practice and getting what they called for: time to answer, the greeting and menu, holds and transfers, whether the request is resolved on that call, and what happens outside office hours. It is measured from the call log and from listening to real calls, not only from satisfaction surveys.

It matters beyond courtesy. For many patients the phone is access. If they cannot get through, they do not have a practice, whatever your clinical quality is. And it is one of the few parts of access a practice controls entirely.

It is also widely felt to be poor. England's national GP Patient Survey, with 702,837 responses in 2025, found that 52.9 percent of patients said it was easy to contact their practice by phone. That is a different health system, but the question it asks is exactly the right one, and almost half of patients did not say yes.

Which numbers measure the patient phone experience?

Seven, and you probably have the data for four of them already in your phone system.

MetricWhat it tells youWhere to get itA reference point
Average speed of answerHow long callers wait for a person or a useful answerPhone system reports, by hourThe VA standard is 80 percent answered within 30 seconds
Abandonment rateShare of callers who hang up before reaching anyonePhone system reportsThe VA standard is 5 percent or less
First call resolutionWhether the caller got what they called for on that callCall review, or repeat calls from the same numberSet your own baseline
Transfers per callHow often callers are passed aroundPhone system reportsSet your own baseline
Time to a person in the menuSeconds of recording and options before a human or the right queueCall your own lineShorter is better; time it
After hours outcomeWhat happens to a call at 7pmCall your own line, review the overnight logAnything better than voicemail
Voicemail callback timeHow long a message waitsStaff log or EHR task timestampsSame day during business hours

The VA standards are useful reference points because they are public and specific, not because a small practice must hit them. They come from the Veterans Health Administration's call center guidance, as described by its Office of Inspector General.

Does speed of answer matter more than abandoned calls?

The best study I know on this suggests it does. Griffith and colleagues, in the American Journal of Managed Care in 2019, linked survey responses from 252,145 VA primary care patients at 285 facilities to each facility's monthly call center data. Facilities in the slowest quartile for average speed of answer had lower odds that patients felt they could get urgent appointments as soon as they needed them, with an odds ratio of 0.85 against the fastest quartile.

Abandonment rate showed no clear relationship with any of the patient outcomes they measured. The authors suggest one reason: abandonment counts include very short calls, which are often misdials, and those dilute the measure.

Two practical lessons. First, track speed of answer by hour, not just the daily average, because the wait at 8am is what patients remember. Second, if you report abandonment, exclude the very short calls so the number means something. The study has limits, including an observational design and a mostly older male patient population, but it points the same way as everything I hear on real calls: waiting is what patients feel.

How do you measure your patient phone experience in one week?

  1. Export a month of call data and chart speed of answer and abandonment by hour of the day. Look for the first hour after opening and the lunch hour
  2. Call your own main line ten times across the week, including at 8:05am, 12:15pm, 4:55pm and after hours. Time the recording, count the menu options and note how long it takes to reach a person
  3. Listen to twenty recorded calls. Count how often the caller repeats information, how often they are transferred, and whether the call ended with the request done
  4. Count repeat calls from the same number within a day. Each one is usually a first call that did not resolve
  5. Ask patients at check in one question for a week: how easy was it to reach us by phone? Very easy, fairly easy, fairly difficult or very difficult

At the end you will have a baseline that is a number rather than a feeling, which is what you need before changing anything or buying anything.

How do you improve the patient phone experience?

Roughly in order of cost.

  • Cut the greeting. Long recorded notices before the menu are the cheapest delay to remove
  • Shrink the menu. Put the most common reasons first, and say the option for new patients and for prescriptions early
  • Staff the peak, not the average. The VA OIG review found waits rose from about three minutes to over ten in the first hour after opening and during breaks. Move one person to phones only for that hour
  • Separate phones from check in during busy periods, so nobody is choosing between the caller and the patient at the window
  • Transfer with context. A warm transfer where the caller does not repeat themselves feels twice as fast
  • Resolve on the first call. Book, rather than take a message and call back
  • Cover after hours with something better than voicemail, and make sure emergency callers hear the instruction to call 911
  • Put routine calls on an AI phone agent, so reschedules and cancellations stop queuing ahead of the callers who need a person

That last item is what MedPhone does. It answers on the first ring, every caller at once and at any hour, books, reschedules and cancels in the EHR, takes refill requests into the provider's inbox for the clinician to approve, and transfers to staff by rules the practice sets. At Synergy Medical, a family medicine practice in Michigan on athenahealth, it handled 2,848 inbound calls between June and mid August 2026. Around 53 percent were finished without staff on day one, improving as call flows were tuned.

What will a better phone experience not fix?

A fast answer to a full schedule is still a wait. If new patients cannot be seen for a month, the phone becomes the place where they find that out quickly. Billing disputes, rude handoffs and a clinic that never calls back are not phone metrics either, though they show up as angry calls.

And an AI agent can make the experience worse if it is set up badly: if it traps callers who need a person, or tries to resolve an emotional call that should go straight to staff. MedPhone gives no clinical advice and does no triage, transfers clinical questions to staff, works in English only and does not make outbound calls. A practice whose patients mostly call in another language, or whose calls are mostly clinical, should fix the phone another way.

See how MedPhone answers a call

Questions people ask

How do you measure the patient phone experience?

Combine phone system data, such as speed of answer, abandonment and transfers by hour, with calling your own line, listening to recorded calls, and asking patients how easy it was to reach you. Data alone misses what callers hear; listening alone misses the patterns.

What is a good call abandonment rate for a medical practice?

The Veterans Health Administration's standard is 5 percent or less, which is a useful public reference point. Exclude very short calls, which are often misdials, and look at abandonment by hour, because a fine daily average can hide a bad first hour.

What is a good average speed of answer for a doctor's office?

The VA standard is 80 percent of calls answered within 30 seconds. Whatever target you set, measure it by hour of day. A 2019 study of VA facilities linked slower speed of answer to patients feeling they could not get urgent appointments when they needed them.

Why do patients find it hard to reach their doctor's office by phone?

Mostly because calls arrive in peaks, at opening and over lunch, while the same staff are also checking patients in. Long recorded greetings, deep phone menus and voicemail after hours add to it.

How can a practice improve phone service without hiring?

Shorten the greeting and menu, move one person to phones only during the first hour, resolve requests on the first call instead of taking messages, and consider an AI phone agent for routine scheduling calls so staff handle the calls that need them.

Does an AI phone agent improve the patient phone experience?

For routine calls it removes the wait, because it answers every caller at once and can finish a reschedule in the EHR. It helps only if callers who need a person can reach one quickly and clinical questions go straight to staff.

Written by

Agni Patel is the founder and CEO of MedPhone, an AI phone agent for medical practices. He spends a lot of his week listening to how practices answer the phone.

Want this run on your practice?

Bring fifteen minutes and your call reports. We will use your call volume and your payer mix rather than the benchmarks in this post.

Book a 15 minute demo