Medical Front Desk and Patient Phone Statistics (2026)

No US dataset publishes a national missed-call rate for medical practices. The closest national figure is patient-reported: in AHRQ's 2019 CAHPS Clinician and Group chartbook, 58% of patients said they always got a same-day answer when they contacted the office during business hours, and 15% said never or sometimes.

Quick answer:

The honest number: nobody measures missed calls nationally. The nearest thing is patient-reported same-day answer rates from AHRQ, and the last national collection covers 2019.
What practice leaders say now: phone access was the third-ranked patient access priority for 2026 at 22% of medical groups polled by MGMA in December 2025, behind no-shows at 27% and online scheduling at 24%, and calls were the second most common target for front-office AI in a February 2026 poll.
What individual practices report: one health system published that some clinics were sending more than half of incoming calls to voicemail, and that fixing it tracked with higher schedule utilization.

Every number below is named, dated and linked. Where the public data is thin, we say so.

Why there is no national missed-call rate

Nobody collects it. Practices own their phone data, and no federal program asks a practice how many calls it dropped.

The one national instrument that measured phone access at practice level has stopped. AHRQ states plainly on its own page that the CAHPS Clinician & Group Survey Database "was operational from 2010 through 2020" and that "In 2021, AHRQ suspended new data submissions for the Clinician & Group Survey Database due to a decline in participation." Its chartbooks run 2012 to 2019 and nothing has replaced them.

That vacuum is why unsourced figures circulate. Weave's own AI receptionist page states "1 in 3 calls are missed" and "47% of bookings occur after hours" with no study, sample or date attached. Those may hold for Weave's customers, but they are not published research.

The closest thing to a national number

The CAHPS Clinician & Group Survey Database 2019 Chartbook reports responses from 330,390 patients at 2,002 medical practice sites, collected January 2019 through June 2020 about encounters in 2019.

The question that matters is asked word for word: "In the last 6 months, when you contacted this provider's office during regular office hours, how often did you get an answer to your medical question that same day?" Across 130,425 respondents:

Always: 58%
Usually: 27%
Never or sometimes: 15%

So 42% of patients did not always get a same-day answer, and about one in seven said it rarely or never happened. That is a patient-experience measure, not a phone log: it covers the whole loop, including the callback that never came.

Getting timely appointments, care and information scored 65%, the lowest of the five measures. Helpful, courteous and respectful office staff scored 79%. Patients rate the people at the front desk far higher than their ability to reach them.

AHRQ is explicit: organizations submit voluntarily and "are not representative of all U.S. medical practices." Treat it as the best available national reading, not a census.

Phone access gets worse as the group gets bigger

This finding undercuts the framing of phones as a small-practice problem. In the same chartbook, top-box scores on timely appointments, care and information fall almost monotonically with practice size:

1 provider: 76%
2 to 3 providers: 69%
4 to 9 providers: 65%
10 to 13 providers: 62%
14 to 19 providers: 61%
More than 20 providers: 62%

AHRQ's own highlight for 2019 reads: "Practices with only one provider scored higher on all measures than practices with more than one provider."

A solo clinician answering her own phone outperforms a 20-provider group with a switchboard. What a larger front office gains in headcount it loses in routing, transfers, queues and handoffs. Ona is a general ambulatory EHR whose own homepage FAQ says it "scales from solo clinics to enterprise systems".

What practice leaders reported in 2025 and 2026

MGMA Stat polls are small and self-selected, but MGMA publishes the date and applicable responses for each, which makes them the most usable recent operator-side data in the US.

December 9, 2025, 236 responses. Top patient access focus for 2026: no-shows 27%, online scheduling 24%, phone access 22%, wait times 21%, other 5%.
March 10, 2026, 294 responses. Most time-intensive phone work for staff: eligibility and prior authorization 45%, scheduling 31%, intake 9%, prescription refills 6%, other 9%.
February 10, 2026, 177 responses. Where practices are pointing front-office AI: scheduling 31%, calls 27%, registration and eligibility 23%, prior authorization 16%, other 5%.
April 8, 2025, 375 responses. Asked "Do you use some version of chatbot or virtual assistant for patient communication?", 19% said yes and 81% said no.
July 29, 2025, 244 responses. Share of patients self-scheduling digitally: 71% of groups said less than 25%, 21% said 25% to 50%, 5% said 51% to 75%, 3% said more than 75%.
May 26, 2026, 303 responses. Staff turnover against last year: about the same 39%, lower 30%, higher 28%, unsure 2%, with turnover "concentrated in the jobs that keep clinic flow, access and cash moving - MAs, front-desk and patient access staff."

The picture is consistent. Seven in ten groups (71%) still have fewer than one patient in four self-scheduling, so the phone carries the appointment. The work on that phone is mostly eligibility, prior authorization and scheduling: slow, script-heavy, and exactly what turns over fastest at the front desk.

What happens to the call nobody picks up

Direct evidence comes from studies that called practices and recorded what happened.

A mystery shopper study published in SSM Qualitative Research in Health on January 28, 2025 by Melgoza and colleagues placed 129 calls to primary care offices listed in Medicaid managed care plan directories in the Los Angeles, Houston and New York metro areas to schedule a first-time appointment, with a stated preference for telehealth. Of 66 calls in English, 31 (46.9%) were completed. Of 63 calls in Spanish, 18 (28.5%) were completed. It is a Medicaid directory sample, not a cross-section of US practices, so read it as a floor on a hard population rather than a general answer rate.

The Senate Committee on Finance published a secret shopper study on May 3, 2023 in which staff called 120 provider listings drawn from 12 different health plans. As the American Medical Association summarized it on May 18, 2023, "33% were inaccurate, had nonworking numbers, or had unreturned calls," and appointments were available only 18% of the time. It measured directory accuracy in mental health networks, not practice phone performance, so it is not a general answer rate. It does show how often a patient holding a correct-looking number still reaches nobody.

The one operator account with before-and-after numbers is Josh Anderson's write-up for MGMA on September 26, 2023, describing phone system work at Valley View Hospital in Colorado. It defines the metric usefully: "Missed calls were calculated as the number of calls that went to voicemail divided by the total number of calls." Before the changes, some clinics had "more than 50% of their incoming calls going to voicemail." The target was "a missed call rate of 10% or lower," and "it took almost a year to do this." Cutting voicemail cycles freed about 48 hours a month. The organization reported a 17% increase in work RVUs and a strong negative correlation between missed calls and schedule utilization.

One site, published by the people who did the work, so read it as a case and not a benchmark, and still the most specific public account of what a missed call costs.

What it costs to answer the calls you are missing

One axis only: how much of an unanswered call ends up as a booked appointment inside the patient record, without adding another vendor. It is not a quality ranking. Every price comes from the vendor's own pricing page.

Coverage order	Option	Who can use it today	What happens to the call	Published price, checked September 17, 2026
1	Ona	Any ambulatory practice, solo clinician to multi-location group to enterprise system; not inpatient or hospital workflows	Answered 24/7, booked against the live calendar, with transcript and structured intake written to the chart, and urgent calls warm-transferred to a human	AI receptionist $100 per month for 400 minutes, $150 for 600, $200 for 800, $400 for 1,500, then $0.20 per minute; seats $130 per practitioner, $45 per staff member
2	Smith.ai	Any business; HIPAA is not mentioned on its pricing page	Answered 24/7 by live agents or AI and passed to your team; nothing reaches a clinical record	Starter $300 per month for 30 calls, Basic $810 for 90, Pro $2,100 for 300, with overage of $11.50, $10.50 and $8.50 per call
3	Weave AI Receptionist	Weave customers across dental, optometry, medical and other practice types, with veterinary on a waitlist; automatic booking only with qualified dental practice management systems	Answered and booked where a supported dental system exists, otherwise captured as a request for staff to work	No price published for the AI Receptionist add-on; Weave platform plans start from $199 per month, plus a 25% early-access discount for the rest of 2026
4	CloudTalk	Any business; its pricing page carries general compliance statements including HIPAA	Queued, routed, or answered by an AI voice agent add-on; no clinical record	From EUR 19 per user per month billed annually; AI Receptionist add-on from EUR 99 per month for 200 minutes
5	SimplePractice	Any practice on the platform	Nothing on the phone. Inbound call answering does not appear in its published plans or add-ons; Care Aide is clinical AI for session support, note drafting and client summaries	Starter $49, Essential $79, Plus $99 per month; add-ons Care Aide $59, Note Taker $35, ePrescribe $49 plus a one-time $89 setup fee

Do the per-unit arithmetic before any demo. Ona's first three tiers are $0.25 per included minute. CloudTalk's entry AI add-on is roughly EUR 0.50 per included minute. Smith.ai's Starter plan is $10 a call. Weave publishes no unit price.

Be honest about Ona's own floor. The receptionist is an add-on on top of a seat, so a solo clinician starts at $230 a month: one $130 practitioner seat plus the $100 tier. Cheapest headline price is not the claim. The claim is that the answered call lands in the same record as the visit, the note and the bill, which is what the AI receptionist does.

Two smaller facts change the arithmetic. Omnichannel puts calls, texts, email and website chat in one inbox, where missed and AI-answered calls "mark the thread unread and raise a notification for the answering group, so the callback happens." And online scheduling removes calls rather than answering them, which matters when 71% of groups have fewer than a quarter of patients self-scheduling.

Measure your own rate before you buy anything

Take one ordinary week and pull five numbers out of your phone system.

Total inbound calls.
Calls that ended in voicemail or an abandoned queue. Divide by total calls, using the Valley View definition, for your missed-call rate.
The same split for business hours and after hours. Different problems, different fixes.
Median time to return a voicemail.
Share of answered calls that ended in a booked appointment. Without it, no vendor's booking claim can be tested.

Compare a vendor's numbers with yours, not a statistic from a landing page. Our buyer guide lists what to test in a trial, and the comparison of AI receptionists for medical clinics covers who publishes a price and who does not.

Frequently asked questions

How many patient calls go unanswered?

No US dataset answers that question directly. The closest national measure is patient-reported: in the AHRQ CAHPS Clinician and Group 2019 Chartbook, 58% of patients said they always got a same-day answer when they contacted the office during regular office hours, 27% said usually, and 15% said never or sometimes. Individual published cases run far worse. Valley View Hospital reported in MGMA on September 26, 2023 that some of its clinics were sending more than half of incoming calls to voicemail before it changed its phone system.

What is a good missed-call rate for a medical practice?

There is no published national benchmark. The only public target we can point to by name is Valley View Hospital's, described in MGMA on September 26, 2023: an organizational goal of a missed call rate of 10% or lower, where missed calls were counted as calls that went to voicemail divided by total calls. The same account says it took almost a year to get there.

Is there a national statistic on missed calls at doctors' offices?

No. No federal collection reports a missed-call rate for ambulatory practices. The national patient-experience survey that measured phone access, the CAHPS Clinician and Group Survey Database, was operational from 2010 through 2020, and AHRQ suspended new data submissions in 2021 due to a decline in participation. Its 2019 chartbook is the most recent national picture available, and the widely repeated figures circulating online are vendor marketing without published methodology.

Do bigger practices answer the phone better than solo practices?

The AHRQ data says the opposite. In the 2019 chartbook, top-box scores on getting timely appointments, care and information fell as practices got larger: 76% at practices with one provider, 69% at two to three, 65% at four to nine, 62% at 10 to 13, 61% at 14 to 19, and 62% above 20. Phone access is not a small-practice problem, and the multi-site groups score lowest.

How much does it cost to have AI answer a practice's phone?

Ona publishes minute tiers: $100 per month for 400 minutes, $150 for 600, $200 for 800, $400 for 1,500, then $0.20 per minute beyond the tier. That is $0.25 per included minute at the first three tiers, on top of seats at $130 per practitioner and $45 per staff member, so a solo clinician's entry point is $230 per month. Smith.ai publishes $300 per month for 30 calls, which is $10 a call. Weave publishes a platform starting price of $199 per month but no price for the AI Receptionist add-on itself, and most healthcare AI receptionist vendors publish nothing you can divide by minutes or calls.

How do I measure my practice's missed-call rate?

Take one ordinary week. Count total inbound calls, then count how many ended in voicemail or an abandoned queue, and divide. Split the result into business hours and after hours, because the two need different fixes. Then count how many answered calls produced a booked appointment. Most phone systems export all four numbers, and you cannot judge any vendor's claim until you know your own.

Next step

Measure your own week first. If the number is worse than you expected, the fastest test is hearing a call answered, booked and written to the chart in one pass. Book a 15-minute demo with your real call volume in hand, or start the 14-day free trial, full access, no credit card.