What does a medical answering service cost?
Nobody quotes a medical answering service cost as a single number, and that is not evasiveness. The service is sold by the minute, by the call or by the month, and the same practice will pay three different amounts depending on which of those units its own calls happen to suit.
This page describes the units and what moves them, for a practice of one or two treatment rooms rather than a hospital switchboard. It does not reproduce anybody's published rate card: those belong to the companies that publish them, and a figure copied out of one is a sales document rather than evidence. What follows comes instead from the public data that does exist — what the work underneath the service pays, how long healthcare telephone calls actually run, and what the rules oblige of anyone handling patient information on your behalf.
What does a medical answering service cost?
It costs your own call pattern multiplied by whichever unit the contract counts, which is why two practices on identical terms receive different bills. The rate is the smaller half of the question; the unit, and what it is allowed to count, is the larger half.
An answering service for medical offices is sold by the minute, by the call or by the month, but underneath every one of those units is the same thing: a person wearing a headset for an hour. The occupation has its own line in the United States labour statistics — switchboard operators, including answering service, defined as workers who “operate telephone business systems equipment or switchboards to relay incoming, outgoing, and interoffice calls” and who “may supply information to callers and record messages”. As of May 2023 there were 43,830 of them, at a mean hourly wage of $19.24 and a median of $17.67.
That figure is not a price and should not be read as one — it is wage, not invoice, and it carries none of the supervision, systems, premises or margin that sit on top of it. It is useful for one thing only, which is knowing what the floor is made of. A quote far below the cost of the hour it covers is not a bargain; it is a different product, usually one where the hour is shared across many more clients than yours.
What is actually being priced?
Four things, usually bundled and rarely itemised: connected time, the call itself, a window of coverage, and the handling rules that apply inside it. A quote that names only one of them is not cheaper than a quote that names all four — it is less specific.
Two medical answering service pricing models can produce the same monthly bill in one practice and a third more in another, without either supplier changing a rate. What changes is the shape of the calls going into them, and that shape belongs to the practice. The table below separates what is charged from the unit it is charged in, because the second column is where the money actually moves.
What answering services charge for, and what the unit is
| What is charged | The unit it is charged in | What makes it move |
|---|---|---|
| Connected time | The minute, plus whatever rounding rule is applied to the last part of one | How long a healthcare telephone call runs, which varies far more than it averages. Across 128,717 out-of-hours consultations the mean call length was 7.78 minutes with a standard deviation of 4.77 — a spread more than half as large as the average itself |
| The answered call | The call, whatever its length turns out to be | What the calls are about, not who is making them. In the same dataset calls designated as mental health calls ran a mean of 11.16 minutes against 7.15 for calls taken by a GP, while the study found no equivalent effect of patient age or sex |
| Being covered at all | The month, for a named window of hours and a named number of lines | Capacity rather than usage: the window has to be staffed whether or not it is used. A survey of telephone access in general practice found an average of 3,659 patients per incoming line, with 848 of the responding doctors estimating four or fewer patient calls a day — the same line, very different loads |
| Set-up | A one-off charge for scripts, escalation rules and training the people who will answer | How much of your practice has to be written down before the first call is taken, and whether patient information is involved — in which case a written agreement has to exist before any of it is disclosed, not after the service goes live |
| After-hours cover | The out-of-hours window, priced apart from the working day | What arrives in that window and what it has to do. Of 970 out-of-hours calls in general practice, 86% were managed directly by the practice rather than by a deputising service, and the share resolved by telephone advice alone ranged from 5% to 57% |
| Compliance handling | The written agreement, and the safeguards it obliges behind it | Whether the service creates, receives, maintains or transmits protected health information on your behalf. If it does, it is a business associate as the regulation defines one, and the obligations that follow are contractual rather than optional |
What does an after-hours answering service for a medical practice add?
A second staffing window, and a different job inside it. The daytime call mostly wants a booking; the call that arrives once you have closed more often wants an answer, and the two are not the same service even when the same voice takes them.
The distinction shows up in the only out-of-hours data that is public. In the study of 970 calls above, the proportion of contacts closed by telephone advice alone varied between practices from 5% to 57%, with a mean of 37%; the authors also recorded that use of deputies rose at night while the doctors who stayed on call kept their advice rates. A window in which more than a third of contacts end in a conversation rather than an appointment is being staffed for judgement, and judgement is the expensive kind of minute.
A private manual therapy practice is not an out-of-hours co-operative, and the parallel should not be stretched: nobody is triaging an emergency at eleven at night over a sports injury. But the pricing lesson carries over intact. If your evening calls are mostly people asking to be booked, you are paying for a message to be taken. If they are people asking whether they should come in at all, you are paying for something the person answering may not be qualified to give, and the cheapest version of the service is the one that will give it anyway.
Ambra Anfosso, an osteopath who had opened her own practice only months before her interview, describes what happens when the volume arrives before the arrangement does: she began by giving out her personal mobile number and had changed it within weeks, and the paper diary she was attached to lasted about a week after that. Her published comment is about the demand rather than the phone — “You are amazing, and you support the operator in every way! My first availability after 15 days is in more than two weeks.” — but the sequence behind it is the ordinary one. The cover gets bought after the week it was needed.
What does a HIPAA-compliant medical answering service actually have to do?
It has to be bound by a written agreement before it hears a single patient name, and that agreement has a specified content rather than a reassuring tone. The phrase on a supplier's website is a claim about a document; the document is the thing being bought.
The regulation is unambiguous about who is caught. A person who, on behalf of a covered entity, “creates, receives, maintains, or transmits protected health information for a function or activity regulated by this subchapter” is a business associate, and the definition extends to anyone providing “management, administrative, accreditation, or financial services” involving that information. Taking a patient's name, number and reason for calling is squarely inside it. The same rule pulls in a subcontractor who handles the information on the answering service's behalf, which matters when calls overflow to a second site at two in the morning.
What the contract must require is itemised and worth reading once in full, because it is the checklist a quote can be held against: appropriate safeguards, no use or disclosure beyond what the contract or the law permits, reporting of unauthorised use or disclosure including breaches, the same restrictions passed down to subcontractors, information made available for access and amendment and for an accounting of disclosures, internal practices and records made available to the Secretary, and return or destruction of the information when the contract ends. The last item is the one most often discovered late: leaving a supplier is a data question, not only a notice period.
None of this applies to an Italian or European practice in the form written above — the framework there is a different one, with its own processor agreement. The structure of the cost, though, is the same wherever you are: the handling obligations are a separate line from the answering, and a supplier who has not priced them has not read them.
What does a clinic pay for twice without noticing?
The same call, whenever the message it produces has to be re-entered by somebody at the practice. And the same hour, whenever the person answering has no way to write into the diary the practice actually runs on.
This is the quiet half of the bill and it never appears on the quote. A call is billed at its full length, a message arrives, and then a second person — usually the therapist, usually between patients — reads it, telephones back, finds the slot and writes it down. The minute was bought once and paid for twice, and the second payment comes out of the treatment day rather than the invoice. Whether it happens at all depends on one thing: whether whoever answers can write directly into the record the practice keeps, which is a question about a crm for medical clinics rather than about the phone line.
Put the other way round, the difference between an answering service and a virtual receptionist for clinics is not the voice at the other end but what it is permitted to finish. One hands you work; the other closes it. That distinction is worth more than a difference of a few cents a minute, and it is the one comparison a rate card cannot help you make.
It is also where the honest caveat belongs. Veronica Zamboni, an osteopath working across three locations, is direct about the automatic handling of her schedule — “Of all the methods and portals, this is the only one that brings new patients in no time. Such results are not normal; automatic management of the schedule thanks to Artificial Intelligence has allowed me to focus solely on the patients in the office.” — and, in the Italian interview from which that is drawn, she is equally direct that the automatic replies were not always precise and that she read the conversations afterwards. Reported rather than quoted, because those words were said in another language: an automated answer still needs somebody to look at it, and any budget that assumes otherwise is short by the time it takes to look.
What should be counted before comparing two quotes?
Four numbers out of your own diary, none of which a supplier can produce for you. Without them a quote is a rate; with them it is a forecast, and two forecasts can be compared even when the terms behind them are not alike.
- How many calls arrive on an ordinary working day, counted for a fortnight rather than remembered.
- What share of them land outside your opening hours, which is the number that decides whether a separate out-of-hours line is worth its own fee.
- How long the answered ones actually run — and how wide the spread is, because on per-minute terms the spread is yours and not the supplier's.
- What share end in a booking rather than in a message somebody has to action later, since only the first kind has been finished.
With those four in hand the arithmetic is short. Multiply the daily volume by the mean length for a per-minute quote, by the call count for a per-call quote, and compare both against the flat monthly figure — then add the fourth number back as the work the practice still has to do itself. The quote that wins on the first calculation often loses on the last one, and the last one is the one paid in treatment hours.
An honest note on the evidence used here. The call-length and out-of-hours figures come from general practice and from clinician telephone consultations, not from commercial answering services taking messages for private clinics: the conversations measured are longer and more clinical than the ones a message-taking service handles. They are cited for what they establish, which is that the length of a healthcare telephone call is driven by what the call is about and varies widely around its own mean — not as a forecast of your own minutes. The wage data is United States national data for one occupation, and no part of it is a price.
Where the answering is one part of a larger question about who reaches the practice and what happens next, it stops being a phone-line decision and becomes the same conversation a healthcare marketing agency would have about acquisition: an enquiry that is never answered was paid for twice over, once to arrive and once to be lost.
If the part you would rather change is that the phone rings while you are with a patient and nothing happens after it stops, you can Apply as a customer.
