Key takeaways
- A portal message can be billed only when it becomes an e-visit: patient-initiated, an established patient, and at least five minutes of clinician time over seven days[2].
- Medicare patients pay 20% of the approved amount after the Part B deductible[3], which is $283 in 2026[4].
- Practices that bill well charge for a small share of messages, often under 1%, and warn patients before they send[1].
- The front desk needs three scripts: before the message, after the statement, and for patients who would rather not pay.
Yes, but only when a message turns into care
The question usually reaches the desk as a phone call from an unhappy patient: “I sent the doctor a message and got a bill for it.” The short answer is that practices can bill for some portal conversations, under rules that were written for exactly that, and most messages will never qualify. Medicare calls these e-visits. They must be started by the patient, the patient must already be established with the practice, and the clinician's time on the problem is added up over seven days and billed in three bands: 5 to 10 minutes, 11 to 20, and 21 or more[2].
The pressure behind the question is real. Patient-authored portal messages per active patient rose 153% between 2020 and 2025 in national EHR data cited by MGMA, and the busiest quarter of primary care physicians receive 53.3 messages a week against a median of 9.6[5]. Inbox management already took 85 minutes of a primary care physician's day in an earlier time-motion study[6]. Billing is one way practices try to recognize that work. It is also the fastest way to start an argument with a patient who thought messages were free.
Source: Medical Group Management Association (MGMA), 2026[5]
Show the data
| Physician group | Messages a week |
|---|---|
| Median physician | 9.6 messages a week |
| Highest-volume quarter | 53.3 messages a week |
What turns a message into a billable e-visit
Four conditions do most of the sorting. The patient started the conversation. The patient is established with the practice. A clinician evaluated and managed a problem, not just passed on information. And that clinician's cumulative time across up to seven days reached at least five minutes[2]. CMS describes the codes as billed by practitioners who can bill evaluation and management visits on their own, such as physicians and nurse practitioners, with a separate code set for clinicians like physical and occupational therapists[2]. Time a medical assistant spends sorting the inbox is not what those codes pay for.
| Patient message | Likely billable as an e-visit? | Why |
|---|---|---|
| “Can I move Tuesday's appointment to Thursday?” | No | Scheduling work, no clinical evaluation |
| “Are my lab results back yet?” | No | A status check; the result discussion happens separately |
| “Please send my usual refill to the new pharmacy.” | Rarely | Usually handled under a standing refill protocol with no new evaluation |
| “My home blood pressure readings are attached. Should my dose change?” | Possibly | A clinician reviews data and decides on treatment; time may pass five minutes |
| “New rash on my arm since Sunday, photo attached.” | Possibly | A new problem evaluated online, unless the clinician asks the patient to come in |
| A two-line question answered in under five minutes | No | Below the five-minute minimum for the lowest code |
The seven-day window matters at the desk. A patient who writes on Monday, answers a follow-up question on Wednesday and gets a plan on Thursday has had one e-visit, not three. If the clinician instead decides the rash needs to be seen, the message becomes the reason for a visit and the conversation is handled as part of it.
What the patient actually pays
Medicare Part B covers e-visits. After the Part B deductible, the patient pays 20% of the Medicare-approved amount[3], and that deductible is $283 in 2026[4]. So a patient in January who has not met the deductible may owe the full approved amount for an e-visit, while the same message in June costs only the coinsurance. Commercial plans set their own cost sharing, and a patient with a high-deductible plan may also owe the full allowed amount early in the year.
Example: a Medicare patient writes about dizziness after a dose change. Over three days the physician spends 14 minutes reading home readings, adjusting the dose and replying. That sits in the 11 to 20 minute band[2]. The patient's share depends on whether the $283 deductible is met[4]: if it is, 20% of the approved amount[3]; if not, potentially all of it. The desk should be able to explain both cases without quoting a dollar figure it cannot see.
Medicare is explicit about who starts it: to get an e-visit, the patient must ask the provider for one[3]. CMS has also said patients must consent to these communication-based services[2]. In practice that means a clear notice before the patient sends a clinical message, and a consent record in the chart.
E-visit, virtual check-in or visit: telling them apart
Patients and staff mix these up, and the statement descriptor rarely helps. An e-visit is the portal conversation described above: written, spread over days, billed on cumulative time[2]. A virtual check-in is brief and happens in real time, usually 10 minutes or less, to decide whether the patient needs a visit at all[7]. The CPT code for it changed in 2025, when code 98016 replaced the old HCPCS code[1]. Medicare asks for the patient's verbal consent to virtual check-ins, documented in the record, and a single consent can cover a year of them[7].
For the desk, the useful distinction is what the patient experienced. Wrote back and forth in the portal over a few days: e-visit. Had a short call or video exchange to decide whether to come in: virtual check-in. Had a scheduled appointment by video or phone: a telehealth visit, billed like the office visit it replaces. If the patient cannot place the charge in one of those three, send the call to billing rather than guessing.
Tell patients before the first bill, not after
The Mayo Clinic experience reported by MGMA is the one most practices cite: in six months of billing eligible messages as e-visits, message thread volume fell 8.8% against the same months a year earlier, and a pop-up disclaimer warning patients of a possible bill may have contributed[1]. MGMA's description of programs that work is consistent: bill only messages that need medical expertise and real time, typically five minutes or more, tell patients clearly which messages may be charged, and expect to bill a small share of messages, often under 1%[1].
Example portal notice, shown before a clinical message is sent: “Messages about scheduling, refills on your current medicines, forms and billing are free. If your message asks your clinician to assess a new or changing problem and takes five minutes or more of their time, it may be billed to your insurance as an online visit, and you may owe a copay, coinsurance or deductible. If you would rather book a visit, call us at (555) 010-4400.”
- Put the notice where the patient writes, not only in the financial policy nobody reads.
- List what is never charged. Patients worry less when the free categories are spelled out.
- Use the same words in the auto-reply, the financial policy and the statement descriptor, so the call to the desk starts from familiar language. The financial policy for new patients is the place to add it.
- Tell the desk the day billing starts. The first statements will produce calls, and a surprised front desk makes it worse.
- Caller
- Your team
- System
01Caller
Patient sees notice, sends
The possible-charge notice appears before a clinical message is submitted.
02Your team
Inbox sorted
Scheduling, refills and forms go to staff pools; clinical questions to the clinician.
03Your team
Clinician evaluates
Replies over up to seven days and documents the cumulative time.
04System
Code chosen by time
5 to 10, 11 to 20 or 21-plus minutes; under five minutes is not billed.
05Your team
Statement and questions
The descriptor matches the portal notice; the desk explains with the same words.
Example workflow. Time bands are from CMS's e-visit codes[2]; each payer's policy is checked before billing.
Scripts for the desk
Most charge calls fall into three types. The first is a patient asking before they write. The second is the statement call. The third is the patient who would rather not pay and wants to know the alternatives.
Before the message: “Messages about appointments, refills of your regular medicines, forms and billing don't cost anything. If you're asking the doctor to look at a new problem and it takes them five minutes or more, it may go to your insurance as an online visit, the same way a short office visit would. If you'd like, I can book you a visit instead.”
After the statement: “That charge is for an online visit. Dr. Patel reviewed your readings and changed your medicine through the portal over a few days, and that was billed to your insurance. What you owe depends on your plan and your deductible. I'll connect you with our billing team, who can see exactly how it was processed.”
The third call needs a straight answer, not a sales pitch for the portal. Administrative messages stay free; a clinical question can become an appointment, a phone visit, or a call back from the nurse line under whatever triage policy the practice runs. Phone visits have billing rules of their own, covered in billing for phone calls. For bill disputes, hand off to billing using the wording in the billing questions phone script, and for cost questions before care, answering “how much will it cost?” on the phone has the tone to use.
Deciding whether to bill at all
Billing is a choice, not a requirement. MGMA lists the trade-offs plainly: fewer messages may mean missed chances for early intervention, patients may start calling instead and add to phone volume, charges may weigh hardest on lower-income patients, and practices in value-based agreements often treat messaging as part of comprehensive care[1]. A practice that starts billing and does nothing about the phones may simply move the work from the inbox to the front desk.
Measure the inbox first
Pull four weeks of messages by type. If most are scheduling, refills and forms, better routing helps more than billing; managing portal message volume shows how to sort them.
Check who already does the work
In an October 2026 MGMA poll, 71% of groups had someone designated to handle other providers' inbox messages, and most of those who commented on time or pay said the work came with neither[5].
Plan for the phones
Decide what happens when patients call instead of writing. Portal directions and the practice's message policy are office questions; a patient's own bill goes to billing.
Review after 90 days
Compare message volume, phone volume and charge complaints against the quarter before. Adjust the notice wording before you adjust the policy.
Phone coverage is where an SI Employee fits. It can answer the office side of these calls, such as how to reach the portal and how the practice's approved message policy reads, and it passes anything about a patient's own statement to the billing line you choose without taking details (see what it does). In the family practice morning rush example, routine booking and results callers are pointed to the portal the practice already runs, which only works if patients trust that using it will not surprise them. Portal use is common enough to make that matter: 65% of individuals were offered and accessed their online records or portal in 2024[8], and getting patients to use the portal depends on keeping that trust.
Frequently asked questions
Can a practice charge for every portal message?
No. Only messages that become e-visits can be billed: patient-initiated, from an established patient, with at least five minutes of clinician time over seven days[2]. Programs MGMA describes as working well bill a small share, often under 1% of messages[1].
Does Medicare cover e-visits?
Yes. Part B covers them, and the patient pays 20% of the approved amount after the Part B deductible[3], which is $283 in 2026[4].
Does the patient have to agree first?
Medicare says the patient must ask the provider for an e-visit[3], and CMS has said patients must consent to these services[2]. A notice before a clinical message is sent, plus a consent note in the chart, covers both.
Can our nurses or medical assistants bill for answering messages?
CMS describes the e-visit codes as billed by practitioners who can bill evaluation and management services independently, with a separate set for clinicians such as therapists[2]. Time spent by staff who sort and route messages is not what those codes pay for.
Will billing reduce the number of messages?
Probably somewhat. MGMA reports an 8.8% drop in message threads over six months of billing at Mayo Clinic, and notes some patients may call instead[1].
References
Medical Group Management Association (MGMA) (2025). Ensuring accurate coding and billing for patient portal messages as e-visits (opens in a new tab) ↑
Centers for Medicare & Medicaid Services (2020). Medicare Telemedicine Health Care Provider Fact Sheet (opens in a new tab) ↑
Medicare.gov (Centers for Medicare & Medicaid Services) (2026). E-visits (opens in a new tab) ↑
Centers for Medicare & Medicaid Services (2026). 2026 Medicare Parts A & B Premiums and Deductibles (opens in a new tab) ↑
Medical Group Management Association (MGMA) (2026). Who owns the inbox? How medical groups can redesign portal message work (opens in a new tab) ↑
Annals of Family Medicine (2017). Tethered to the EHR: Primary Care Physician Workload Assessment Using EHR Event Log Data and Time-Motion Observations (opens in a new tab) ↑
Medicare.gov (Centers for Medicare & Medicaid Services) (2026). Virtual check-ins (opens in a new tab) ↑
Office of the National Coordinator for Health IT (ASTP/ONC), HHS (2024). Individuals' Access and Use of Patient Portals and Smartphone Health Apps, 2024 (Data Brief No. 77) (opens in a new tab) ↑