When a patient doesn't answer, the front desk tends to blame the patient. The chart note says "LM x2," the patient later says nobody called, and the slot sits unconfirmed until the morning of the visit. We think most of that delay starts on the practice side of the conversation.
Patient response times depend on three things: the channel, the wording, and who's watching the inbox. A practice controls all three of them.
Phone calls lose to the patient's workday. A voicemail asks for a callback during office hours, which are also the patient's work hours, and the loop repeats. Portal messages wait for a password the patient may not remember. A text lands on a phone most patients keep within reach, so the first barrier is gone before they read a word.
Channel alone doesn't hold the gain. A text that asks for three things gets a partial answer or none. A reply that lands after the desk logs off sits until morning. And a thread nobody owns goes quiet while the patient picks up the phone anyway.
Portal traffic is also getting heavier on the staff side. MGMA's January 28, 2025 poll found 70% of medical groups saw portal message volume rise in 2024, from 223 responses.
The sections below use real message wording, a send-window rule, an escalation timer, and a small worked set. That set shows why the median is the number to track, and how a practice-wide average can make a good week look bad.
Secure text is a patient message sent through a HIPAA-compliant messaging platform that lands in the patient's regular SMS inbox.
Patients answer on the device they carry, and for most adults that's a smartphone. Pew Research Center puts smartphone ownership at 91% of US adults, from 5,022 adults surveyed February to June 2025.
ONC's analysis of the 2024 HINTS survey (7,278 adults) found 65% accessed their online records or portal at least once.
A reply window is the stretch between sending a message and getting an answer, and its shape depends on the channel. Text windows are front-loaded. A 2016 study of text message response times in UK smoking-cessation trials (198 and 293 people) found median replies came in under half an hour.
Phone replies need two free schedules at once. In the MGMA Stat poll on phone bottlenecks in medical practices (294 responses, March 10, 2026), 31% of leaders named scheduling as their biggest phone drain.
Reply behavior by channel
|
Channel |
What has to happen before a reply |
Shape of the reply window |
|---|---|---|
|
Text message |
Patient looks at the phone |
Front-loaded: most answers soon after the send |
|
|
Patient opens the inbox and finds the message |
Tied to when the patient checks email |
|
Patient portal |
Patient remembers a password and logs in |
Waits for the next login |
|
Phone call |
Patient is free when the phone rings |
Needs both schedules to line up; voicemail adds a round |
Three causes of slow replies sit inside the practice's control.
A message asking for several things gets a partial reply or none, since one line can't answer it all. This one asks for three:
[Practice name]: Please confirm your appointment on 10/14, bring your updated insurance form, and call us about your balance.
The patient replies "ok," and staff can't tell which part it answers. Split it: send the confirmation alone three days out, then ask for the insurance card after the patient replies C. The balance goes out later as its own billing message.
Send timing decides whether a fast patient reply gets a fast staff answer. A reminder goes out at 4:55 PM, the patient asks a question at 5:30 PM, and the desk logged off at 5:00. That reply waits until 8:00 the next morning.
Our send-window rule: routine texts go out only when staff will watch the inbox for at least two more hours. For a desk that closes at 5 PM, batch sends stop at 3 PM.
A thread resets when each campaign opens a new conversation, and the patient loses the context needed to answer. A reminder, a form link, and a recall text that arrive as three separate threads look like three unrelated senders.
The patient who replied C last week now sees "Please complete your intake" with no visit attached, and ignores it. Keep every message to one patient in one continuous secure healthcare messaging thread from the practice's main number. A patient scrolling back should see the visit they're being asked about.
Wording is the part of reply speed a practice can change this afternoon.
The pattern is one question, one action, and one way to answer. Lead with the practice name, state the single fact, then give the reply option. Most text appointment reminders fit the first example below.
Each one asks for a single reply, so a one-letter answer is complete.
A reply keyword is a one-letter answer that tells staff exactly what to do next. C should mark the visit confirmed on the schedule. R should open a reschedule task for a named staff member, with the thread attached.
Anything else, such as "can I come at 10 instead?", goes to a person. When comparing HIPAA-compliant texting platforms, ask whether a C reply writes back to the EHR schedule or needs a staff update. Test every keyword end to end before launch.
A reminder can name the practice and the appointment date and time, but should leave out the diagnosis, test, or visit reason. HHS guidance on leaving messages for patients suggests limiting a message to the practice name, number, and what's needed to confirm an appointment. Leave out a department name that reveals the visit type, such as oncology.
That fits the Privacy Rule's minimum necessary standard at 45 CFR 164.502(b). Keep results questions and other clinical texting inside HIPAA-compliant texting, where the thread is protected on the practice side.
Reply speed holds when a named person is watching the inbox during every open hour.
Each open thread should belong to one named person per shift. In a shared inbox where anyone can answer, each person tends to assume someone else did.
New patient replies go to the front desk lead on that shift. Clinical questions move to the nurse for the patient's provider, and billing questions go to the billing queue. The owner's name shows on the thread, so a covering staff member knows who to ask. Owners swap at shift change, and open threads get handed over by name.
An escalation timer reassigns a thread that sits unanswered past a set window. We use 30 minutes during office hours. If the owner hasn't replied by then, the thread moves to the shift lead, and at 60 minutes it goes to the practice manager.
The timer stops at the first staff reply. A thread that's been opened but not answered keeps counting. After hours, an automatic reply tells the patient when the office reopens and to call 911 for emergencies.
Curogram is a HIPAA-compliant, SOC 2 Type II patient communication platform. Task Management for Workflow lets staff forward a patient message as a task assigned to a named user, so each message has an owner.
After-hours automatic responses tell patients their message arrived and when someone will follow up. Texts go out from the practice's main phone number, so patients recognize the sender.
Response time is the gap between a message and its first reply, and healthcare communication efficiency depends on measuring it well.
Averages distort patient response times because one slow thread drags the whole number up. Take five illustrative first replies: 4, 6, 9, and 12 minutes, plus one thread answered two days later, at 2,880 minutes.
The average of that set is about 9.7 hours. The median is 9 minutes, close to what four of the five patients saw. Track the median, then review the slow thread by name. Show staff both numbers in the first month, so nobody mistakes the average for the typical wait.
Three numbers cover reply speed from both sides of the thread. Median first-reply time shows how fast patients answer outbound messages. The share of patient threads answered inside the same shift shows whether staff keep up.
The share of outbound messages that got no reply at all flags wording or timing problems in specific templates. Each month, rewrite the template with the worst no-reply share. To turn those numbers into staff hours and dollars, see the complete guide to HIPAA texting ROI.
Patient response times shrink when a text reaches the phone, asks for one thing, and lands while someone is watching.
Start with the reminder you send most. Cut it to one action, give its replies a named owner and a 30-minute timer, and watch the median for a month.
Book a Curogram demo and bring the reminder you send most. We'll show it running as a two-way thread with a named owner.
Frequently Asked Questions