Write Appointment Reminders as if the Care Is Reserved


Two vaccination megastudies found that reminders work best when they tell patients the care is already set aside for them, and most health center reminders can be rewritten that way in a week.
A patient gets a text two days before a primary care visit. It reminds her that flu season is here and asks her to consider a flu shot. She reads it, means to ask, and forgets by the time she is in the exam room. The reminder did its job on paper, and the shot never happened. Most appointment reminders work the same way.
Health center messages like this one inform, and they leave the next decision to the patient. Two of the largest messaging studies in health care suggest a better default: tell the patient the care is already set aside for them. Here is how to check yours.
Why the wording of appointment reminders matters
In a megastudy published in PNAS in 2021, researchers sent 19 different text messages to 47,306 patients before their upcoming primary care visits and measured flu vaccination rates. On average, the messages raised vaccination rates by 5 percent.
The authors report that messages performed better when they were framed as reminders about a flu shot that was already reserved for the patient, and when they sounded like the communication patients expected from their provider. The best-performing message reminded patients twice and said the shot was reserved for them.
A second megastudy, published in PNAS in 2022, sent 22 different text reminders to 689,693 pharmacy patients. Across the messages, vaccination rose by an average of 2.0 percentage points, or 6.8 percent, over three months. The top message was sent twice, three days apart, and told patients a vaccine was waiting for them.
The reminders that worked best told patients the care was already theirs.
What this suggests for health centers is that a reserved framing turns the visit into something the patient already holds. Declining it becomes the active choice. Both studies tested vaccination, so applying the findings to other services is an inference; the steps below provide a way to test it.
The steps
Each step lists the action, the owner, how to know it passed, how to check, and how long it takes. They are written for a health center without a marketing department.
Step 1: Collect every reminder you send
Action: Pull the exact text of every automated reminder: appointments, screenings, vaccinations, refills and recalls, in every language you send.
Owner: Operations manager, with the person who manages the scheduling or texting platform.
Pass: One document holds every live message, with its trigger, timing and language.
How to check: Compare the list against the platform's active templates. Every template should appear once.
Time: Half a day.
Step 2: Rewrite the core line around the reservation
Action: Change the main sentence from a suggestion to a reservation. "Remember to ask about your flu shot" becomes "We have a flu shot reserved for you at your visit on Thursday." "It's time to schedule your mammogram" becomes "Your mammogram is set aside for you. Reply 1 to confirm a time."
Owner: Operations manager, with the clinical lead for each service.
Pass: Every rewritten message names the service, states that it is held for the patient, and provides one clear action.
How to check: Read each message aloud. If the patient still has to decide whether to start the process, rewrite it.
Time: One day for drafting, one review meeting.
Step 3: Send it twice
Action: Schedule a second message. In both studies, the best-performing message was a reminder sent twice.
Owner: Scheduling platform administrator.
Pass: Each priority service sends two messages, spaced a few days apart, with the second sent before the visit.
How to check: Send a test appointment to a staff phone and confirm both messages arrive on time.
Time: Two hours.
Step 4: Keep the voice the patient expects
Action: Remove jokes, emojis, quizzes and anything that asks for a chat. The 2021 study found messages did better when they matched what patients expected from their provider, rather than being surprising, casual or interactive.
Owner: Operations manager.
Pass: Every message names the health center, reads like a note from the clinic, and contains no more than one question.
How to check: Show the messages to two front desk staff and ask whether each one sounds like the health center.
Time: One hour.
Step 5: Make the reservation true
Action: Confirm that the front desk and clinical staff will honor every message. If a text says a shot is reserved, the vaccine must be in stock and the medical assistant must know to offer it.
Owner: Clinical operations lead.
Pass: Staff working the day's schedule can see which patients received a reserved message for which service.
How to check: On three random days, ask the medical assistant which patients on the schedule have a reserved service. They should be able to answer from the schedule.
Time: One week to set up a flag in the schedule.
Step 6: Measure it
Action: Compare completion rates for each service for eight weeks before and after the change.
Owner: Quality improvement lead.
Pass: A one-page report shows the completion rate for each rewritten service, before and after.
How to check: Use the same measure your quality reports already use, so that the numbers can be compared.
Time: One hour a month.
The most common failure
The most common failure is a reserved message that is not reserved. A patient who arrives to hear that the shot is out of stock, or that nobody knows about the screening, learns that the health center's messages cannot be trusted. That costs more than the original reminder ever gained. Step 5 exists for this reason and should be completed before the new messages go live.
A second failure is to write the reservation in English and translate it word for word. "Reserved" can read as formal or even as a hold on payment in another language. Have a native speaker on staff write each language version from the meaning, then run the same read-aloud test. The goal is the same promise in every language, in words a patient would use.
A reminder that promises care the clinic is not ready to deliver costs more trust than it gains.
References
Milkman, K. L., et al. (2021). A megastudy of text-based nudges encouraging patients to get vaccinated at an upcoming doctor's appointment. Proceedings of the National Academy of Sciences, 118(20), e2101165118. Field experiment, N = 47,306. https://pubmed.ncbi.nlm.nih.gov/33926993/
Milkman, K. L., et al. (2022). A 680,000-person megastudy of nudges to encourage vaccination in pharmacies. Proceedings of the National Academy of Sciences, 119(6), e2115126119. https://knowledge.uchicago.edu/record/9653/files/milkman-et-al-2022-a-680-000-person-megastudy-of-nudges-to-encourage-vaccination-in-pharmacies.pdf
Behavior Change for Good Initiative, University of Pennsylvania. Vaccination megastudy summaries. https://bcfg.wharton.upenn.edu/vaccination/
Some ideas are worth discussing in the context of your organization.


