From her final month of pregnancy through just after childbirth, a woman received a flurry of 17 texts, reminding her about appointments, prompting her to check in, and requesting satisfaction surveys after each visit. After delivery, even more texts streamed in on postpartum guidance, pediatrician appointment reminders, lactation support, and ‘rating your hospital experience’. And that didn’t include the personal messages celebrating the birth and the requests of marketing and political groups.

She welcomed the initial texts, LDI Senior Fellows Kimberly Waddell and Shivan Mehta reported in a JAMA Internal Medicine Viewpoint, but the messages soon grew tiresome.

The lesson? Texting works because users open 82% of text messages within five minutes, the authors note. The tool has boosted cancer screening and vaccination rates and helped patients avoid unnecessary hospital visits.

But the torrent can quickly become a tsunami, the authors say, turning off the very people the messages are meant to help.

Waddell, an Assistant Professor of Medicine at the Perelman School of Medicine, knows this turf well — studying how text messages, wearable devices, and gamification can improve health. Mehta, Associate Chief Innovation Officer at Penn Medicine, leads programs across the health system that use behavioral science, digital health, and AI to help more patients receive better care.

In their commentary, Waddell and Mehta offer five lessons on how providers can preserve this vital communication channel.

1. Don’t Just Send Texts. Integrate Them Into Care and Test What Works

Text messaging works best when embedded in clinical care and designed using behavioral science principles, recognizing that what motivates patients depends on context. We know little about which messages patients value and which improve care.

Evaluate texting programs using process and outcome measures—not just basic engagement metrics. When possible, also use randomized or strong quasi-experimental studies.

2. Personalized Texts Beat One-Size-Fits-All Messaging

Not all patients need the same reminder.

Someone who has never missed an appointment probably doesn’t need multiple reminders, while another patient with a history of missed visits may benefit from additional outreach.

Use AI and patient data to personalize communication based on previous behaviors, medical history, and scheduling needs. Personalized texts—based on patients’ activity levels and pain scores—have already been shown to reduce rehospitalizations after surgery.

3. Health Systems Need One Texting Strategy—Not Dozens

Hospital departments often operate independently. Primary care, specialty clinics, research studies, population health initiatives, and patient experience teams may all send texts without knowing what others are sending.

Create a centralized system that tracks every patient message across the organization. Comprehensive electronic medical records and multisite clinical trials offer features that could serve as models. Seek to eliminate duplicate communications and give teams a clearer picture of the overall patient experience.

4. One Well-Designed Message Can Be Better Than Three

Instead of sending separate reminders for every appointment or preventive screening, bundle related information into a single message when appropriate.

This can reduce notification fatigue while making it easier for patients to manage their care.

But not every message should be combined. Time-sensitive information or messages requiring immediate action may still warrant their own texts.

5. Conclusion: The Future of Patient Communication is Smarter—Not Louder

Strategies such as centralized tracking, personalization, message bundling, and behaviorally informed design can make communication more relevant and reduce message fatigue. These approaches should be rigorously evaluated to ensure text messages improve patient outcomes.

---

The Viewpoint, “Texting in Health Care—Preserving Impact and Managing Influx,” was published in JAMA Internal Medicine on June 29, 2026 by Kimberly J. Waddell and Shivan J. Mehta.

---

Author

---