Key takeaways
- The average no-show rate across 105 studies is about 23%, ranging from 13% in Oceania to 43% in Africa.
- A single WhatsApp reminder about 24 hours before the visit cut no-shows by about 24%. Adding an earlier second reminder brought no extra attendance benefit.
- Telling patients what a missed slot costs the health service cut missed appointments from 11.1% to 8.4% in a randomised trial.
- Machine-learning models can now flag likely no-shows with an AUROC of 0.85, which is useful for targeting extra reminders or smart overbooking.
How big is the no-show problem?
A no-show is a booked appointment that the patient neither attends nor cancels in time for someone else to use it. A systematic review of 105 studies in Health Policy found an average no-show rate of about 23%, with large regional differences: 43% in Africa, about 28% in South America and 13% in Oceania[1]. The factors most often linked to no-shows were a long lead time between booking and visit and a history of previous no-shows[1].
Even well-organised systems lose a meaningful share. A 2025 study covering more than one million appointments across 15 US family-medicine clinics recorded about 77,000 no-shows and a similar number of late cancellations[2]. In Buenos Aires' public system, more than a third of appointments (34.6%) were missed when no reminder was sent[3].
Every empty slot is lost revenue for the clinic, a longer wait for the next patient, and often a delay in care for the patient who didn't come.
1. Send a reminder, and get the timing right
Reminders are the best-studied fix. A 2026 meta-analysis of 10 randomised trials (8,236 patients) found that reminders raised hospital attendance by 11% (risk ratio 1.11, 95% CI 1.05–1.19)[4]. Phone-call reminders were clearly effective. The pooled SMS result pointed the same way but was not statistically significant on its own[4].
The newest evidence is about WhatsApp, which is how many patients in the Middle East, Africa and Latin America actually communicate. An analysis of 475,214 appointments in Buenos Aires compared four strategies[3]:
| Strategy | Approx. no-show rate | Relative risk vs none |
|---|---|---|
| No reminder | 34.6% | — |
| One reminder ~72 hours before | ~31.5% | 0.91 |
| Two reminders (~72 h + ~24 h) | ~27.6% | 0.80 |
| One reminder ~24 hours before | ~26.4% | 0.76 |
The single reminder sent about a day before worked as well as two reminders and better than one sent three days ahead. The authors' conclusion: timing matters more than frequency[3]. The study is a target-trial emulation posted as a preprint, so it is strong observational evidence but not yet a peer-reviewed randomised trial.
2. Make cancelling and rescheduling effortless
In the same WhatsApp study, the extra early reminder didn't improve attendance, but it increased patient-initiated cancellations[3]. That sounds bad, but a cancellation two days ahead is far better than a silent no-show, because the slot can be offered to someone else. If your clinic has a waiting list, an early reminder with a one-tap "I can't make it" option turns lost slots into filled ones.
The practical rule is to make it as easy to cancel or move a visit as it was to book it. If the only way to cancel is to call during opening hours, many patients won't.
3. Word the message carefully
What a reminder says matters as much as when it arrives. In two randomised trials in an English hospital trust, a reminder that told patients what a missed appointment costs the health service reduced missed appointments from 11.1% to 8.4% (odds ratio 0.74), and the effect was repeated in a second trial[5]. The message cost nothing extra to send.
Good reminders are short and specific: who the appointment is with, date, time, location, any preparation needed, and a clear way to confirm, cancel or reschedule.
4. Predict who is likely to miss, with models staff can understand
No-shows aren't random, and machine learning is now good at spotting them. In the 2025 primary-care study, a gradient-boosting model predicted no-shows with an AUROC of 0.85 and late cancellations with 0.92, and its predictions were not biased by sex or race/ethnicity[2].
Research published by IEEE emphasises that accuracy alone isn't enough: clinic staff must be able to see why a patient was flagged. Toffaha et al. built interpretable tree-based models for hospital no-shows in IEEE Access[6], and their review at the 2024 IEEE ICTMOD conference argues that explainable predictions are what make these tools usable in daily operations[7]. The same group's 2025 review of 52 studies found that simple logistic regression is still the most common model, used in 68% of them[8]. You don't need a deep-learning lab to get value.
5. Focus extra effort on high-risk appointments
A prediction is only useful if it changes what happens next. In a pragmatic randomised study at Kaiser Permanente Washington, sending an additional targeted text reminder to appointments with a high predicted no-show risk reduced missed visits in primary care and mental health[9]. Other targeted options include a personal phone call, offering an earlier or more convenient slot, or careful overbooking of slots with a very high predicted risk.
6. Watch the day in real time
Emirates Health Services in the UAE combined an AI no-show model (86% accuracy) with a live operations dashboard across its primary-care centres. Staff used it to re-allocate patients between clinicians during the day. Comparing the three months before and after rollout, the no-show rate fell by 50.7% from a baseline of about 21%[10]. Because this was a before-and-after study rather than a randomised one, part of the change may have had other causes, but it shows what's possible when prediction and scheduling are connected.
7. Measure your own no-show rate
You can't improve what you don't track. Start with three numbers: the no-show rate, the late-cancellation rate and how far ahead appointments are booked (the lead time). Break them down by doctor, day of week and appointment type. Lead time and previous no-shows are the strongest known predictors[1], so they're the first place to look.
A practical checklist for your clinic
- Send one reminder about 24 hours before each visit, on the channel your patients actually read (often WhatsApp).
- For appointments booked weeks ahead, add an early reminder whose main job is to free up slots.
- Let patients confirm, cancel or reschedule by replying, with no phone call needed.
- Keep the message short, specific and polite, and test wording that explains why attending matters.
- Keep a waiting list so that freed slots get re-booked.
- Track no-shows by doctor, day and lead time every month.
- Get consent before messaging and only send what's needed. Never include clinical details in a reminder.
Where Klinmate fits
This research shaped how we're building Klin, the AI front desk on WhatsApp inside Klinmate. Klin is designed to send reminders, let patients confirm, cancel or reschedule by chat, and hand anything it can't handle to your team. Klinmate is launching soon; join the waitlist to hear when it opens.
This guide is general information for clinic owners and healthcare professionals. It is not medical, legal or regulatory advice, and it is not a substitute for your own professional judgement.
References
- L. F. Dantas, J. L. Fleck, F. L. Cyrino Oliveira, and S. Hamacher, "No-shows in appointment scheduling – a systematic literature review," Health Policy, vol. 122, no. 4, pp. 412–421, 2018. doi.org/10.1016/j.healthpol.2018.02.002
- W.-J. Tuan, Y. Yan, B. Abou Al Ardat, T. Felix, and Q. Chen, "Predicting missed appointments in primary care: a personalized machine learning approach," Ann. Fam. Med., vol. 23, no. 4, pp. 294–301, 2025. doi.org/10.1370/afm.240316
- S. Esteban, G. Quintana, M. Sanchez, A. Rubinstein, and A. Szmulewicz, "Comparative effectiveness of single vs. dual WhatsApp reminders on no-shows: a target trial emulation within the public health system of Buenos Aires, Argentina," medRxiv (preprint, not peer reviewed), 2026. doi.org/10.64898/2026.08.17.26360609
- M. Al-Turbag, M. Mooney, and M. Corry, "A systematic review and meta-analysis of appointment reminders for enhancing hospital attendance," J. Hosp. Manag. Health Policy, vol. 10, art. 2, 2026. doi.org/10.21037/jhmhp-25-51
- M. Hallsworth, D. Berry, M. Sanders, A. Sallis, et al., "Stating appointment costs in SMS reminders reduces missed hospital appointments: findings from two randomised controlled trials," PLOS ONE, vol. 10, no. 9, e0137306, 2015. doi.org/10.1371/journal.pone.0137306
- K. M. Toffaha, M. C. E. Simsekler, A. Alshehhi, and M. A. Omar, "Predicting hospital no-shows: interpretable machine learning models approach," IEEE Access, vol. 12, pp. 166058–166067, 2024. doi.org/10.1109/ACCESS.2024.3490662
- K. Toffaha, M. C. E. Simsekler, and M. Omar, "Review of hospital outpatient no-show explainable prediction using machine learning," in Proc. 2024 IEEE Int. Conf. Technology Management, Operations and Decisions (ICTMOD), 2024, pp. 1–6. doi.org/10.1109/ICTMOD63116.2024.10878195
- K. M. Toffaha, M. C. E. Simsekler, M. A. Omar, and I. ElKebbi, "Predicting patient no-shows using machine learning: a comprehensive review and future research agenda," Intell.-Based Med., vol. 11, art. 100229, 2025. doi.org/10.1016/j.ibmed.2025.100229
- E. Ulloa-Pérez, P. R. Blasi, E. O. Westbrook, P. Lozano, et al., "Pragmatic randomized study of targeted text message reminders to reduce missed clinic visits," Perm. J., vol. 26, no. 1, pp. 64–72, 2022. doi.org/10.7812/TPP/21.078
- Y. M. AlSerkal, N. M. Ibrahim, A. S. Alsereidi, M. Ibrahim, et al., "Real-time analytics and AI for managing no-show appointments in primary health care in the United Arab Emirates: before-and-after study," JMIR Form. Res., vol. 9, e64936, 2025. doi.org/10.2196/64936
How we write guides: every figure is taken from the cited source, and we say where evidence is weaker (preprints, before-and-after or single-site studies). Spotted an error? Email [email protected] and we'll correct it.