Skip to Content

Digital Alzheimer’s Risk Disclosure Meets Trial’s Main Comparison

The November 18 eSMARTER trial supports a structured digital disclosure process, not an unexplained laboratory result appearing in a portal.
November 19, 2026 by
Michael Sinclair
| No comments yet

A structured digital platform for returning Alzheimer’s risk information met the main comparison with clinician-led telehealth in the eSMARTER trial, published in JAMA Neurology on November 18. The result concerns a designed disclosure process with education and support. It does not establish that an unexplained laboratory report appearing in a patient portal has the same effects.

The trial randomized 666 adults; 651 received APOE genetic disclosure and were included in that analysis. Its main measures concerned anxiety, disease-specific distress and knowledge. Digital disclosure met the prespecified noninferiority criteria. Optional later disclosure of the p-tau217 biomarker showed no significant between-group differences, but that should not be described as the same separately demonstrated noninferiority result.

The communication process was the intervention

A result can be delivered as a number, a short message or a structured explanation. Those are different experiences. The trial tested a particular platform against a particular clinician-led process, not the idea that any screen can replace any conversation. The content and route of communication are part of what was evaluated.

This is a useful distinction for digital health more generally. A service may be online while still having been carefully designed around the information people need. Conversely, putting information online does not automatically make the explanation clear or complete. The delivery format and the quality of the communication should not be confused.

Noninferiority has a specific meaning

A noninferiority study asks whether the tested approach performs no worse than the comparison by more than a prespecified acceptable margin on particular outcomes. It is not a declaration that every person had the same experience or that every possible effect was equal. The outcome and margin are essential parts of the claim.

That also explains why the main APOE finding and the optional biomarker analysis need different wording. A nonsignificant difference, by itself, does not prove equivalence. Keeping the statistical question visible prevents an encouraging main result from expanding into a stronger claim about every part of the study.

Similar main outcomes do not erase preferences

The paper reports higher satisfaction with clinician-led disclosure and a difference in recall at one follow-up point. Those observations remind readers that meeting a main endpoint does not make all other aspects of the experience irrelevant. A service can achieve its primary comparison while still having features that deserve improvement.

For patients, understanding a result includes knowing what questions remain and how to ask them. A person who reads an explanation but remains uncertain may need clarification, even if the average trial outcome looks favorable. Digital delivery should be judged by its whole support pathway, not only whether the information technically appeared on a screen.

The participants were not every possible clinic population

The trial screened psychological readiness and enrolled a largely White, highly educated, self-selected sample. Those limits matter when considering whether the process will work as well in other settings. A person’s language, reading needs, access to technology and available support can affect how a disclosure pathway is experienced.

Recognizing those questions is not a reason to dismiss digital communication. It is a reason to test and adapt it thoughtfully. A scalable service should not assume that people who were less represented in a study will automatically have an identical experience simply because the platform can reach them.

A risk result still needs its clinical purpose

The trial does not tell everyone to obtain these tests or interpret them independently. Why testing is being considered, what the result may change and what follow-up is available remain clinical questions. A study of communication does not turn a test into a universal screening recommendation.

Before receiving sensitive results, a useful conversation concerns how they will be explained and who can help afterward. That can apply whether the explanation comes through a clinician, a structured platform or a combination. The key is a clear route from receiving information to understanding what it means in the person’s own care.

eSMARTER provides evidence for one carefully designed digital approach. Its strongest message is not that clinicians are unnecessary, but that communication methods can be tested. Keep the tested process, selected participants and specific outcomes attached to the conclusion rather than reducing the finding to “online results are just as good.”


Sources and further reading

Michael Sinclair November 19, 2026
Share this post
Archive
Sign in to leave a comment