Currently submitted to: JMIR mHealth and uHealth
Date Submitted: Jul 6, 2026
Open Peer Review Period: Jul 10, 2026 - Sep 4, 2026
(closed for review but you can still tweet)
NOTE: This is an unreviewed Preprint
Warning: This is a unreviewed preprint (What is a preprint?). Readers are warned that the document has not been peer-reviewed by expert/patient reviewers or an academic editor, may contain misleading claims, and is likely to undergo changes before final publication, if accepted, or may have been rejected/withdrawn (a note "no longer under consideration" will appear above).
Peer review me: Readers with interest and expertise are encouraged to sign up as peer-reviewer, if the paper is within an open peer-review period (in this case, a "Peer Review Me" button to sign up as reviewer is displayed above). All preprints currently open for review are listed here. Outside of the formal open peer-review period we encourage you to tweet about the preprint.
Citation: Please cite this preprint only for review purposes or for grant applications and CVs (if you are the author).
Final version: If our system detects a final peer-reviewed "version of record" (VoR) published in any journal, a link to that VoR will appear below. Readers are then encourage to cite the VoR instead of this preprint.
Settings: If you are the author, you can login and change the preprint display settings, but the preprint URL/DOI is supposed to be stable and citable, so it should not be removed once posted.
Submit: To post your own preprint, simply submit to any JMIR journal, and choose the appropriate settings to expose your submitted version as preprint.
Warning: This is an author submission that is not peer-reviewed or edited. Preprints - unless they show as "accepted" - should not be relied on to guide clinical practice or health-related behavior and should not be reported in news media as established information.
Fluorescein-Based Delivery Validation of an Artificial Intelligence–Enabled Eye Drop Sensor for Medication Adherence Monitoring: Cross-Sectional Validation Study
ABSTRACT
Background:
Objective adherence monitoring is a major unmet need in mobile and connected health. Accelerometer-based smart eye drop bottles can classify instillation-like bottle motion, but a positive motion event does not prove that fluid reached the ocular surface. This dose-delivery confirmation gap is especially important for adherence monitoring because a sensor-positive event without actual delivery can provide false reassurance.
Objective:
We aimed to quantify agreement between artificial intelligence/sensor-classified eye drop instillation events and an objective fluorescein-based reference standard for ocular-surface delivery, emphasizing the false-assurance rate, defined as sensor-positive but delivery-absent events.
Methods:
In this prospective cross-sectional validation study, 100 ophthalmology outpatients performed one unaided instillation from a sensor-equipped 0.2% fluorescein bottle. Participants were assigned to bilateral, right-eye, or left-eye instillation. Blue-light slit-lamp photographs obtained immediately before and within 30 seconds after instillation served as the reference standard. Two ophthalmologists independently graded delivery while masked to the sensor output and to each other. The artificial intelligence/sensor classification was generated independently and masked to the photographic result. Each participant contributed one paired observation. Overall agreement, sensitivity, positive predictive value, and the false-assurance rate were calculated with exact 95% confidence intervals; bootstrap and Jeffreys-prior Bayesian analyses were prespecified as auxiliary uncertainty analyses. Reporting followed STARD 2015.
Results:
The two graders agreed for all 100 participants (Cohen kappa=1.00). The validation matrix included 98 sensor-positive/delivery-confirmed, 1 sensor-negative/delivery-confirmed, 1 sensor-positive/delivery-absent, and 0 sensor-negative/delivery-absent observations. Overall agreement was 98.0% (95% CI 93.0%-99.8%). Sensitivity was 99.0% (98/99; 95% CI 94.5%-100.0%), and positive predictive value was 99.0% (98/99; 95% CI 94.5%-100.0%). The false-assurance rate was 1.0% (1/99; 95% CI 0.03%-5.5%). The bootstrap 95% percentile interval for the false-assurance rate was 0.0%-3.1%, and the posterior probability that the true rate was below 5% was 98.3%.
Conclusions:
A positive artificial intelligence/sensor event from this smart eye drop bottle corresponded to objectively confirmed ocular-surface delivery in 99% of cases, with a false-assurance rate near 1%. These results validate the dose-delivery layer that motion-only adherence sensors typically lack, while the single supervised encounter establishes a controlled-condition performance ceiling and rationale for subsequent unsupervised home-use studies.
Citation
Request queued. Please wait while the file is being generated. It may take some time.
Copyright
© The authors. All rights reserved. This is a privileged document currently under peer-review/community review (or an accepted/rejected manuscript). Authors have provided JMIR Publications with an exclusive license to publish this preprint on it's website for review and ahead-of-print citation purposes only. While the final peer-reviewed paper may be licensed under a cc-by license on publication, at this stage authors and publisher expressively prohibit redistribution of this draft paper other than for review purposes.