Accepted for/Published in: JMIR Formative Research
Date Submitted: Nov 3, 2025
Date Accepted: Jul 31, 2026
Effectiveness of the prospective prescription review system for reducing irrational prescriptions: a retrospective cohort study in a tertiary specialty hospital
ABSTRACT
Background:
Artificial intelligence–based prospective prescription review systems (AI-PPRS) can improve prescription rationality rate, but their effectiveness in high-volume specialty care settings is not well established.
Objective:
We aimed to evaluate the effectiveness of the AI-PPRS in reducing irrational prescriptions and to identify factors associated with successful interception.
Methods:
This retrospective cohort study analyzed all outpatient and emergency prescriptions issued between January 1 and December 31, 2024, at a tertiary EENT hospital in Shanghai, China. Among 2 559 342 prescriptions, 123 914 (4.84%) flagged as irrational by the AI-PPRS were included.
Results:
Of the 123 914 irrational prescriptions identified, 40 877 (32.99%) were successfully intercepted. The prescription rationality rate increased from 95.16% to 96.76% (absolute increase, 1.60%). In multivariable analysis, review level was the strongest predictor of interception (mandatory vs reminder: adjusted odds ratio, 5241.44; 95% CI, 3978.67–6976.77). The AI-PPRS intercepted 88.29 times more irrational prescriptions than pharmacists. After pharmacists revised the dosage rules of mometasone furoate nasal spray for children aged 3–11 years, the number of irrational prescriptions identified peaked at 1 511 times the pre-revision level in a single month.
Conclusions:
The AI-PPRS has great effect in reducing irrational prescriptions and improving prescriptions rationality. The system functioned as a valuable adjunct to pharmacists, with review level being the primary determinant of success. These findings suggest that AI-PPRS can enhance medication safety when integrated with appropriately configured rules and review levels, though future enhancements using machine learning may further optimize performance.
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.