International journal of medical informatics · 2025
Clinical effectiveness of a cloud-based dual-layer prescription review system: provincial integration across internet and outpatient care.
СтатьяПоддержка NIH: внешняя программаНазвание и аннотация: Перевод в очереди
Обработка статьи · этапы, ошибки и управление
Загрузка состояния…
Ошибки и ограничения
История · было → стало
Состояние обновляется каждые 5 секунд, пока блок открыт. Процент показывается только там, где известен объём работы. Пауза останавливает продолжение этапа; текущий запрос к источнику может завершиться.
Shi Jinming, Sun Dongxu, Kang Jian, Hao Wenjie, Huang Guoshu, Wang Feifei, He Xianying, Li Jia, Zhao Jie, Cui Fangfang
Аннотация
PURPOSE: Ensuring medication safety remains a pressing challenge in fragmented healthcare systems, particularly with the rapid growth of Internet Hospitals and limited pharmacist resources. Existing prescription review tools are often siloed and lack cross-institutional scalability. This study presents a cloud-based, dual-layer prescription review system (CEPR) designed to support provincial integration across Internet and outpatient care.
METHODS: CEPR adopts a microservices-based architecture and a two-tier rule engine encompassing 146,309 rules across 13,080 medications. It enables asynchronous, high-throughput review workflows with both standardized and institution-specific configurations. A retrospective dataset of 5,216 outpatient prescriptions and a prospective dataset of 922 real-world Internet Hospital prescriptions were used to evaluate clinical performance, diagnostic accuracy, and system efficiency.
RESULTS: In retrospective validation, CEPR achieved a precision of 94.23 % and F1-score of 96.96 %. Prospective validation yielded a sensitivity of 80.77 %, specificity of 99.33 %, and accuracy of 98.48 %, with 21 clinically inappropriate prescriptions successfully intercepted. System throughput reached 3,647 TPS, with 95 % of requests processed within 173 ms. The average prescription review time was reduced from 20 min (manual) to 2 min and 43 s (system-assisted), reflecting an 86.6 % gain in efficiency. False alerts were primarily due to terminology inconsistencies, which have been partially mitigated through synonym mapping and pharmacist feedback loops.
CONCLUSION: The CEPR system offers a scalable, clinically validated infrastructure for cross-institutional prescription oversight. Its dual-layer rule engine, asynchronous architecture, and interoperability with Internet and outpatient services make it a promising model for large-scale prescription governance. Planned enhancements include NLP-based semantic parsing to improve alert precision and reduce false positives.
METHODS: CEPR adopts a microservices-based architecture and a two-tier rule engine encompassing 146,309 rules across 13,080 medications. It enables asynchronous, high-throughput review workflows with both standardized and institution-specific configurations. A retrospective dataset of 5,216 outpatient prescriptions and a prospective dataset of 922 real-world Internet Hospital prescriptions were used to evaluate clinical performance, diagnostic accuracy, and system efficiency.
RESULTS: In retrospective validation, CEPR achieved a precision of 94.23 % and F1-score of 96.96 %. Prospective validation yielded a sensitivity of 80.77 %, specificity of 99.33 %, and accuracy of 98.48 %, with 21 clinically inappropriate prescriptions successfully intercepted. System throughput reached 3,647 TPS, with 95 % of requests processed within 173 ms. The average prescription review time was reduced from 20 min (manual) to 2 min and 43 s (system-assisted), reflecting an 86.6 % gain in efficiency. False alerts were primarily due to terminology inconsistencies, which have been partially mitigated through synonym mapping and pharmacist feedback loops.
CONCLUSION: The CEPR system offers a scalable, clinically validated infrastructure for cross-institutional prescription oversight. Its dual-layer rule engine, asynchronous architecture, and interoperability with Internet and outpatient services make it a promising model for large-scale prescription governance. Planned enhancements include NLP-based semantic parsing to improve alert precision and reduce false positives.