REAL-TIME IMPROVEMENT IN THE RELEVANCE OF PRESCRIPTIONS FOR ELDERLY PATIENTS: IMPACT OF AN AUTOMATED SYSTEM
Author(s)
Anne Bertheu, PharmD, PhD1, Cécile Carron de la Carrière, MSc1, Maryam Khiali, MSc1, Arnaud Boisgontier, MSc1, Marie-Ange RASENDRA, MSc2, Eléonore Herquelot, PhD2, Aurélie Schmidt, MSc2, Olivier Cuvillier, MD1.
1Vidal France, Issy-les-Moulineaux, France, 2Heva, Lyon, France.
1Vidal France, Issy-les-Moulineaux, France, 2Heva, Lyon, France.
OBJECTIVES: Optimizing the appropriateness of prescriptions is a key factor in improving healthcare efficiency. The Vidal Focus system continuously monitors prescribing data and uses rule-based algorithms to identify potentially risky situations in patients aged over 65. When an inappropriate prescription is identified, the prescriber receives a notification, enabling real-time adjustements. This study aimed to assess the feasibility, acceptability, and impact of the system under real-world conditions
METHODS: The study was conducted based on anonymised prescriptions issued between February and October 2025 by 8,000 general practitioners.
RESULTS: Out of the 1,163,857 anonymized prescriptions analyzed, 37.3% triggered a notification, representing 434,212 prescriptions identified as potentially at risk. The system led to a substantial number of corrections: 30,968 prescriptions were modified in real time, corresponding to 7.1%. The most frequent notifications involved PPIs (159,684), NSAIDs (134,693), benzodiazepines (110,572), and statins (93,611). For these four categories, the number of inappropriate prescriptions corrected was 7,246 (4.5%), 6,872 (5.1%), 5,126 (4.6%), and 2,437 (2.6%), respectively. Other situations associated with high correction rates were also identified: statins combined with fibrates (15.8%) and domperidone (12.0%). A very low disconnection rate was retrieved (remaining below 1%), demonstrating excellent integration into clinical practice.
CONCLUSIONS: This study demonstrated the feasibility of the automated system across a cohort of 8,000 physicians, highliting both its technical robustness, and its immediate effect on improving prescriptions appropriateness. Its high level of acceptability and its ability to handle a large volume of prescriptions demonstrate measurable improvements in patient care. These findings suggest that an enhanced version of the system, incorporating actionable recommendations, rule prioritization and reinforced learning mechanisms, could significantly contribute to addressing national challenges in terms of relevance, harmonization of practices, and efficiency of care.
METHODS: The study was conducted based on anonymised prescriptions issued between February and October 2025 by 8,000 general practitioners.
RESULTS: Out of the 1,163,857 anonymized prescriptions analyzed, 37.3% triggered a notification, representing 434,212 prescriptions identified as potentially at risk. The system led to a substantial number of corrections: 30,968 prescriptions were modified in real time, corresponding to 7.1%. The most frequent notifications involved PPIs (159,684), NSAIDs (134,693), benzodiazepines (110,572), and statins (93,611). For these four categories, the number of inappropriate prescriptions corrected was 7,246 (4.5%), 6,872 (5.1%), 5,126 (4.6%), and 2,437 (2.6%), respectively. Other situations associated with high correction rates were also identified: statins combined with fibrates (15.8%) and domperidone (12.0%). A very low disconnection rate was retrieved (remaining below 1%), demonstrating excellent integration into clinical practice.
CONCLUSIONS: This study demonstrated the feasibility of the automated system across a cohort of 8,000 physicians, highliting both its technical robustness, and its immediate effect on improving prescriptions appropriateness. Its high level of acceptability and its ability to handle a large volume of prescriptions demonstrate measurable improvements in patient care. These findings suggest that an enhanced version of the system, incorporating actionable recommendations, rule prioritization and reinforced learning mechanisms, could significantly contribute to addressing national challenges in terms of relevance, harmonization of practices, and efficiency of care.
Conference/Value in Health Info
2026-11, ISPOR Europe 2026, Vienna, Austria
Value in Health, Volume 29, Issue 12S
Code
MT13
Topic
Epidemiology & Public Health, Medical Technologies
Disease
No Additional Disease & Conditions/Specialized Treatment Areas