Medical errors are 3rd leading cause of death. Cost estimated at $42B USD annually. MedGuard based on 1.3B of prescription big data, unsupervised learning and reinforcement learning, has implemented to 3 hospitals. >200 physicians and 250K patient encounters are using MedGuard every year, with >60% user acceptance rate and >85% by expert review, save 300-600M NTD hospital/year.

MedGuard is alert system running on CPOE, based on 1.3 billions of prescription big data, unsupervised learning and reinforcement learning, has implemented to three teaching hospitals. >200 physicians and 250,000 patient encounters are using MedGuard every year, with >60% user acceptance rate and >85% by expert review, based on reinforced learning for physician behavior resulted in 3% alert rate.

Medication errors cost a lot but preventable. Cost estimated 3.5B USD in the US, 126B NTD in Taiwan. Besides direct cost leading by errors, hospitals in Taiwan also suffer from punitive fine from NHI by 2-4M NTD. For example a young girl with cervical problem, was prescribed Filgrastim, defined as inappropriate by NHI, received 6.6M NTD fine. MedGuard can reduce errors and cost save 300-600M NTD estimated.