OBJECTIVE

The aim of this study is to create a national database to record incidents that endanger patient safety. We try to identify systemic problems in hospitals in order to avoid safety incidents in the future and improve the quality of healthcare.

METHOD

The Taiwan Patient Safety Reporting System employs a voluntary notification model. We define 13 types of patient safety incidents, and the reports of different types of incidents are recorded using common terminology. Statistical analysis is used to identify the incident type, time of occurrence, location, person who reported the incident, and possible reasons for frequently occurring incidents.

RESULTS

There were 340 hospitals that joined this program from 2005 to 2010. Over 128,271 incident events were reported and analyzed. The three most common incidents were drug-related incidents, falls, and endo tube related incidents. By analyzing the time of occurrence of incidents, we found that drug-related incidents usually occurred between 8 and 10 am. Falls and endo tube incidents usually occurred between 4 and 6 am. The most common location was wards (57.6%), followed by intensive care areas (13.5%), and pharmacies (9.1%). Among hospital staff, nurses reported the highest number of incidents (68.9%), followed by pharmacists (14.5%) and administrative staff (5.5%). The number of incidents reported by doctors was much lower (1.2%). Most staff members who reported incidents had been working for less than five years (58.1%).

CONCLUSION

The unified reporting system was found to improve the recording and analysis of patient safety incidents. To encourage hospital staff to report incidents, hospitals need to be assisted in establishing an internal report and management system for safety incidents. Hospitals also need a protection mechanism to allow staff members to report incidents without the fear of punishment. By identifying the root causes of safety incidents and sharing the lessons learned across hospitals is the only way such incidents can be stopped from happening again.