SYSTEM FAILURE - AN ANALYSIS OF 2000 INCIDENT REPORTS

Although 70-80% of problems have some component of human error, its ov erall contribution to many problems may be small, studies of complex s ystems have revealed that up to 85% are primarily due to deficiencies in the lay-out and processes of the system. The anaesthetist has to op erate in a complex system, many problems originate from deficiencies i n this system. Information of relevance to system failure was extracte d from the first 2000 incidents reported to the Australian Incident Mo nitoring Study (AIMS). A system-based deficiency directly contributed to one-quarter of problems (four-fifths if human factors are included) , some aspect of the system minimized the adverse outcome in over half of all cases (four-fifths if human factors are included), and in two- thirds (three-quarters if human factors are included) a system-based s trategy would have been helpful; the system was implicated in 90% of a ll incidents (97% if human factors are included). Regardless of whethe r or not all human error should be regarded as part of the ''system'', attempts to modify its incidence and nature have to emanate from the system. AIMS reporting pathways and the organizations involved in deve loping and implementing strategies to improve the system operate at fo ur levels. Level I involves the use of AIMS reports by hospitals and g roup practices for audit at a local level. Level II involves AIMS part icipants sending forms to the AIMS central office, collated informatio n is then sent back to contributors by newsletter. Level III involves interaction between AIMS and the major professional bodies and level I V interaction between AIMS, these bodies and a variety of national and international agencies. Over 100 topics were identified from the AIMS data for consideration at one or more of these levels. AIMS has the p otential not only to play a vital practical role in the continued enha ncement of the quality of anaesthetic practice, but also to provide a valuable resource for research at the increasingly important interface between human behaviour and complex systems.