Table 2.  The sequence of events in the investigation of four operating room incidents involving misadministrations via an infusion device.

Background: Investigating incidents related to human performance in anesthesia Case protocols collected prospectively based on cases presented at the internal morbidity and mortality conference Investigation I: Informal notification of an OR incident involving infusion device (no patient consequences) Interviews with participants within hours of incident to reconstruct case Bench testing of device behavior to corroborate sequence of events and to identify underlying contributors Results identified classic deficiencies in practitioner-computer cooperation   Reactions to incident by practitioners and management: Human error — "can’t make devices idiotproof" Coping strategies —"yeah, there are some device weaknesses, but I can handle it" Untrustworthy device —"you have to be careful, it can burn you" Investigation II: Exactly one week later a second incident occurs: another near miss Device captured in "impossible" state Interviews with participants immediately following incident to reconstruct case Investigation identifies central role of breakdowns in practitioner-computer cooperation Other methods employed to understand device use and breakdowns in context; observations of device use in context; more testing of device behavior unable to get any useful data on other incidents involving this device or similar devices from incident reporting systems   Reactions by stakeholders: cryptic incident report by device manufacturer refers to "custom" setup, states device worked as designed, implies erratic human behavior was responsible Device manufacturer reaction focused on, is there a patient injury? Will we be sued? Practitioners now see device interface as the source of incidents and difficulties; they report more cases   Investigation III: Two more incidents are reported and investigated "fresh" Report of the results of all of the investigations and studies documents the problems in practitioner-computer cooperation and how they contributed to incidents (Moll van Charante et al., 1993) Results lead to predictions of other problems; later, incidents occur where these problems are one contributor to the sequence of events (e.g., an event during transport to ICU) Researchers begin a follow-up project to redesign the device interface; the goal of the redesign is to show how to correct deficiencies in practitioner-computer cooperation (a) with this device, (b) with this class of devices and, (c) in general (Yue, Woods and Cook, 1992)   Aftermath: Investigators report incidents and results in specialty journal (Cook, Woods and Howie, 1992); give talks to research-oriented and technology-oriented anesthesiologists Use of device is reduced Leads to studies of different classes of infusion devices used in other contexts (Obradovich and Woods, 1996)