Quality of Quality Data (QoQ)

Reliable quality data are an important basis for attempts to improve quality and safety of patient care. For anesthetic practice in Switzerland, an "Absolute Minimal Data Set" (AMDS) of preoperative patient characteristics and intra- and postoperative quality indicators is provided by the Institute of Social and Preventive Medicine (IUMSP, University of Lausanne) in cooperation with the Swiss Society of Anaesthesiology and Reanimation (SGAR-SSAR). Data are electronically forwarded by the participating institutions to IUMSP, whereas primary collection can be achieved by traditional paper records or electronic records as part of anesthesia information management systems (AIMS).

In the investigator's institution, physician and nurse anesthetists are supposed to use a window in the electronic anesthesia record for this purpose. This form should be completed at the end of each case. If an event according to the AMDS definitions occurs at least once during anesthesia, the respective box (e.g., "intraoperative hypotension") should be ticked in the form. The anesthesia record cannot be closed unless the quality form is filled, which can notably be done even in advance "on the quick" by ticking "no events". Considering the numerous duties of anesthesia staff at the end of a case, the investigators questioned the reliability of data generated during this busy phase.

A pilot study of 50 consecutive unselected cases of the year 2010 revealed a low rate of reporting (10.8%) of selected perioperative events related to anesthesia (specifically: hypotensive, hypertensive, bradycardic, tachycardic, and hypoxemic episodes). Consequently, the current extensive study with more representative sample size was initiated. To gain insight into possible causes (among others: time pressure, unclear definitions, fear of litigation), interviews with anesthesia staff are performed and will hopefully provide a basis for possible improvements. For the time being and considering the common nature of possible causes, the investigators suspect that their results may not be specific for their institution. The incidence of perioperative events may be grossly underestimated if the process of data collection is not properly designed and monitored.