To the Editor:-Sanborn et al. [1]suggests that automated anesthesia records can identify, track, and report deviations from specific limits for physiologic variables, and that this might be preferable to self-reporting of adverse events. Copper [2]concluded in his accompanying editorial that "anesthesiologists...do not report most events meeting criteria that they themselves had defined as relevant to QA." We strongly disagree and published evidence to the contrary (Lagasse et al. [3]). In fact, we demonstrated that with a non-threatening QA system with 100% concurrent medical record review, anesthesiologists reported approximately 90% of adverse clinical outcomes. Our present QA system relies on self-reporting of adverse outcomes associated with 35,000 anesthetics performed annually. The resultant database contains over 1,200 reports since January 1995 and again approximately 90% of these were self-reported by the anesthesiologists.
We agree with Cooper that at least two steps are necessary for effective self-reporting: (1) demonstrate the value of reporting, and (2) change the culture that attributes error to negligence. [2]W. Edwards Deming, an industrial quality manager, published these ideas over a decade ago. [4]At our institution, adverse outcomes are analyzed by a structured peer review process and statistical process control is applied to the adverse outcomes as a measure of the quality of our perioperative care. The outcomes data is also used by members of the department for clinical investigations which are considered part of our quality management program and have resulted in improved care. This leads us to further disagree with Cooper who writes that "the benefit to patient care of anesthesia QA systems has not been established rigorously...." [2]In fact, several authors have demonstrated the benefit of quality management programs to perioperative patient care. [5-7]
The difference between our experience and that of Sanborn may lie in the nature of the quality management programs. Our peer review process looks at errors in the system as critically as we look at human errors. Thus, peer review is less threatening thereby encouraging anesthesiologists to share the responsibility with management for delivering quality health care. This does not appear to be one of the objectives of the quality management program described by Sanborn which excluded incidents detected by his automated record if they were part of the system's limitations (e.g., the patient's clinical condition or other factors necessitating acceptance of deviations in physiologic variables). [1]In our experience, system errors account for the vast majority of all adverse outcomes. Because human error contributes only a small portion to adverse outcomes (5-15%), programs which focus QA measures on human error to the exclusion of system error may misdirect resources.
In summary, we have published evidence that demonstrates reliable self-reporting of adverse outcomes originating from active participation in a nonthreatening quality management program. Our results show that such a system functions best when dependent upon self-reporting by anesthesiologists.
Robert S. Lagasse, M.D.
Department of Anesthesiology; Montefiore Medical Center; Jack D. Weiler Hospital; 1825 Eastchester Road; Bronx, New York 10461-2373
(Accepted for publication June 25, 1997.)