In Reply:—

Thank you for the opportunity to respond to the letters by Dr. Leslie and by Dr. Sebel et al.  We welcome the opportunity to clear up some misconceptions about our study1that are put forward in those letters.

Dr. Leslie expressed concern that we had not defined awareness in our study. Dr. Brice et al.  2offered this definition: “awareness has been taken to mean the ability to recall, with or without prompting, any events which occurred during the period at which it was thought the patient was fully unconscious.” This has been an accepted definition for several decades and we, perhaps mistakenly, did not think it needed to be repeated.

Dr. Leslie and Dr. Sebel et al.  expressed concerns over the questions used in the study. We used the questions designed to elicit awareness as defined by Dr. Brice in 1970.2As was described in the article, we substituted two nonawareness questions with two anesthesia-related ones to increase sensitivity. Despite the correspondents' assertions, it is nonsensical to suggest that this substitution would cause a greater than 10-fold decrease in the incidence of anesthesia recall. Dr. Sebel et al.  themselves admit that there is no evidence that differing modifications of the Brice questions are any more or less effective at detecting awareness. We also believe that the correspondents are mistaken in their belief that direct questions are required to elicit recall. The use of nonthreatening questions has been previously advocated as a valid technique of eliciting the incidence of recall in patients,3thus obviating the need for the “crucial” question. It is our clinical experience that patients respond more openly to indirect and nonthreatening questions. We would also like to mention that with two interviews, a follow-up survey on anesthesia experience, and heightened public awareness of this problem, we still have not heard of any other cases.

Dr. Leslie and Dr. Sebel et al.  expressed some confusion with regard to the mention of dreaming in the study. It was not our intention to discuss the incidence or categorization of dreams in the article. The issue was only mentioned because an in-depth analysis of dreaming during anesthesia was not undertaken in the postoperative period.

Dr. Leslie and Dr. Sebel et al.  expressed concern about our use of a quality assurance program to collect data on patients. This process attempts to collect data on all patients undergoing anesthesia. The continuous quality improvement team completes this evaluation tool and ensures that all questions are answered and documented at several points during the patients' hospital stay. As was mentioned in the article, the validity of the quality assurance data are confirmed by an independent audit performed retrospectively, using a statistically significant sample size to ensure the accuracy of collected data. The article reports all cases and suspected cases uncovered for the study period. Readers can therefore examine the reported cases and make their own assessment as to whether possible cases of recall should have been included in this study. The article emphasizes that the purpose of such a program is to identify and remedy the causes of such instances through education and system process modification, not through obfuscation. Dr. Leslie is therefore correct in that a quality assurance program is designed to minimize the incidence of adverse events.

Dr. Sebel et al.  question the robustness of the data. We would like to reiterate that the purpose of a quality assurance process is to identify and monitor adverse outcomes to decrease these incidences through education and systems processes modification. This study illustrates the ongoing institution of a quality assurance process in a hospital system where new locations are added as the anesthesia group expands its practice. The majority of the patients included in the study were from the level I trauma and tertiary referral hospital where the system originated. Here, the quality assurance system failed to capture less than 10% of all cases. This compares favorably with the study of Dr. Sebel et al. ,4where they were unable to complete 33% of their interviews on enrolled patients. If the correspondents are concerned about the loss of comparatively few data points in our study, what must this say about their own? After all, a quality assurance database strives to collect information on all patients, whereas a research database is under no such constraint. The general impression that we are left with is that the use of a passive research database using recruited patients does a disservice to those same patients we are charged to protect.

Dr. Leslie suggests that the lower incidence of intraoperative awareness demonstrated in this study is due to lower risk patients in a community-based setting. This assertion conveniently overlooks the fact that the primary site in the study is a level I trauma center and tertiary referral hospital serving a population of several million. It is possible, but very unlikely, that this is one of the nation's busiest hospitals serving “not really sick” patients. The article discussed the fact that this hospital has a similar acuity, but a higher volume, than the three other teaching hospitals in North Carolina.

Dr. Leslie finishes by alleging that the study in question may have been influenced by investigational biases. Unlike the correspondents, none of the authors work for, or are compensated in any fashion by, a corporation that stands to make money from the data. Nor is there any intent or ability to sell or profit from the study. There have been multiple articles on the use of corporate funds to conduct research.5,6Lexchin et al.  7goes as far as to state that systemic bias favors products that are made by the company funding the research. One could rephrase one of Dr. Leslie's comments by saying that the aim of previous studies was to inflate the incidences of intraoperative awareness.

In summary, we understand that Dr. Leslie and Dr. Sebel et al.  are concerned about the implications of the study; however, we believe that these data represent a true picture of anesthesia awareness in clinical practice. The use of an independently audited quality assurance program allows practitioners to rapidly identify flaws in anesthetic techniques and practices, and to make the appropriate changes to decrease their incidence.

*Southeast Anesthesiology Consultants, Charlotte, North Carolina.


Pollard R, Coyle J, Gilbert R, Beck J: Intraoperative awareness in a regional medical system: A review of 3 years' data. Anesthesiology 2007; 106:269–74
Brice DD, Hetherington RR, Utting JE: A simple study of awareness and dreaming during anaesthesia. Br J Anaesth 1970; 42:535–41
Edited by Fleisher LA. Philadelphia, WB Saunders, 2004, pp 223–7Fleisher LA
WB Saunders
Sebel PS, Bowdle TA, Ghoneim MM, Rampil IJ, Padilla RE, Gan TJ, Domino KB: The incidence of awareness during anesthesia: A multi-center United States study. Anesth Analg 2004; 99:833–9
Tallon D, Chard J, Dieppe P: Relation between agendas of the research community and the research consumer. Lancet 2000; 355:2037–40
Bekelman JE, Yan L, Gross CP: Scope and impact of financial conflicts of interest in biomedical research. JAMA 2003; 289:454–65
Lexchin J, Bero LA, Djuulbegovic B, Clark O: Pharmaceutical industry sponsorship and research outcome and quality: Systemic review. BMJ 2003; 326:1167–70