False Negative BIS? Maybe, Maybe Not!

We appreciate Dr. Rampil’s interest and comments on our case report. We hope that the following will address some of his questions, as well as the questions of others who have contacted us directly. The case report was written after a retrospective examination of the medical record after the patient’s comments during the postanesthesia evaluation. The data presented was based primarily on the anesthesia record, which was handwritten by the anesthesiology resident during the case. Automated record keeping is not available at our institution, but it is our experience that the residents’ notes are usually quite accurate when compared to data saved in other monitors. Our residents are extensively educated in the proper use and interpretation of the Bispectral Index (BIS) monitor. They are instructed to observe the signal quality index and record a representative BIS value every 15 min. At the time of incision, the value recorded was 47; the subsequent value was 63. As mentioned, the concentration of sevoflurane had been decreased before incision, and it continued to decrease after sternotomy for hemodynamic reasons. In the 30 min after sternotomy, inspired sevoflurane was recorded as 0.8% in 70% nitrous oxide and 30% oxygen. End-tidal concentration was not recorded, but was monitored by the Ohmeda RASCAL (Datex Ohmeda, Helsinki, Finland). The BIS remained in the range of 50–69 for the next 90 min before the initiation of cardiopulmonary bypass. During this period the inspired sevoflurane varied from 0.8 to 1.7%. The patient reported no recall aside from the brief moment described in the case report. We believe Dr. Rampil may be correct in his implication that given the anesthetic conditions, recall would not be expected unless equilibration had not occurred. The studies cited, by Mollestad et al. (1) and Chortkoff et al. (2), however, do not support such a broad conclusion. Both studies examined the possibility for “wakefulness” or awareness in only 22 patients under controlled, prospective conditions. Given the rarity of intraoperative recall, this number falls far short of that needed to achieve statistical power. A recent study by Sandin et al. (3) prospectively examined 11,785 patients for intraoperative awareness, demonstrating a 0.18% incidence when muscle relaxants were used. They concluded that 861 patients would need to be monitored for awareness to avoid one incident of recall. Although statisticians may argue, thousands of patients would have to be studied before one could conclude that recall was impossible under the conditions studied by Mollestad et al. (1) and Chortkoff et al. (2). Nevertheless, we do feel that our patient had equilibrated, as he had been maintained on 3% sevoflurane in 70% nitrous oxide for 30 min, but it should be noted that the sevoflurane concentration was decreasing at the time of the event. Although this is speculation, considering the solubility characteristics of sevoflurane and its possible neuroexcitatory potential (4), rapidly changing brain pressures of sevoflurane might predispose susceptible individuals to brief moments of awareness. The BIS is updated every 2 s, but such a brief moment of EEG activity would likely not be detected, given the unit’s 30-s averaging time in calculation of the index. Indeed, it is unlikely that any monitor can be made that will detect these brief intervals of possible awareness and even more unlikely that an anesthesia provider could respond effectively. The Aspect A-2000 monitor (Aspect Medical Systems, Newton, MA) does have approximately 720 h of nonvolatile memory, and our event was recorded and the data subsequently extracted by Aspect. Dr. Paul Manberg has provided the graph (Fig. 1) that summarizes both the data from the monitor and the data on the anesthetic record. Sternotomy occurred at or about 1300 h. What Dr. Rampil does not mention is that this memory function is neither well known by the anesthesia community nor directly accessible to the user. For the data in Figure 1 to be generated, the monitor had to be downloaded by an Aspect representative using special software and transmitted to Aspect’s offices in Massachusetts, where the graph was produced. As we were not aware that this function even existed, the monitor had not been synchronized to the clock in the operating room. The time on the bottom of the graph has been corrected for a daylight savings difference of 1 h, but there is obviously ambiguity regarding the exact minute-to-minute times that events occurred. We have urged Aspect to make this capability available on-site or at least to better publicize its existence and stress the importance of synchronizing the monitor with the operating room clock.Figure 1: Graph generated by Aspect Medical Systems combining downloaded monitor data with data from medical record.Finally, our intent in writing the case report was to generate discussion about the characteristics of low-solubility volatile anesthetics and about possible limitations of the BIS monitor, which may lead to prospective studies. Because of the tendency of some practitioners to alter their practices based on a single case report, prominent authors have advocated that single case reports not be published (personal communication, Kenneth Mattox, MD, Baylor College of Medicine). Although we feel that they are valuable in generating discussion, the ambiguities discussed above illustrate Dr. Mattox’s point. George Mychaskiw II, DO Marc Horowitz, MD