Memory function during propofol and alfentanil anesthesia: predictive value of individual differences

BACKGROUND: Conscious recall and implicit memory have been shown to depend on hypnotic state as measured by electroencephalographic (EEG) bispectral index (BIS). A third expression of memory (unconscious-controlled memory) was recently observed after moderate to light sedation (BIS, 70-80). The present study investigated memory function during deep sedation (BIS, 60-70). As memory effects are small, the authors studied potential predictors of individual differences in memory performance. METHODS: Memory function and speed of information processing were assessed in 56 outpatients before surgery. During propofol anesthesia supplemented with alfentanil, patients heard a series of words while anesthesia was titrated to BIS, 60-70. In between words, response to command was assessed using the Isolated Forearm Technique. The authors tested memory with a word stem completion task and process dissociation procedure to distinguish explicit from implicit effects. RESULTS: Mean (+/- SD) BIS during word presentation was 64.0 +/- 3. Patients with conscious recall of verbal commands (n = 9) did not recall or recognize presented words. Even so, the process dissociation procedure revealed evidence of memory by a significantly higher hit rate in the inclusion condition (0.26) than in the exclusion condition (0.12). Patients without conscious recall showed no evidence of memory for presented words. Hit scores correlated significantly with scores in the preoperative memory test (r = 0.35). CONCLUSIONS: The authors found evidence of weak explicit memory function during anesthesia titrated to BIS, 60-70. The observations strongly suggest that postoperative memory relates to awareness during anesthesia, but the nature of this relation remains unclear. Memory seems more likely in patients with good preoperative memory performance.