32. Perioperative monitoring of cognitive functions by event-related potentials and psychometric tests

Abstract

Introduction Postoperative cognitive dysfunction (POCD) is well recognized, but poorly understood syndrome. The role of used anaesthetics remains unclear. Neurophysiologic method – auditory event-related potentials (ERPs) are widely used for assessment of cognitive brain functions. Objectives To test the difference in occurrence of POCD between patients undergoing sevoflurane and propofol anaesthesia using standard psychometric tests and ERPs. Methods We present results of scheduled interim analysis of the prospective randomized trial after inclusion of 30 patients undergoing lumbar discectomy. Patients were randomized to receive either sevoflurane (group S; n = 18) or propofol anaesthesia (group P; n = 12). Anaesthesia depth was controlled by BIS monitoring. POCD was assessed using modified standard psychometric test battery and ERPs (wave N100 and P3) in defined time-points (preoperatively and postoperatively on day 1, 7 and 40). POCD was defined as a decline of more than one standard deviation in three or more tests. Results ERPs analysis of the S group showed increase in the latency (84.9 ± 1.4 vs. 81.4 ± 1, p = 0.03) and decrease of the amplitude (6.3 ± 0.6 vs. 8.0 ± 0.8, p POCD using standard tests was diagnosed in 6 patients (33 %) of S group versus 2 (17 %) in P, p = 0.41. Temporary occurrence of new POCD development did not differ between groups (4, 5 and 4 pts. vs. 1, 1 and 0 pts.; on day 1, 7 and 40 in groups S and P respectively; p = 0.31). In the S arm a significant decrease in Categorical Verbal Fluency score (CVF) was observed on day 1 (13.7 ± 0.8; p Conclusions ERPs analysis showed early auditory sensory impairment and late cognitive decline after sevoflurane use, however these changes were subclinical. POCD occurrence was non-significantly more frequent among patients after sevoflurane anaesthesia, possibly associated with short-term semantic memory impairment assessed by CVF decline. The study is supported by Charles University grant PRVOUK P36 and grant GACR P407/12/1525.