Enhanced visual memory effect for negative versus positive emotional content is potentiated at sub-anaesthetic concentrations of thiopental.

BACKGROUND: Emotional information has the ability to alter the formation and strength of a memory ('memory modulation'). Memory modulation by negative emotion is mediated by the amygdala. It is not known how gamma aminobutyric acid (GABA)ergic drugs affect the processes involved in memory modulation. This study investigates whether memory for negative emotional stimuli is more refractory to the effects of GABAergic drugs. METHODS: Eighty-three healthy volunteers were shown a randomized sequence of 60 visual stimuli consisting of negative, positive and neutral emotive pictures, while receiving a controlled infusion of thiopental (n=31), propofol (n=31), dexmedetomidine (n=10) or placebo (n=11). After a 5 h retention interval, when drug concentration was negligible, subjects performed a recognition task with 'old' pictures randomly mixed with 'new' pictures. Drug effect was calculated as the proportionate reduction in recognition for images of each emotional valence. RESULTS: Forty-eight subjects were included in a within-subject logistic dose-response model analysis. In the thiopental group there was a smaller drug effect seen for negative vs positive images (proportional memory reduction from baseline 0.27 (SD 0.20) vs 0.56 (0.25), P<0.001, n=20 included in analysis). A similar trend was seen in the propofol group (0.25 (0.28) vs 0.54 (0.30), n=10), but this did not attain statistical significance. No trend was seen in the dexmedetomidine group (0.33 (0.26) vs 0.24 (0.22), n=7). CONCLUSIONS: Over a specific dose range of thiopental (target serum concentration 2-7 micro g ml(-1)), impairment of explicit memory for images with negative emotional valence is less than that for images with positive emotional valence. There is a strong possibility that propofol (target serum concentration 0.3-2.4 micro g ml(-1)) causes a similar effect. Modulation of visual memory by negative emotional content continues at sub-anaesthetic concentrations of GABAergic drugs associated with explicit memory impairment.

Drug-induced amnesia is a separate phenomenon from sedation: electrophysiologic evidence.

BACKGROUND: Sedative-hypnotic drugs not only increase sedation, but also impair memory as serum concentration increases. These drugs also produce profound changes in the auditory event-related potential (ERP). The ability of various ERP components to predict changes in sedation and memory produced by various drugs was tested. METHODS: Sixty-five healthy volunteers randomly received intravenous placebo, midazolam, propofol, thiopental, fentanyl with ondansetron, or ondansetron alone at five different stable target concentrations (three increasing, two decreasing) using a computer-controlled infusion pump to produce varying degrees of sedation without loss of consciousness. ERPs were recorded while volunteer participants detected a deviant auditory stimulus and made a button-press response to a target tone (standard oddball paradigm, 80:20 ratio, to elicit a P3 response). At each target concentration, volunteers learned a list of 16 words. The predictive probabilities (Pk) of various ERP components were determined for word recognition at the end of the day (memory) and log reaction time to the deviant stimulus (sedation). RESULTS: The N2 latency of the ERP consistently predicted log reaction time in all groups (Pk +/- SE from 0.58 +/- 0.04 to 0.71 +/- 0.04). The N2P3 amplitude of the ERP was the best predictor of memory performance for midazolam (Pk, 0.63 +/- 0.04), propofol (Pk, 0.62 +/- 0.05), and thiopental (Pk, 0.66 +/- 0.04). There was a differential ability to predict memory performance from sedation for midazolam and propofol. CONCLUSIONS: Midazolam and propofol affect memory differentially from their sedative effects, and these are indexed by specific components of the auditory ERP. These components of the ERP are associated with specific, but not necessarily unique, neuroanatomic structures. Thus, these drugs act by additional mechanisms beyond general central nervous system depression to produce the effects of sedation and memory impairment.

Multiplicity of the alpha rhythm in normal humans.

Previous analyses of the alpha rhythm in a given single derivation have shown that it is a result of narrowband filtration of a broadband process. As many as three distinct alpha rhythms within each hemisphere could be differentiated in 65 resting, awake subjects by considering the spatial properties of these rhythms along with their dynamics. The analysis was based on cross-correlation measurements of rhythmic and broadband processes, and comparison of the dynamic characteristics of oscillatory systems underlying the alpha rhythms. Five-minute epochs of the EEG were recorded to provide 10% precision of the statistical estimates of the variables measured. A frontal alpha rhythm independent of occipital rhythmic activity was present in 20% of subjects. This third rhythm is distinct from the more familiar alpha and mu rhythms described in the literature, and is attenuated when the eyes are open. The authors demonstrate that the dynamic characteristics of the oscillatory systems underlying the alpha rhythm, as well as intrahemispheric coefficients of cross-correlation, are reproducible over time in a single subject. These parameters can form the basis for reliable individual EEG characteristics in the description of the "normal" EEG. The high correlation of the alpha rhythm in symmetric derivations could be explained by symmetric afferent impulse flows rather than by structural interconnection between the oscillatory systems.

Midazolam changes cerebral blood flow in discrete brain regions: an H2(15)O positron emission tomography study.

BACKGROUND: Changes in regional cerebral blood flow (rCBF) determined with H2(15)O positron emission tomographic imaging can identify neural circuits affected by centrally acting drugs. METHODS: Fourteen volunteers received one of two midazolam infusions adjusted according to electroencephalographic response. Low or high midazolam effects were identified using post-hoc spectral analysis of the electroencephalographic response obtained during positron emission tomographic imaging based on the absence or presence of 14-Hz spindle activity. The absolute change in global CBF was calculated, and relative changes in rCBF were determined using statistical parametric mapping with localization to standard stereotactic coordinates. RESULTS: The low-effect group received 7.5 +/- 1.7 mg midazolam (serum concentrations, 74 +/- 24 ng/ml), and the high-effect group received 9.7 +/- 1.3 mg midazolam (serum concentrations, 129 +/- 48 ng/ml). Midazolam decreased global CBF by 12% from 39.2 +/- 4.1 to 34.4 +/- 6.1 ml x 100 g(-1) x min(-1) (P < 0.02 at a partial pressure of carbon dioxide of 40 mmHg). The rCBF changes in the low-effect group were a subset of the high-effect group. Decreased rCBF (P < 0.001) occurred in the insula, the cingulate gyrus, multiple areas in the prefrontal cortex, the thalamus, and parietal and temporal association areas. Asymmetric changes occurred, particularly in the low-effect group, and were more significant in the left frontal cortex and thalamus and the right insula. Relative rCBF was increased in the occipital areas. CONCLUSION: Midazolam causes dose-related changes in rCBF in brain regions associated with the normal functioning of arousal, attention, and memory.

The comparative amnestic effects of midazolam, propofol, thiopental, and fentanyl at equisedative concentrations.

BACKGROUND: The authors evaluated the effects of midazolam, propofol, thiopental, and fentanyl on volunteer participants' memory for words and pictures at equisedative concentrations. METHODS: Sixty-seven healthy volunteers were randomized to receive intravenous infusions of midazolam (n = 11), propofol (n = 11), thiopental (n = 10), fentanyl with ondansetron pretreatment (n = 11), ondansetron alone (n = 8), or placebo (n = 16) in a double-blind design. Three increasing and then two decreasing sedative concentrations were achieved by computer-controlled infusion in each volunteer. Measures of sedation, memory, and drug concentration were obtained at each target concentration. Drug concentrations were normalized to equisedative effects using both Emax and logistic regression methods of pharmacodynamic modeling. The serum concentrations at 50% memory effect (Cp50s) were determined using four different memory end points. The relative potencies compared with midazolam for memory impairment were determined. RESULTS: Equisedative concentrations were midazolam, 64.5 +/- 9.4 ng/ml; propofol, 0.7 +/- 0.2 microg/ml; thiopental, 2.9 /- 1.0 microg/ml; and fentanyl, 0.9 +/- 0.2 ng/ml. The Cp50s for 50% loss of memory for words were midazolam, 56 +/- 4 ng/ml; propofol, 0.62 +/- 0.04 microg/ml; thiopental, 4.5 +/- 0.3 microg/ml; and fentanyl, 3.2 +/- 0.4 ng/ml. Compared with midazolam, relative potencies (with 95% confidence intervals) were propofol, 0.96 (0.44-1.78); thiopental, 0.76 (0.52-0.94); and fentanyl, 0.34 (0.05-0.76). Large effects on memory were only produced by propofol and midazolam. CONCLUSIONS: At equal sedation, propofol produces the same degree of memory impairment as midazolam. Thiopental has mild memory effects whereas fentanyl has none. Ondansetron alone has no sedative or amnesic effects.

Comparison of the EEG effects of midazolam, thiopental, and propofol: the role of underlying oscillatory systems.

The EEG effects of 3 intravenous sedative drugs from different chemical families were studied during conscious sedation in 47 normal volunteers. The drugs studied were midazolam (a benzodiazepine), propofol (an alkylphenol) and thiopental (a barbiturate). Though these drugs cause different degrees of amnesia, they have the common EEG effects of suppressing alpha-rhythm and increasing total beta-power. A large portion of the increase in beta-power can be accounted for by beta-rhythms. We used the UNIFAC-EEG technique to differentiate oscillatory systems underlying the rhythms induced by these drugs in a quantitative fashion. While thiopental induced beta-rhythms which were similar to those appearing during drowsiness, midazolam and propofol induced beta-rhythms with substantially different characteristics. The differences between the beta-rhythms induced by drug infusion and previously described 'sleep spindles' are discussed. We conclude that a quantitative analysis of beta-rhythms can differentiate the effects of these drugs on the EEG.

Relationship of early postoperative dysrhythmias and long-term outcome after resection of non-small cell lung cancer.

STUDY OBJECTIVES: To determine whether supraventricular tachydysrhythmias (SVTs) occurring early after thoracic surgery for non-small cell lung cancer (NSCLC) are associated with poor long-term survival. DESIGN: Prospective, cohort. SETTING: Referral cancer center. PATIENTS: Seventy-eight patients undergoing resection of NSCLC. INTERVENTIONS: Examination of univariate and multivariate effects of factors that might influence long-term survival: advanced age, sex, perioperative chemotherapy, extent of pulmonary resection, tumor stage, and SVT occurrence. RESULTS: In this group of patients, 10 of 78 (13%) developed early postoperative SVT. Log-rank analysis showed SVT occurrence (p = 0.01), age of 70 years or older (p = 0.04), and perioperative chemotherapy (p = 0.005) to predict poor long-term survival. Multivariate Cox regression analysis identified SVT occurrence (p = 0.007; relative risk [RR], 2.8; 95% confidence interval [CI], 1.3 to 6.1) and perioperative chemotherapy (p = 0.004; RR, 2.6; 95% CI, 1.4 to 5.1) to be independently associated with decreased survival. No other clinical or laboratory characteristic tested differentiated those patients who did or did not develop postoperative SVT. CONCLUSIONS: Early SVT occurrence after resection of NSCLC is associated with poor long-term survival. Although the etiology for this is unclear, this intriguing observation, not previously reported (to our knowledge), may be used in larger trials examining the effects of these and other factors on survival from lung cancer surgery.

Value of perioperative Doppler echocardiography in patients undergoing major lung resection.

BACKGROUND: The effects of major lung resection on right heart function have not been well established. Our goal was to evaluate these effects using serial Doppler echocardiography in the perioperative period. METHODS: In 86 patients undergoing lobectomy (n = 47) and pneumonectomy (n = 39), we examined the effects of pulmonary resection on perioperative changes in right heart function by transthoracic echocardiography. Serial echocardiograms were performed preoperatively, on postoperative day 1, and again between postoperative days 2 and 6 (median, 3 days) to evaluate cardiovascular function and to estimate right ventricular systolic pressure by the tricuspid regurgitation jet Doppler velocity method. RESULTS: Right or left atrial size, right atrial pressure, and estimated right ventricular systolic pressure did not differ between groups on the preoperative or postoperative day 1 examinations. However, on postoperative days 2 through 6 patients who underwent pneumonectomy had higher (mean +/- standard deviation) right ventricular systolic pressure values than lobectomy patients (31 +/- 15 versus 25 +/- 10 mm Hg, respectively; p < 0.05 by analysis of variance). In the subset of patients with percent predicted forced expiratory volume in 1 second less than 60% undergoing pneumonectomy (9/39), preoperative right ventricular systolic pressure was inversely correlated with percent predicted forced expiratory volume in 1 second values (r = -0.78; p < 0.04). This correlation was not significant in corresponding lobectomy patients. Postoperative right ventricular enlargement determined by echocardiography occurred with similar frequency in both groups and was associated with poor short-term prognosis in patients in whom severe respiratory failure developed. CONCLUSIONS: Preoperative indices of right heart function were within the normal range in both groups. Pneumonectomy but not lobectomy was associated with mild postoperative pulmonary hypertension that was not accompanied by significant right ventricular systolic dysfunction. Postoperative echocardiography may be useful to evaluate right heart function in critically ill patients after lung resection.

Brain tumors do not affect thiopental dosing requirements.

We used the biphasic electroencephalographic (EEG) response to increasing concentrations of thiopental to measure regional brain responses to thiopental. Eight patients with cortical parietal brain tumors, 3.3 (SD 1.3) cm in diameter, and eight control patients with lung cancer and normal brain computed tomography scans received thiopental by infusion (50-75 mg/min) until burst suppression (50% isoelectric activity) on the EEG occurred. Infusion lasted 10.7 (SD 2.4) min, and the average dose of thiopental administered was 810 (SD 170) mg [11.2 (SD 1.9) mg/kg]. During infusion the EEG was continuously recorded from the F3, F4, P3, and P4 electrodes. On-line power spectral analysis was performed, and data were saved for later analysis. Four EEG parameters [log beta (15-30 Hz) power, percent beta power, spectral edge 95% and spectral edge 70%] were plotted against calculated brain concentration of thiopental [using an assumed plasma-effect site rate constant (ke0) of 0.58] for each individual. Three points were measured on each curve (50% upslope, peak, and zero intercept) to quantitate the EEG response. Statistical comparisons were performed between the following sets of data: EEG response at electrode closest to brain tumor versus electrode farthest from tumor (in the same patient); and electrodes closest to brain tumors (parietal P3 and P4) versus same electrode pair in control patients (patients with thoracic tumors) using analysis. No differences were found in any comparison. Thus, the presence of a brain tumor does not affect the response of the brain in this region to thiopental as measured using EEG.

Clinical and echocardiographic correlates of symptomatic tachydysrhythmias after noncardiac thoracic surgery.

BACKGROUND: Supraventricular tachydysrhythmias (SVTs) following thoracic surgery occur with significant frequency and may be associated with increased morbidity. Prospective data on the etiology and importance of these dysrhythmias are sparse. METHODS: In 100 patients undergoing pulmonary resection without history of atrial dysrhythmias or previous thoracic surgery, we examined the effects of predefined risk factors by history, pulmonary function, and echocardiography on the incidence of postoperative SVT. Serial echocardiograms were performed preoperatively, on postoperative day 1, and again between postoperative days 2 to 6 (median = 3) to evaluate cardiovascular function and to estimate right ventricular systolic pressure (RVSP) by the tricuspid regurgitation jet (TRJ) Doppler velocity method. RESULTS: Symptomatic postoperative SVT occurred in 18 (18%) of the 100 patients studied at a median of 3 days after surgery and was disabling in 12 of 18 (67%). Digoxin loading was ineffective in controlling the ventricular response in 16 of 17 episodes. In the patients developing SVT, postoperative echocardiography revealed significant elevation of TRJ Doppler velocity (2.7 +/- 0.6 m/s vs 2.3 +/- 0.6 m/s, p < 0.05) but not right atrial or ventricular enlargement or right atrial pressure increase when compared with patients without SVT. Independent correlates of SVT determined in a stepwise logistic regression included intraoperative blood loss > or = 1 L (p = 0.0001) and a postoperative TRJ Doppler velocity > or = 2.7 m/s (p < 0.05). Patients who developed SVT had a higher rate of intensive care unit admission (p < 0.004), a longer hospital stay (p < 0.02), and higher 30-day mortality (p < 0.02). CONCLUSIONS: These prospective data suggest that increased right heart pressure but not fluid overload or right heart enlargement predisposes to clinically significant SVT after pulmonary resection. SVT may be an important marker of poor cardiopulmonary reserve in patients who develop significant morbidity after thoracic surgery. Early interventions to reduce right heart pressure may decrease the incidence of postoperative SVT and potentially improve overall surgical outcomes.

The P300 event-related potential during propofol sedation: a possible marker for amnesia?

We have studied the effects of conscious sedation with propofol on long latency components of the auditory event-related potential (ERP) in 10 normal volunteers (aged 21-41 yr) receiving propofol 75 micrograms kg-1 min-1 i.v. We examined the effects of propofol on ERP amplitudes and latencies, and their relationship to delayed recognition performance using a verbal memory test, a selective attention task (button pushing) and serum concentrations of propofol. During infusion of propofol, subjects were mildly sedated, oriented and readily responsive to verbal commands. ERP were recorded from monopolar FZ, CZ and PZ electrodes. We used a standard paradigm requiring selective attention to randomly occurring stimuli associated with a task (button push). The peak-to-peak amplitudes and latencies of the N2 and P3 waves were obtained before and during infusion, and 15, 100 and 170 min after infusion. Propofol produced a 70% decrease in the amplitude of P3 (P < 0.0001) from baseline and a 50% increase in reaction time. The differential response to target compared with non-target stimuli was maintained during infusion for both N2 and P3. Memory performance correlated more strongly with changes in P3 amplitude (r = 0.59) than with serum propofol concentrations (r = -0.07), although this correlation with memory did not reach statistical significance (P = 0.08). We conclude that P3 amplitude was profoundly affected by propofol given in sedative concentrations.

Impaired memory and behavioral performance with fentanyl at low plasma concentrations.

Fentanyl is commonly administered to conscious patients by continuous epidural or intravenous (i.v.) infusions, or by the transdermal route, which result in relatively constant, low, concentrations of the drug. Previous studies of memory and cognitive effects have not been performed at constant plasma concentrations of fentanyl. Based on simulated infusions using the pharmacokinetic modeling program IV-SIM, we administered fentanyl or placebo to nine healthy volunteers (aged 21-45 yr) by continuous i.v. infusion, targeting plasma concentrations of 1, 1.5, and 2.5 ng/mL in succession. A battery of memory and psychomotor tasks was administered at each plasma concentration of fentanyl, and at two points in the recovery phase while drug levels were decreasing. At increasing plasma concentrations of fentanyl, we found the following effects on memory (in comparison with placebo): a progressive decline in verbal learning (P < 0.03); decreased delayed recognition of words presented at different test times (P < 0.02); and decreased spontaneous recall of pictures shown during infusion (P < 0.03). Fentanyl at concentrations above 2.5 ng/mL caused a performance decrement of 15%-30% relative to baseline on all the psychomotor tests administered. Plasma concentrations less than 2.25 ng/mL had negligible effects on performance with the exception of the critical flicker fusion frequency, which decreased by 5 Hz at plasma concentrations between 1.5 and 2.25 ng/mL. Visual analog scale (VAS) measures of mental and physical sedation were significantly affected by fentanyl, but euphoria was not demonstrable. All subjects receiving fentanyl experienced severe nausea and four of six had one or more episodes of emesis (P < 0.03).(ABSTRACT TRUNCATED AT 250 WORDS)

Transcutaneous cardiac pacing during thoracic surgery. Feasibility and hemodynamic evaluation by transesophageal echocardiography.

BACKGROUND: Occasionally, emergency perioperative pacing is necessary. Transcutaneous cardiac pacing is noninvasive, safe, and readily available. Its feasibility and hemodynamic effects during thoracic surgery and one-lung ventilation have not been established. METHODS: Twenty anesthetized patients (aged 25-70 yr) without cardiac disease undergoing elective pulmonary resection (right n = 10, left n = 10) were studied in normal sinus rhythm and during transcutaneous cardiac pacing. Patients were paced in supine and lateral decubitus positions (with closed and opened chest) at the minimal current necessary to produce ventricular capture. Invasive arterial monitoring permitted calculation of mean arterial pressure, and transesophageal echocardiography was used to assess atrial and ventricular wall motion and the evaluation of transmitral flow. Twelve patients underwent Doppler analysis of pulmonary venous flow. RESULTS: Pacing was achieved in all patients, with a mean threshold of 86.9 +/- 20.6 mA for the right thoracotomy group, and 106.7 +/- 16.2 mA for the left thoracotomy group. The mean paced heart rates for the right and left thoracotomy groups were 101.6 +/- 18.2 and 105.4 +/- 11.5 beats/min, respectively. During pacing, all patients sustained reversible transient decrements in mean arterial pressure (9-19%) from baseline, the loss of AV synchrony, and the development of paradoxical ventricular septal wall motion. No patient had significant mitral regurgitation during sinus or paced rhythms. Decreased systolic pulmonary venous flow velocity and abnormal systolic flow reversal were seen during pacing in 11 of the 12 patients studied. CONCLUSIONS: Transcutaneous cardiac pacing is effective in patients undergoing thoracotomy and one-lung ventilation. Its use in patients in normal sinus rhythm induces reversible decrements in mean arterial pressure because of the effects of altered atrioventricular association, ventricular wall motion, and pulmonary venous return.

Propofol versus midazolam for monitored sedation: a comparison of intraoperative and recovery parameters.

STUDY OBJECTIVE: To compare intraoperative and recovery parameters in patients who received either propofol infusion (PI), propofol bolus (PB), or midazolam bolus (MZ) for sedation. DESIGN: Randomized clinical study. SETTING: Medical/surgical patients in a specialized hospital. PATIENTS: Ninety patients, aged 18 to 85 years, scheduled for central venous access for chemotherapy and/or total parenteral nutrition. INTERVENTIONS: In 30 patients, sedation was induced with MZ 0.02 mg/kg intravenously (i.v.), repeated every 2 to 3 minutes to achieve a sedation level of 3 (eyes closed, responds to verbal stimulus) (SL3). Maintenance was with MZ 0.005 mg/kg i.v. repeated as necessary to maintain SL3. In both propofol groups (30 patients each), induction of sedation was with a bolus of propofol 0.75 to 1.0 mg/kg i.v. Maintenance in the PB group was with propofol 0.25 mg/kg IV, repeated as necessary to maintain SL3. Maintenance in the PI group was with propofol 2 to 4 mg/kg/hr or 33 to 66 micrograms/kg/min to maintain SL3. MEASUREMENTS AND MAIN RESULTS: Blood pressure, heart rate, respiratory rate, oxygen saturation, and sedation level were monitored each minute for 5 minutes and then at 5-minute intervals during the procedure. A right atrial blood sample was taken for pH and partial pressure of carbon dioxide at maximum sedation. Adequate sedation was achieved in all three groups. The time to reach SL3 was significantly shorter in the PB group than in the PI and MZ groups (p < 0.05 and p < 0.01, respectively). Cardiovascular and respiratory parameters were remarkably stable. Immediate recovery, as judged by spontaneous eye opening, response to commands, and ability to state date of birth, was significantly shorter in both the PB and PI groups than in the MZ group (p < 0.0001). Intermediate recovery, as measured by sedation score at recovery entry, Aldrete score, and time to standing, was slower in the MZ group (p < 0.05 for the MZ group vs. the PB and PI groups for sedation score and Aldrete score; p < 0.05 for the MZ group vs. the PI group in time to standing). Psychomotor recovery, judged by digit symbol substitution tests, was significantly faster in the PB and PI groups (p < 0.05 vs. the MZ group). Amnesia, measured by picture recall, was significantly greater in the MZ group than in the PI and PB groups (p < 0.05). Mood changes were measured on a visual analog scale. All groups showed improvement. Nausea, headache, dizziness, blurred vision, appetite, tension, pain, depression, drowsiness, and ability to concentrate were evaluated in the preoperative and postoperative periods. The frequency did not differ significantly between groups due to confounding factors such as postoperative chemotherapy and premedicant drugs. CONCLUSION: The PI, PB, and MZ groups all gave excellent sedation for patients undergoing surgical procedures with local anesthesia. Amnesia was greatest with midazolam, and recovery was more rapid with propofol.

EEG and memory effects of low-dose infusions of propofol.

The purpose of this study was to identify EEG changes associated with low-dose propofol infusion producing only sedative effects, and to describe the memory effects of low-dose propofol infusion. Ten healthy volunteers underwent EEG monitoring (at Fz, Cz, Pz and Oz electrode sites) before, during and after propofol 0.5 mg kg-1 i.v. bolus and 75 micrograms kg-1 min-1 as an infusion. Mean serum concentration of propofol during infusion was 0.86 (SD 0.14) micrograms ml-1. The EEG changed significantly during infusion, with increased power in the beta 1 (15-20 Hz), beta 2 (20.5-30 Hz) and delta (1-3.5 Hz) frequencies. Beta 1 and beta 2 power changes were most marked at the Fz and Cz electrodes. Subjects were sedated, but able to complete cognitive tasks. Visual analogue scales of attention and sleepiness were obtained throughout the study and demonstrated a sedative effect during propofol infusion, but were not a significant factor in memory performance or EEG changes. A verbal learning task (Rey Auditory-Verbal Learning Task) administered before, during and after infusion showed a marked reduction in short-term memory capacity and dramatically impaired free recall and recognition during infusion. Nine of 10 subjects had partial amnesia for complex visual scenes presented during infusion, recalling less than 50% of the material. Stronger cueing was required to retrieve information presented during propofol infusion, with an increase in mean retrieval time from 95.4 (41.2) s to 426.8 (83.1) s. EEG and memory effects resolved quickly after the end of infusion.(ABSTRACT TRUNCATED AT 250 WORDS)

Effect of midazolam on the auditory event-related potential: measures of selective attention.

To elucidate the delayed effects of midazolam, we assessed electrophysiologic and motor responses by measuring auditory event-related potentials and a button-press reaction time response in 10 normal volunteers (aged 25-36 yr). Fifty minutes after intravenous infusion of 0.07 mg/kg of midazolam, subjects were mildly sedated, oriented, and readily responsive to verbal commands. To obtain ERPs, frequent tones (85%: 1000 Hz) and rare tones (15%: 2500 Hz) were presented at intervals of 1.5 s. Electroencephalographic signals were collected from FZ, CZ, and PZ for 1000 ms after stimulus presentation until 40 artifact-free rare-tone responses were obtained (average time, 6 min). Peak-to-peak amplitudes and latencies for N2, P3, and the subsequent negative slow wave (N3) were averaged within condition and were analyzed by repeated measures analysis of variance. After midazolam infusion, there was a 50% decrease in amplitude of P3 in response to target tones (P less than 0.006), whereas N3 latency increased by 40 ms (P less than 0.05). Event-related potential amplitudes were still significantly larger to rare (target) stimuli (P less than 0.003) after midazolam infusion. Although reaction time increased by 70 ms (P = 0.031), performance accuracy remained unchanged. Self-ratings of sleepiness and concentration show that a significant sedation effect was still present 50 min after infusion. Although routine clinical examination may be normal, full recovery from the effects of a typical intravenous dose of midazolam requires more than 50 min. The potential for adverse drug interaction, particularly with narcotics, is still present at this time.