Midazolam is often used as an anxiolytic premedication before surgery. Although the sedative, anxiolytic, and amnestic properties of midazolam may be desirable before the induction of general anesthesia, residual effects in the immediate postoperative period may contribute to postoperative sedation, as well as to delayed recovery and discharge-readiness after brief outpatient surgery.
A single study examining these effects after brief outpatient general anesthesia found delayed emergence but no differences in recovery times between patients who did and did not receive midazolam premedication . Objective measures of postoperative sedation were not reported. Furthermore, the subjects of that study underwent a surgical procedure (dilatation and curettage) associated with minimum postoperative pain, and as a result, few patients received opioid analgesics postoperatively. Since opioids are known to potentiate the sedative effects of benzodiazepines [2,3], this study tests the hypothesis that midazolam premedication contributes to postoperative sedation and delayed recovery when administered to outpatients who typically experience postoperative pain requiring significant doses of opioid analgesics during recovery.
This prospective, double-blind study was approved by our institutional human subjects review board. Healthy ASA physical status I and II women scheduled for outpatient laparoscopic tubal ligation using Falope rings were considered for the study. Women with a history of alcohol or illicit drug abuse and those with allergy to study medications were excluded. Patients taking sedatives, antidepressants, and antiepileptic medications were excluded. Patients whose surgeons used local anesthetics or planned any surgery in addition to tubal ligation were also excluded.
After written, informed consent, subjects were randomized in a double-blind fashion to one of two groups. The midazolam group received intravenous (IV) midazolam 0.04 mg/kg, while the placebo group received IV normal saline solution. The study drug was prepared in 5 mL of saline and administered 10 min before the induction of general anesthesia. Anesthesia was induced with fentanyl 1.5 micro g/kg, propofol 2 mg/kg, and mivacurium 0.2 mg/kg. After tracheal intubation, anesthesia was maintained with oxygen, nitrous oxide 70%, and isoflurane. Ketorolac 0.5 mg/kg IV was administered after the induction of anesthesia. Inhaled anesthetics were discontinued at the end of the procedure, and upon emergence, the trachea was extubated and the patient was transported to the postanesthesia care unit (PACU). All patients received morphine sulfate 2 mg IV every 5 min as needed (PRN) for analgesia. Treatment for nausea was standardized: ondansetron 2 mg IV q 5 min PRN to a maximum of 6 mg, followed by metoclopromide 10 mg IV PRN. Unaware of study drug assignment, PACU nurses evaluated PACU discharge criteria in each patient every 10 min. Upon meeting institutional PACU discharge criteria (awake and aware of surroundings, able to breathe deeply and cough freely, blood pressure within 20% of preoperative values, minimum pain and no requests for further analgesia, minimum nausea and no requests for further antiemetic therapy), patients were considered PACU discharge-ready and were transported to the ambulatory surgery care unit (ASCU) prior to discharge home. Patients were considered ASCU discharge-ready upon meeting ASCU discharge criteria, which include vital signs (heart rate, blood pressure, and respiratory rate) within 20% of preoperative values, temperature <or=to38[degree sign]C, unobstructed respiration and ability to cough, no vomiting within 30 min of discharge, ability to take oral fluids, pain controlled with oral analgesics, ability to walk without becoming dizzy, and being awake and conscious of surroundings.
Demographic data recorded included age, weight, surgery time (incision to surgery end), and anesthesia time (induction to emergence). PACU arrival, PACU discharge-ready, and ASCU discharge-ready times were noted. Surgery end time to PACU arrival time was recorded. PACU recovery time was defined as PACU arrival to PACU discharge-ready, and ASCU recovery time as PACU discharge-ready to ASCU discharge-ready time. Postoperative pain and nausea were quantitated using a 100-mm visual analog scale (VAS), and they were assessed at 15, 30, and 60 min after PACU arrival. Total postoperative morphine sulfate and ondansetron doses were noted.
The digit-symbol substitution test (DSST)  and Trieger dot test (TDT)  were used to quantitatively assess sedation levels and psychomotor recovery. Patients performed the tests before (baseline) and 5 min after study drug administration and then 15, 30, and 60 min after arrival in the PACU. All patients were positioned with 30[degree sign] head elevation and used the same writing implement (ball-point pen, medium point, black) for all tests. The DSST score represented the number of correct symbol substitutions made in 60 s. The TDT score represented the total number of dots (of 42) connected. TDT deviation represented the cumulative shortest distance (in millimeters) between the drawn line and missed dots. To account for interpatient differences in test-taking ability, DSST and TDT scores and TDT deviations were normalized to baseline scores and deviation for each patient. Changes in test scores and TDT deviation relative to baseline values were compared. A single investigator blinded to group assignment performed all test scoring and calculations.
A sample size of 30 patients was determined by a power analysis based on the following assumptions: (a) a difference in Phase I (PACU) discharge time of 20 min (SD = 16 min) would be clinically significant; (b) alpha = 0.05; (c) beta = 0.10. All results are expressed as mean +/- SD. Demographic data and outcome data were analyzed and compared using the t-test or rank sum test, as appropriate (Sigmastat for Windows, version 1.0; Jandel Corp., San Rafael, CA). Fisher's exact test was used to compare incidence of antiemetic therapy. Pain VAS scores were analyzed by performing inference hypothesis testing from 95% confidence intervals for the means . Values of P < 0.05 were considered statistically significant.
Fourteen patients in the control group and 16 in the midazolam group were studied. A single patient in the midazolam group was excluded from analysis due to inability to perform the DSST and TDT during the postoperative period due to extreme somnolence. She was admitted overnight, and her somnolence was attributed to being awake working throughout the night before surgery, to which she attested postoperatively. There were no differences between control and midazolam groups with regard to age (34 +/- 6 vs 31 +/- 6 yr), weight (66 +/- 16 vs 66 +/- 13 kg), surgery time (28 +/- 10 vs 27 +/- 10 min), or anesthesia time (46 +/- 11 vs 46 +/- 13 min).
Outcome results are presented in Table 1 and Figure 1, Figure 2, and Figure 3. The surgery end to PACU arrival times were not different between control and midazolam groups (6 +/- 4 vs 8 +/- 4 min, respectively, P = 0.14). There were no differences in PACU or ASCU recovery times or mean cumulative morphine and ondansetron doses between groups (Table 1). The incidence of antiemetic therapy was not different between control and midazolam groups (6 of 14 vs 4 of 15, P = 0.45). Neither pain nor nausea VAS scores were different at any time point. There were no differences between control and midazolam groups in baseline DSST score (36 +/- 6 vs 34 +/- 7, P = 0.31), TDT score (36 +/- 6 vs 36 +/- 6, P = 0.59), or TDT deviation (8 +/- 10 vs 8 +/- 8, P = 0.65). Five minutes after study drug administration, patients in the midazolam group had significantly poorer performance on both tests compared with the control group (Figure 1, Figure 2, and Figure 3). Despite no difference in recovery times, the control group performed the DSST significantly better at 15 and 30 min in PACU (Figure 1). There were no differences in TDT scores throughout recovery (Figure 2), but mean TDT deviation was greater (dots missed by greater distance) in the midazolam group 15 min into recovery (Figure 3).
Preoperatively, DSST scores improved in the control group five minutes after saline placebo administration, which reflects learning behavior in that group (Figure 1). The opposite was observed in patients receiving midazolam, who experienced significant impairment in performance of both DSST and TDT relative to baseline (Figure 1, Figure 2, and Figure 3). Because study drug administration was the only intervention between baseline and post-premedication testing, midazolam alone accounts for the observed differences. This result is consistent with other studies of the psychomotor effects produced by midazolam [7-9].
Postoperatively, patients receiving midazolam premedication had increased levels of sedation 15 minutes (TDT deviation and DSST) and 30 minutes (DSST) after emergence from anesthesia compared with controls. The DSST has been compared with other tests and is highly sensitive in detecting psychomotor impairment due to midazolam for up to 45 minutes after a dose of 0.029 mg/kg IV . As would be expected, TDT deviation was more sensitive than TDT score in detecting psychomotor impairment (Figure 2 and Figure 3). The latter is determined only by the number of dots missed, whereas the former is influenced by both the number and magnitude of deviation. Residual systemic midazolam in the immediate postoperative period (approximately 60 minutes after premedication administration) likely accounts for the observed differences between groups. The additive sedative effects of midazolam premedication and IV morphine administered soon after PACU arrival in most patients may have further contributed to the observed differences [2,3]. The differences in sedation, as measured by DSST and TDT, could not be accounted for by differences in age, weight, duration of surgery, duration of anesthesia, or total cumulative doses of morphine administered in the PACU.
The presence of midazolam would be expected to potentiate the sedative effects of residual inhaled anesthetics and fentanyl during the immediate recovery period [2,3,10]. The general anesthetic was carefully controlled and standardized, although we did not measure or attempt to control the dose of isoflurane. Differences in isoflurane administration are unlikely to have contributed to increased sedation in the midazolam group. Larger isoflurane doses would be expected to prolong emergence; however, comparable surgery end to PACU arrival times between groups suggest that this was not so. Even if the midazolam group did have delayed emergence, residual isoflurane levels should be similar to or less than those of the control group at the time of emergence.
Despite greater impairment in the performance of tests of psychomotor recovery in the group receiving midazolam premedication, there were no differences in times to recovery and discharge-readiness. A potential limitation of this finding is the sample size studied. Our results revealed a SD of 22 minutes in PACU recovery time, which is larger than the SD assumed in the original sample size calculation. The actual sample size and SD of Phase I PACU recovery times provide an 80% likelihood of detecting a difference of 24 minutes. Differences less than 24 minutes are unlikely to be clinically significant, as institutional costs do not appear to be affected by such differences [11,12]. Furthermore, we observed no differences in recovery times despite the administration of a generous dose of midazolam (0.04 mg/kg), a short duration of general anesthetic exposure, and moderate doses of postoperative morphine. If postoperative sedation due to midazolam premedication does delay recovery, our choice of midazolam dose and surgical patient population would be expected to maximize the likelihood of observing such a difference. Finally, mean recovery times were 55 and 57 minutes. That there were no differences in psychomotor test performance 60 minutes into PACU recovery would suggest that midazolam is unlikely to be responsible for any differences in discharge times, if they exist.
Other investigators have reported little difference in recovery between groups premedicated with midazolam versus placebo [1,13,14]. Elwood et al  observed delayed emergence from anesthesia but found no effect of midazolam premedication (0.03 mg/kg or 0.06 mg/kg versus placebo) on recovery after very brief general anesthesia via a mask using alfentanil, propofol, and nitrous oxide. However, their patients underwent dilatation and curettage, a procedure associated with minimum postoperative analgesic requirements (only 4 of 64 patients required postoperative opioid analgesics). Also, their patients received only a small dose of alfentanil (10 micro g/kg) on induction and no volatile anesthetics. Objective tests of sedation were not reported. In a similar surgical population, Aantaa et al.  reported that midazolam 0.08 mg/kg given intramuscularly 60 minutes before induction of anesthesia delayed emergence compared with placebo controls, despite reduced intraoperative anesthetic requirements . They did not detect differences in psychometric tests of recovery (Maddox Wing and critical flicker fusion tests). Oral and intranasal midazolam premedication has been reported not to affect recovery times after general anesthesia in pediatric outpatients [13,14,16]. Compared with propofol, midazolam prolongs recovery times and impairs performance of tests of psychomotor recovery (DSST) after conscious sedation in adult outpatients [8,17].
Because benzodiazepines potentiate the sedative and hypnotic effects of opioids and volatile anesthetics, we sought to determine whether midazolam premedication augments postoperative sedation, which results in delayed recovery. We specifically chose to study this interaction in patients undergoing outpatient laparoscopic tubal ligation because of the short duration of surgery and the tendency of these patients to have high postoperative opioid requirements, with pain control the factor that most often limits discharge. Although midazolam premedication was associated with increased sedation up to 30 minutes into the recovery period, there were no differences in recovery times. These results do not support concerns regarding delayed recovery due to midazolam premedication after brief general anesthetics for painful outpatient procedures.
1. Elwood T, Hutchcroft S, MacAdams C. Midazolam coinduction does not delay discharge after very brief propofol anesthesia. Can J Anaesth 1995;42:114-8.
2. Tverskoy M, Fleyshman G, Ezry J, et al. Midazolam-morphine sedative interaction in patients. Anesth Analg 1989;68:282-5.
3. Kissin I, Brown PT, Bradley EL. Sedative and hypnotic midazolam-morphine interactions in rats. Anesth Analg 1990;71:137-43.
4. Hindmarch I. Psychomotor function and psychoactive drugs. Br J Clin Pharmacol 1980;10:189-209.
5. Newman MG, Trieger N, Miller JC. Measuring recovery from anesthesia: a simple test. Anesth Analg 1969;48:136-40.
6. Mantha S, Thisted R, Foss J, et al. A proposal to use confidence intervals for visual analog scale data for pain measurement to determine clinical significance. Anesth Analg 1993;77:1041-7.
7. Thapar P, Zachny JP, Thompson W, Apfelbaum JL. Using alcohol as a standard to assess the degree of impairment induced by sedative and analgesic drugs in ambulatory surgery. Anesthesiology 1995;82:53-9.
8. Ghouri AF, Ruiz MA, White PF. Effect of flumazenil on recovery after midazolam and propofol sedation. Anesthesiology 1994;81:333-9.
9. Claffey L, Plourde G, Morris J, et al. Sedation with midazolam during regional anaesthesia: is there a role for flumazenil? Can J Anaesth 1994;41:1084-90.
10. Melvin MA, Johnson BH, Quasha AL, Eger EI II. Induction of anesthesia with midazolam decreases halothane MAC in humans. Anesthesiology 1982;57:238-41.
11. Lubarsky DA. Understanding cost analysis. Part 1. A practitioner's guide to cost behavior. J Clin Anesth 1995;7:519-21.
12. Dexter F, Tinker JH. Analysis of strategies to decrease postanesthesia care unit costs. Anesthesiology 1995;82:94-101.
13. Vetter TR. A comparison of midazolam, diazepam, and placebo as oral anesthetic premedicants in younger children. J Clin Anesth 1993;5:58-61.
14. McMillan CO, Spahr-Schopfer IA, Sikich N, et al. Premedication of children with oral midazolam. Can J Anaesth 1992;39:545-50.
15. Aantaa R, Jaakola ML, Kallio A, et al. A comparison of dexmedetomidine, an alpha2-adrenoreceptor agonist, and midazolam as i.m. premedication for minor gynaecological surgery. Br J Anaesth 1991;67:402-9.
16. Davis PJ, Tome JA, McGowan FX, et al. Preanesthetic medication with intranasal midazolam for brief pediatric surgical procedures: effect on recovery and hospital discharge times. Anesthesiology 1995;82:2-5.
17. Pratila MG, Fischer ME, Alagesan R, et al. Propofol versus midazolam for monitored sedation: a comparison of intraoperative and recovery parameters. J Clin Anesth 1993;5:268-74.