The Effects of Midazolam on Propofol-Induced Anesthesia

Propofol Dose Requirements, Mood Profiles, and Perioperative Dreams

Abstract

This study examined the effects of midazolam on the doses of propofol required for the induction of hypnosis and the maintenance of propofol/nitrous oxide anesthesia.In addition, the effects of midazolam on the time to patient recovery, perioperative mood profiles, incidence of perioperative dreams, patient satisfaction scores, and requirement for postoperative analgesics were assessed. This investigation was a prospective, randomized, and double-blind study of female patients undergoing dilatation and curettage. Patients received midazolam (30 micro g/kg, n = 30) or an equal volume of placebo (n = 30) immediately before the induction of anesthesia. Recall of dreams was assessed immediately postoperatively, in the postanesthesia care unit (PACU), and on the day after surgery using a questionnaire designed for surgical patients. Mood profiles were quantified using the Multiple Affect Adjective Check List-Revised, which was completed preoperatively and 1 h postoperatively. The Client Satisfaction Questionnaire-8, an eight-item self-administered version of the Client Satisfaction Questionnaire, was used to assess patient satisfaction on the day after surgery. Our results indicate that although the time to the loss of the lid reflex was significantly shorter in patients receiving midazolam (43.8 +/- 2.7 vs 74.7 +/- 7.6 s, P < 0.0003), there was no significant difference in the dose of propofol required to induce hypnosis or maintain anesthesia. There were no group differences in postoperative sedation and orientation scores, perioperative mood profiles, incidence of dreams, and patient satisfaction scores. More patients who received midazolam requested analgesics in the PACU (11 vs 4, P < 0.05). In conclusion, midazolam did not reduce the anesthetic dose requirement of propofol in patients undergoing anesthesia with nitrous oxide, nor did it accelerate patient recovery. Our results call into question the benefit of coinducing anesthesia with propofol and midazolam. Implications: Midazolam, administered immediately before anesthetic induction with propofol, did not decrease the dose of propofol necessary for hypnosis, nor the maintenance of surgical anesthesia, in female patients undergoing diagnostic dilatation and curettage. In addition, midazolam did not alter patient recovery characteristics, postoperative mood, incidence of perioperative dreams, or patient satisfaction. The use of midazolam was associated with an increased need for postoperative analgesics. Our study calls into question the benefit of administering midazolam immediately before anesthetic induction with propofol.

(Anesth Analg 1997;85:553-9)

Midazolam is used before the induction of general anesthesia for its sedative, anxiolytic, and amnestic properties ^[1]. Midazolam is also used for coinduction of anesthesia to accelerate the onset of hypnosis and to reduce the dose requirements of other drugs used to induce anesthesia ^[2,3]. The therapeutic advantage of coinduction is based on the principle of synergism: drugs interact synergistically when the effect produced by a combination of two drugs is greater than the sum of the effects produced by each drug alone ^[4]. By using drugs of two different pharmacological classes, the dose of each drug can be reduced so that side effects are minimized. Clinical and animal studies indicate that the combinations of propofol and midazolam ^[5-9]1, thiopental and midazolam ^[10,11], or methohexitone and midazolam ^[12] have a synergistic interaction with respect to their ability to produce hypnosis. In these studies, hypnosis was defined as either the inability to open eyes in response to a verbal command, or the loss of the lid reflex. In contrast to hypnotic effects, synergism between midazolam and propofol has not demonstrated when the effect end point is movement in response to a surgical stimulus. Indeed, hypnosis and surgical anesthesia are likely two distinct clinical effects that are mediated by different cellular mechanisms.

Clinical investigations have reported that midazolam (0.13 mg/kg) reduced the dose of propofol required to prevent movement in response to a tetanic stimulus ^[8]. However, the large doses of midazolam were inconsistent with those used in routine clinical practice. Other clinical studies have failed to demonstrate that midazolam, administered before the induction of anesthesia, reduced the doses of propofol required to maintain anesthesia ^[6]1. A confounding factor in these studies was that patients received a supplemental dose of another benzodiazepine or a short-acting opioid at the time of induction. In addition, the anesthesiologists administering the drugs were not blinded with respect to the treatment groups.

(¹) DeLucia JA, White PF. Effect of midazolam on induction and recovery characteristics of propofol [abstract]. Anesth Analg 1992;74:S63.

In recent years, numerous reports have documented subjective feelings of euphoria ^[13,14] and pleasant dreams ^[15] that occasionally accompany propofol anesthesia and sedation. We previously examined the perioperative mood profiles and incidence of dreams in patients receiving either propofol or thiopental anesthesia, in combination with nitrous oxide ^[16]. Using the Multiple Affect Adjective Check List-Revised (MAACL-R) ^[17], a self-administered questionnaire used to describe affect, we found that the patients who received propofol were more likely to exhibit sensation-seeking tendencies (i.e., adventurous, enthusiastic, and energetic) in the postoperative period. The incidence of perioperative dreams was small and similar in both groups.

The purpose of this study was to determine whether midazolam, administered immediately before the induction of anesthesia, influenced the dose of propofol required to maintain anesthesia in patients undergoing minor gynecological surgery. A midazolam dose of 30 micro g/kg was selected because it is a typical dose used in patients undergoing outpatient surgical procedures. In addition, we were interested in determining whether midazolam influenced the subjective phenomena associated with propofol anesthesia, as midazolam mitigates the dysphoria and unpleasant dreams associated with the dissociative anesthetic ketamine ^[18]. Therefore, we sought to determine whether midazolam influenced the perioperative mood alterations, incidence of dreams, time to recovery, and satisfaction in patients undergoing propofol-induced anesthesia. Finally, because benzodiazepines are known influence the threshold for pain and narcotic requirements ^[19,20], we wished to determine whether the administration of midazolam alters the requirements for analgesics in the postoperative period.

Methods

Approval from the Sunnybrook Health Science Centre Ethics Committee was obtained, as was informed consent from participants. Patients were considered eligible if they were female, ASA class I or II, more than 19 yr of age, and scheduled for outpatient diagnostic dilatation and curettage (D&C) with or without hysteroscopy. Patients were excluded from participating if any of the following conditions were identified: pregnancy; preexisting neuromuscular, hepatic, renal or cardiovascular disease; allergy or other contraindication to the use of the drugs used in the study; anatomic abnormalities that might contribute to difficult endotracheal intubation; a history of gastroesophageal reflux; inability to understand the consent form or research measures; and previous psychiatric disturbance or the use of mood-altering drugs. Patients were also excluded from analysis if there was deviation from the anesthetic plan. The patient's age, body mass index, and previous anesthetic experiences (major or minor problems) were noted. Education level (0 = some or all grade school, 1 = some or all high school, 2 = some college/university) was assessed to eliminate a comprehension bias. Sleep and dream patterns (Appendix) were recorded.

Sixty sealed envelopes were prepared: 30 were labeled midazolam and 30 were labeled placebo. On enrollment, patients were randomly allocated to either the midazolam or placebo group. A syringe labeled study drug was prepared on the day of surgery by an anesthesiologist not involved with the conduct of the anesthetic. The syringe contained either midazolam (30 micro g/kg) or an equivalent volume of normal saline (placebo; 0.03 mL/kg). The anesthesiologist administering anesthesia was blinded to the patients' group.

On the day of surgery, patients were taken to the operating room, and the standard monitors were applied. Immediately before the induction of anesthesia, the study drug was injected into a rapidly flowing intravenous (IV) line. This was followed by a bolus of propofol (1 mg/kg), which was administered over 0.5-1 min. The bolus of propofol was followed by incremental boluses of propofol (10-mg increments) every 5-10 s until the loss of the lid reflex was detected and the patient failed to open her eyes in response to a verbal command. Propofol was administered immediately after the study drug to prevent the anesthesiologist and the study nurse from determining the patient's treatment group. Anesthesia was maintained with intermittent injections of propofol (0.1-0.2 mg [center dot] kg^-1 [center dot] min^-1) and mask inhalation of 70% nitrous oxide and 30% oxygen. Dose adjustments were made in accordance with anesthetic depth as judged by heart rate, systolic blood pressure, and evidence of patient movement. No other anesthetic drugs were used.

Data were collected at seven time periods: Time 0, preoperatively in the outpatient department; Time 1, the time of induction of anesthesia; Time 2, when the lid reflex was lost and patient did not open her eyes to verbal command; Time 3, postoperatively when the patient first opened her eyes in response to a verbal stimulus presented every 5-10 s; Time 4, postoperatively, as soon as a sedation score of 2-3 had been reached (Appendix); Time 5, 30 min postoperatively in the postanesthesia care unit (PACU); and Time 6, the following day, by telephone. The patient's responses were assessed by the same nurse, who was blinded to the patient's treatment group.

The perioperative mood profiles, incidence of dreams, time to postoperative eye opening, and sedation and orientation scores were assessed in the following manner.

Mood was assessed using the MAACL-R ^[17]. This is a self-administered questionnaire in which patients are asked to choose from a list of 132 adjectives to describe their mood. The affect parameters that are measured are anxiety (A), depression (D), hostility (H), positive affect (PA), and sensation-seeking (SS). There are two summary scores: dysphoria (A+D+H), and positive affect/sensation-seeking (PA+SS). The checklist has two versions: one measuring trait and the other measuring state. The trait version asks respondents to mark the appropriate answers, which consider how they generally feel, whereas the state version asks them to consider how they feel today. The trait version measures normal affective traits, whereas the state version measures fluctuations resulting from different clinical situations. The MAACL-R generates raw scores, which are dependent on the adjectives the patient selects; for example, a higher anxiety score will result from the selection of adjectives that describe that affect and omission of those that do not. The MAACL-R was administered at Times 0 and 5.

Perioperative dreaming was assessed at Times 4, 5, and 6 with a questionnaire previously applied to general surgical patients ^[16]. We adapted the questions so that the patients were unaware that they were being questioned specifically about dreams. The time interval between Time 1 (the induction of anesthesia) and Time 3 (patients opened their eyes in response to a verbal command) was recorded by the study nurse. Sedation scores were measured at Times 4 and 5. The time between Time 1 (from the induction of anesthesia) until Time 4 (when the patient reached a sedation score of 2-3) was also recorded. Orientation scores were measured at Time 5 (30 min postoperatively). The induction and maintenance doses of propofol were recorded.

Patient satisfaction was measured at Time 6 using the Client Satisfaction Questionnaire-8, an eight-item self-administered version of the Client Satisfaction Questionnaire. This scale is a direct measure of patient satisfaction. That is, patients were asked to evaluate their satisfaction with specific encounters as opposed to satisfaction with health care in general. This measure has established psychometric properties, and has been used extensively in evaluation studies ^[21]. Respondents were asked to rate their perception on a scale of 1-4, based on the overall quality of service received. In addition, patients were asked questions about how satisfied they were with specific care. Each of the eight items had four response categories so that the higher the number, the more satisfied the respondent was with the care they received. The scale was additive so that the higher the overall score, the greater the patient satisfaction ^[22].

In the PACU, patients who complained of pain were given IV morphine. The PACU nurses were blinded to the patient's group and were unaware that postoperative pain was a measured outcome of the study.

Differences between the midazolam and placebo groups with respect to age, body mass index, propofol dose, time to the loss of the lid reflex, eye opening time, and MAACL-R trait scores were analyzed using an unpaired Student's t-test. A P value of less than 0.05 was considered significant. chi squared analysis was used to examine intergroup differences in previous anesthetic experience, education levels, sleep and dream patterns, incidence and nature of perioperative dreams, performance of hysteroscopy, requirements for analgesics in the recovery room, sedation scores, orientation scores, and patient satisfaction scores. The differences in state scores between the two groups were analyzed using an analysis of variance with a square root transformation for state sensation-seeking and anxiety to stabilize the variance. With 56 patients, and the alpha for the F test set at 0.05, there was an 80% likelihood (power) of showing a statistically significant interaction in the analysis of variance if the interaction accounted for 12% of the total variance (medium size effect).

Results

All 60 patients recruited for the study completed the protocol, with the exception of 4 patients, 2 from each group, who did not return the satisfaction questionnaire. No significant differences were observed between the treatment groups with respect to patient age, body mass index, education level, sleep and dream patterns, previous anesthetic experiences, or number of patients undergoing hysteroscopy (Table 1).

Table 1: Clinical and Pharmacologic Characteristics

Further, there were no significant differences in the doses of propofol used for the induction or maintenance of anesthesia between the two groups. Surprisingly, more of the patients who received midazolam requested analgesics in the PACU (11 vs 4, P < 0.05) (Table 1).

The data regarding patient awakening, sedation, and orientation are presented in Table 2. The time to the loss of the lid reflex was shorter in those patients who had received midazolam compared with those who had received the placebo (43.8 +/- 2.7 vs 74.7 +/- 7.6 s, respectively, P < 0.0003). The eye opening times, the interval between the induction of anesthesia and a sedation score or 2-3, the sedation scores at Time 4 and 5, and the orientation score at Time 5 were not significantly different between the midazolam and placebo groups.

Table 2: Patient Awakening, Sedation, and Orientation

No significant differences were observed in preoperative trait scores between the two groups.

The pre- and postoperative state scores are listed in Table 3. All patients were significantly less anxious, reported less sensation-seeking tendencies, and had a greater positive affect in the postoperative period compared with the preoperative period. The only significant differences between the midazolam and placebo groups were in the summary scores of dysphoria and positive affect/sensation-seeking; patients in the midazolam group had significantly lower dysphoric scores and significantly lower positive affect/sensation-seeking scores postoperatively compared with their scores preoperatively.

Table 3: Preoperative and Postoperative State Scores

The recall of dreams and the nature of the dreams were similar between the midazolam and placebo groups. In both treatment groups, the reported recall of dreams on awakening (midazolam group 27% vs placebo 22%) was greater than the reported recall of dreams in the PACU (midazolam group 12% vs placebo 12%) and the day after surgery (midazolam group 13% vs placebo 12%); however, the differences at these time intervals were not statistically significant.

Patient satisfaction scores are presented in Table 4. There were no significant differences between the groups for any of the measured variables.

Table 4: Patient Satisfaction

Discussion

Midazolam (30 micro g/kg) did not alter the induction or maintenance dose requirements for propofol in patients undergoing minor gynecological surgery. Midazolam did decrease the time to loss of consciousness; however, this decrease was small (approximately 31 seconds) and of little clinical consequence. In both groups, mood profiles showed predictable changes from the preoperative to the postoperative period; patients were less anxious and demonstrated a more positive affect. There were no significant differences between groups with respect to the times to awakening, the postoperative sedation scores and orientation scores, and patient satisfaction. The vast majority of perioperative dreams were pleasant in nature, but the incidence was not significantly different between the midazolam and placebo groups. Surprisingly, patients who had received midazolam were more likely to request postoperative analgesics in the PACU.

Previous studies reported a supraadditive interaction between midazolam and propofol for hypnotic effects ^[7-9]. In contrast, we observed that midazolam (30 micro g/kg) did not influence the dose of propofol required for either the induction of hypnosis or the maintenance of anesthesia. Pharmacokinetic factors could have contributed to the apparent lack of effect of midazolam on the induction dose of propofol. Sedation from midazolam occurs three to six minutes after IV injection ^[8], whereas propofol-induced sedation occurs more rapidly, within one arm-brain circulation time ^[23]. Synergistic interactions may only occur at specific dose ratios of the two drugs and, hence, only at limited time intervals after IV injection ^[3]. The dose of midazolam could have been too small, or, alternatively, insufficient time could have elapsed between the injections of midazolam and propofol. In the study that demonstrated synergistic interactions between midazolam and propofol ^[8], midazolam was administered two minutes before the injection of propofol; another two minutes elapsed before hypnosis was assessed. From a clinical perspective, the time required for midazolam to influence the dose requirements of propofol may not be justified in the face of busy operating room schedules. Also, in our study, the peak plasma concentration of midazolam would have been reached during the course of surgery, yet midazolam had no effect on the maintenance dose of propofol. As previously described, larger doses of midazolam would likely have reduced the anesthetic requirements for propofol. However, the increased doses of midazolam would likely be associated with side effects such as delayed emergence and prolonged recovery time ^[8].

The lack of effect of midazolam on the maintenance dose requirements of propofol could also be due to pharmacodynamic effects. Synergism is not simply a property of two drugs; it is highly dependent on the mixture or dose ratio ^[24,25]. Moreover, synergism may not be a feature of all clinical effects of the drug, such as hypnosis, anesthesia, or anxiolysis. Therefore, to fully explore drug interactions, a variety of drug doses should be examined. This was not the purpose of this study. Rather, we selected an effect-addition model, which determines whether a relatively inactive dose of one drug (midazolam) causes a significant increase in the effect produced by a dose of a second drug (propofol) ^[24]. In humans, a synergistic interaction for benzodiazepines and propofol has never been demonstrated for the clinical end point of anesthesia.

An unexpected finding was the increased analgesic requirement in patients who had received midazolam. The effects of benzodiazepines on nociception have been investigated in humans and animals. These studies have generated contradictory results ^[26]; some investigators report that benzodiazepines cause spinally mediated analgesia and enhance opioid-induced antinociception ^[27,28]. However, other studies in humans indicate that benzodiazepines do not augment the analgesic effects of morphine ^[29]. Furthermore, midazolam antagonizes the analgesic effects of morphine in rats ^[26], and benzodiazepines reduce the antinociceptive effects of systemic opioids ^[20]. Similar to barbiturates, the antinociceptive effects of midazolam may depend on the route of administration and the type of painful stimulus ^[30]. Regardless of the mechanisms, the increased requirement for narcotics in patients who have undergone gynecologic surgery and received midazolam warrants further investigation.

As in our previous work, patients were generally more positive and less anxious after diagnostic D&C. Consistent with this observation, propofol produced a pleasant affective state when administered in anesthetic and subanesthetic doses to rats ^[31]. As previously discussed, the mechanisms that underlie the mood changes with propofol are currently not known ^[16]. The incidence of dreams was higher in both groups in the present study compared with our previous investigation ^[16]. Patients in the present study were questioned immediately after awakening rather than 30 minutes postoperatively. The diminished incidence of dreams over time after propofol anesthesia has previously been described ^[15].

The patient satisfaction scores were all uniformly high. Positively skewed data with high favorable responses are often found in patient satisfaction measures ^[32], and our study was no exception. Neither the overall score nor the scores to each of the eight items yielded a statistically significant difference between the two groups, which suggests that either there were no differences in satisfaction or the dimensions selected to ascertain satisfaction were not sensitive to this clinical population.

Coinduction of general anesthesia with midazolam and propofol has recently gained popularity. A purported advantage is the reduction in the required dose of each drug. Our results suggest that the benefits of coinduction need to be seriously reevaluated. The addition of an unnecessary drug increases the possibility of a drug reaction, increases the cost of the anesthetic, and provides an additional opportunity for a drug error. To justify the use of an additional drug, meaningful changes to clearly defined clinical end points must be demonstrated. Presently, midazolam is not approved for coinduction, and no official dosing recommendations are available. The cost-savings associated with the use of midazolam are questionable, particularly when a vial of propofol is opened, and the remaining drug is discarded. Anesthesiologists have been cautioned against using the contents of vials that have been open for a prolonged time period because of the potential for bacterial contamination.

In summary, midazolam 30 micro g/kg compared with placebo did not reduce the anesthetic dose requirements of propofol in patients undergoing general anesthesia with N₂ O. The time to patient awakening, postoperative sedation and orientation, perioperative mood scores, patient satisfaction, and incidence of dreams was not different between the midazolam and placebo groups. Moreover, the incidence of patients requiring analgesics in the PACU was greater in the group that received midazolam. Our results call into question the benefit of coinduction of anesthesia with propofol and midazolam.

Appendix

Questions To Assess Sleep and Dream Patterns

1. In general, how much sleep do you feel you need in a night?

2. Do you usually have difficulty going to sleep?

3. Do you wake up during the night?

4. How often does this occur in a night?

5. Do you usually have difficulty going back to sleep?

6. Do you usually remember you dreams immediately after waking?

7. Do you usually remember your dreams throughout the day?

8. How often would you say that you have nightmares or unpleasant dreams?

9. Do you usually remember these immediately after waking?

10. Do you usually remember them throughout the day?

Sedation Score

1 = Awake

2 = Drowsy

3 = Asleep, but arousable

4 = Asleep, not arousable

Orientation Score

0 = None

1 = Orientation in either time or place

2 = Orientation in both

We thank Mr. Marko Katic and Dr. John Paul Szalai for the statistical analyses, and Ms. Patricia Watson and Dr. Richard McLean for reviewing the manuscript.