The introduction of synthetic opioids in the late 1970s enhanced the anesthesiologist's ability to provide hemodynamic stability during cardiac surgery. Although high-dose opioid anesthesia provides excellent intraoperative hemodynamic stability, recovery is frequently prolonged. The long-standing practice of mechanically ventilating coronary revascularization patients through the night after surgery is now being challenged [1-3]. There is increasing evidence that prolonged respiratory support does not improve patient outcome, and may actually increase the incidence of perioperative cardiopulmonary morbidity [4-6]. Additionally, as medical resources become more limited, expensive services, such as intensive care unit (ICU) admission, must be reserved for those patients in greatest need.
Few studies have compared the recovery characteristics associated with inhalational versus opioid-based cardiac anesthesia techniques. In a noncomparative study of 100 consecutive patients undergoing various types of cardiac surgery, Lichtenthal et al.  demonstrated that a halothane-based anesthetic permitted patients to be tracheally extubated within a mean time of 4.5 +/- 6.1 h after their operative procedures. An editorial suggested that inhalational anesthetics were the key to early extubation in the cardiac surgical patient . However, propofol is a rapid and short-acting intravenous (IV) sedative-hypnotic that can be used for maintenance of general anesthesia as part of a balanced anesthetic technique and may be an acceptable alternative to the volatile agents for cardiac surgery . Russell et al.  reported that the use of propofol could facilitate recovery and permit earlier extubation in patients undergoing coronary artery surgery.
This study was designed to determine whether the use of propofol for maintenance of anesthesia during coronary artery revascularization would shorten recovery times compared to those of enflurane, fentanyl, or thiopental without compromising intraoperative hemodynamic stability. Our hypothesis examined the effect of the maintenance anesthetic drug on cardiovascular stability and recovery profiles after coronary artery surgery.
We studied 90 patients scheduled for elective coronary artery bypass grafting and excluded patients with a left ventricular ejection fraction less than 0.4, patients requiring preoperative inotropic support, and patients with uncontrolled systemic disease (e.g., diabetes, hypertension, or renal failure). The Thomas Jefferson University Institutional Review Board approved this protocol, and written, informed consent was obtained from each patient. In this open-label, randomized study, patients received either enflurane, fentanyl, propofol, or thiopental for maintenance of anesthesia. Either of two surgeons (RNE, JDM) and one anesthesiologist (CTM) provided all surgical and anesthetic care.
All patients were premedicated with intramuscular injections of morphine (MS), 0.1 mg/kg, scopolamine, 0.3 to 0.4 mg, and oral diazepam, 0.15 mg/kg, 60 to 90 min prior to induction of anesthesia. IV and radial and pulmonary arterial catheters were placed using local anesthesia while patients breathed nasal oxygen, 3 L/min. Leads II and V5 of the electrocardiogram (ECG) were monitored continuously using a real-time oscilloscopic display, and a separate set of leads was used to obtain a magnetic tape recording of ECG leads II and CS5. ETCO2 was monitored before and after cardiopulmonary bypass (CPB) and maintained between 35 and 40 mm Hg. Cardiac output (CO) was determined using the thermodilution method, and triplicate values were averaged. Measurements of CO were repeated if intermeasurement variability was >or=to10%. Fifteen minutes prior to the induction of anesthesia, lactated Ringer's solution, 7-10 mL/kg IV, was administered, and a nitroglycerin infusion was initiated at 50 micro gram/min. Anesthesia was initiated with fentanyl, 25 micro gram/kg IV, and endotracheal intubation was facilitated with metocurine, 50 micro gram/kg IV, and pancuronium, 50 micro gram/kg IV. When the mean arterial pressure (MAP) exceeded the average ward MAP value by 10% or more, or if the absolute MAP value exceeded 95 mm Hg, the patients were randomly assigned to receive one of the four maintenance anesthetic regimens: 1) enflurane, 0.25%-2.0% (administered using a standard vaporizer); 2) fentanyl, 10 to 20 micro gram/kg IV, given in intermittent IV bolus doses (to a maximum dose of 150 micro gram/kg); 3) propofol, 50 to 250 micro gram centered dot kg-1 centered dot min-1 IV; and 4) thiopental, 100 to 750 micro gram centered dot kg-1 centered dot min-1 IV. The latter two drugs were administered as continuous IV infusions. Anesthetic and hemodynamic management algorithms are outlined in Figure 1.
During CPB, all study drugs were administered via the extracorporeal circuit (i.e., IV drugs were administered into the venous reservoir, and the enflurane vaporizer was in line with the bypass circuit sweep gas). All anesthetics were discontinued immediately after removal of the aortic cross-clamp. After discontinuing extracorporeal support, fentanyl, 10 micro gram/kg IV, was given to all patients, and the study drug was administered as needed to maintain hemodynamic stability for the duration of the surgical procedure.
After induction of anesthesia with fentanyl, 25 micro gram/kg, the study medication was titrated to maintain MAP values within 10% of the baseline values (defined as the average of the last three ward values) or between absolute MAP values of 65 and 95 mm Hg. In the enflurane, propofol, and thiopental groups, acute hypertensive reactions were initially treated by increasing the study drug rate of administration by 25% at 30 to 45-s intervals. In the fentanyl group, patients who became hypertensive were treated with a bolus of fentanyl, 10 micro gram/kg (if the MAP was >or=to110% to 120% of the averaged ward value) or 20 micro gram/kg (if the MAP was >120% of the averaged ward value). Hypertensive responses (i.e., MAP >110% of baseline MAP or MAP >95 mm Hg) that were not controlled with the maximum dose of the study drug Figure 1 were treated with sodium nitroprusside (SNP), 0.25 to 0.8 micro gram centered dot kg-1 centered dot min-1 IV. Hypotension was defined as an MAP less than 90% of the baseline value (or a minimum of 65 mm Hg) and was treated with an IV bolus of phenylephrine, 50 to 100 micro gram, or by an infusion of phenylephrine, 25 to 50 micro gram/min. Prior to CPB, tachycardia (heart rate [HR] >20% of the last three averaged ward values) was treated by increasing enflurane, propofol, or thiopental administration rates by 25% every 30-45 s or with a 10-micro gram/kg bolus of fentanyl. Tachycardia refractory to a doubling of the drug administration rates (enflurane, propofol, or thiopental) or to two supplemental doses of fentanyl was treated with esmolol, 0.1 to 0.3 micro gram/kg IV boluses. Bradycardia (HR <80% of the baseline values) associated with hypotension (MAP <65 mm Hg) and a decrease in the cardiac index (CI) (CI <2.5 L centered dot min-1 centered dot m-2) was treated with ephedrine, 5 to 15 mg, in the pre-CPB period. If necessary, external atrial or atrioventricular pacing was used to maintain the HR at 90 bpm at the end of CPB. If the pulmonary artery wedge pressure was more than 16 mm Hg, or in the presence of myocardial ischemia (defined as an ST segment depression of >or=to1 mm or elevation >or=to2 mm), the nitroglycerin infusion was increased to 100 to 300 micro gram/min IV. If myocardial ischemia was associated with a HR >100 bpm, it was treated with esmolol, 0.1 to 0.3 micro gram/kg IV.
The extracorporeal circuit included a hollow-fiber oxygenator primed with Normosol R (Abbott Labs, Chicago, IL), 1300 mL; 25% albumin, 200 mL; and mannitol, 12.5 g. After cannulation of the aortic root, a single venous cannula was placed, and hypothermic (28 degrees C) CPB was initiated. Pump flows were maintained at a CI of 2.1 to 2.3 L centered dot min-1 centered dot m-2, with an MAP between 40 and 80 mm Hg. Myocardial arrest and protection was achieved with cold, potassiumenriched (30 mEq/L) blood cardioplegia and topical saline/ice slush. Distal coronary anastomoses were completed during aortic occlusion, and the proximal anastomoses were placed during partial aortic occlusion.
Patients were considered successfully weaned from extracorporeal support when the CI was more than 2.3 L centered dot min-1 centered dot m-2 and the MAP exceeded 65 mm Hg. If necessary, inotropic and vasoconstrictor drugs were infused in the post-CPB period to achieve these hemodynamic end-points Figure 1. If the CI was >2.3 L centered dot min-1 centered dot m-2 but the MAP was <65 mm Hg, phenylephrine, 50 to 150 micro gram/min IV, was infused. An inadequate response to phenylephrine was treated with norepinephrine, 2 to 16 micro gram/min IV. If the MAP was >65 mm Hg, but the CI was decreased (CI <2.3 L centered dot min (-1) centered dot m-2), epinephrine, 2 to 10 micro gram/min IV, was administered. If the CI did not sufficiently improve, a loading dose of amrinone, 1.5 mg/kg IV, was given, followed by an infusion of 5 to 10 micro gram centered dot kg-1 centered dot min-1. An intraaortic balloon pump (IABP) was placed if pharmacologic interventions failed to provide adequate cardiovascular support. Decisions regarding the administration of postoperative inotropes and/or vasoconstrictors were made by the ICU staff.
We used a laptop computer to record systolic, diastolic, and mean systemic and pulmonary arterial pressures, central venous pressure, and HR values every 30 s. Using this computer record, we determined the amount of time that the MAP or HR was outside the predefined limits. The time elapsed from initiation of the maintenance anesthetic drugs to the return of the MAP and HR values to the predefined limits, as well as the total amount of time the patient was considered either hypertensive or hypotensive, was noted.
Systolic arterial pressure (SAP), MAP, diastolic arterial pressure, central venous pressure, pulmonary artery systolic pressure, pulmonary artery mean pressure, and pulmonary artery diastolic pressure, were obtained during the following times: 1) prior to anesthesia induction (control); 2) postendotracheal intubation; 3) 1 min after skin incision; 4) 1 min after sternotomy; 5) 2-5 min prior to aortic cannulation; 6) at the removal of the aortic cross-clamp; 7) 2 min after discontinuation of extracorporeal support; 8) at the end of sternal closure; and 9) at the end of skin closure. Additionally, CO determinations and pulmonary artery wedge pressure values were obtained at the above intervals: 1, 5, 6, 7, 8, and 9.
On postoperative days 1 and 2, we obtained 12-lead ECGs for evidence of perioperative myocardial infarction. Within 24 h of endotracheal extubation, all patients were questioned about intraoperative awareness by the principal investigator (CTM).
All postoperative patient management decisions were made by the ICU physicians and nurses, who were unaware of the study treatments. The ICU staff evaluated patients every 30 min and recorded the time at which patients first opened their eyes, responded to verbal communication (e.g., opened their eyes or moved in response to hearing their name), and could correctly respond to specific commands (e.g., take a deep breath or move a specific extremity). Data on postoperative cardiovascular drug requirements and the frequency of meperidine and diazepam administration were obtained from the hospital record. The times at which patients were extubated and discharged from the ICU were obtained retrospectively from the hospital record.
Intraoperative magnetic ECG recordings were analyzed on a Marquette Holter Analyzer System 8000 (Marquette Electronics, Milwaukee, WI). All tapes were reviewed by a cardiologist who was not associated with the investigation. The criteria for ischemia included the presence of an ST segment shift from baseline equal to or greater than 1 mm of ST depression measured 60 ms after the J-point, or 2 mm of ST elevation at the J-point persisting for at least 1 min. Two cardiologists (not participating in this study) independently reviewed all perioperative 12-lead ECG recordings. If a postoperative ECG had a new Q wave in one or more leads equal to or greater than 0.04 s, the patient was diagnosed as having a perioperative myocardial infarction.
Data were analyzed using the Systat Intelligent Software (Systat Inc., Evanston, IL) and are reported as mean +/- SD. Demographic data were analyzed using chi squared analysis (for discrete variables), the Kruskal-Wallis test, and one-way analysis of variance (for continuous variables). The hemodynamic data for the four groups were analyzed by univariate and multivariate analysis of variance for repeated measures. Since drug-induced hemodynamic changes over time were of clinical importance, we have also looked at drug-time interaction and reported these effects. A protected least significant difference test was used to perform multiple intra- and intergroup comparisons when significance was found. A P value of <0.05 was considered statistically significant.
Of the 90 patients studied in this investigation, 22 received only fentanyl for maintenance of anesthesia. Patients in the other three treatment groups received either enflurane (n = 24), propofol (n = 23), or thiopental (n = 21). The groups did not differ significantly with respect to age, gender, weight, body surface area, chronic (noncardiac) diseases, antianginal therapy, preoperative hemodynamic values, or intraoperative variables Table 1. Patients in the four treatment groups received either enflurane, 0.25%-2.0%, (average 0.75%); fentanyl, 108 +/- 33 micro gram/kg (6141 +/- 2588 micro gram); propofol, 10.8 +/- 5.0 mg/kg (884 +/- 466 mg); or thiopental, 33.5 +/- 17.3 mg/kg (2076 +/- 1390 mg). Seven of the 22 patients in the fentanyl group required the maximum allowable fentanyl dose (150 micro gram/kg) and required SNP to maintain the MAP within the predetermined limits. In the other three treatment groups, the MAP was adequately controlled with the study drug and SNP was not administered (P = 0.01). Three patients in the enflurane and fentanyl groups and two patients in the thiopental and propofol groups required esmolol for control of tachycardia not responding to the study drug (P was not significant).
Analyses of intraoperative hemodynamic variables revealed several intragroup differences Table 2. Prior to CPB, HR was increased at the time of intubation and at sternotomy in the fentanyl and thiopental groups. SAP and MAP were decreased after intubation in the enflurane and fentanyl groups. In enflurane-treated patients, the SAP and MAP remained below control values at each time period. With thiopental, diastolic blood pressure (DBP) and systemic vascular resistance index were significantly lower than control during the entire post-CPB period Table 2.
Intergroup differences for systolic blood pressure, MAP, and DBP are summarized in Table 2. After intubation, SAP was higher in the propofol and thiopental groups than in the enflurane and fentanyl groups. Immediately after sternotomy, SAP was significantly lower in the enflurane-treated patients than in the other three treatment groups. Propofol-treated patients had a significantly lower SAP just prior to aortic cannulation, compared to fentanyl patients. At the end of sternal closure, thiopental patients had significantly lower SAP, MAP, and DBP values, compared to the other three study groups.
There was no significant difference among the groups in the amount of time required to achieve blood pressure and HR control after the initial increase above the predefined limit Table 3. Similarly, there was no significant difference among the groups in the amount of time that patients were hypotensive either before, during, or after CPB Table 3. However, the duration of intraoperative hypertension was significantly shorter in the propofol group than in the other three treatment groups (P = 0.04) Table 3.
Prior to induction of anesthesia, four patients in the propofol and thiopental groups and five patients in the enflurane and fentanyl groups had Holter monitor evidence of myocardial ischemia. After anesthesia induction and prior to CPB, there were no significant differences between the groups in the incidence of myocardial ischemia. Four patients in the fentanyl group, three patients in the enflurane and propofol groups, and two patients in the thiopental group had evidence of myocardial ischemia.
The need for inotropic support and vasoconstrictor therapy in each of the four treatment groups is indicated in Table 4. Patients in the propofol and fentanyl groups more frequently required vasopressor support prior to CPB than did thiopental patients (P < 0.05). During CPB, fentanyl-treated patients more frequently required vasoconstrictors to maintain the MAP above 40 mm Hg than did patients treated with enflurane, propofol, or thiopental (P < 0.01). Fewer propofol (versus fentanyl) patients required inotropic drug support in order to be weaned from CPB (P < 0.05). Overall, 40%-65% of patients in all treatment groups required inotropic support in the post-CPB period; however, no differences were found among the four treatment groups. An IABP was placed in five patients, with one enflurane-treated patient (who died on postoperative day 2) and four fentanyl-treated patients (one of whom died intraoperatively) requiring IABP counterpulsation (P < 0.02 fentanyl versus propofol and thiopental).
Spontaneous eye opening occurred earlier in the propofol (123 +/- 77 min) and enflurane (161 +/- 112 min) groups than in either the fentanyl (222 +/- 149 min) or the thiopental group (337 +/- 221 min) Figure 2. Patients in the propofol group were able to respond to verbal stimuli and appropriately follow commands more rapidly than were patients in the other three treatment groups Figure 2. Propofol patients were also tracheally extubated significantly sooner (14.0 +/- 5.2 h) after arrival in the ICU than were patients in either the fentanyl (28.9 +/- 19.2 h) or the thiopental (19.9 +/- 3.8 h) group. There was no significant difference between extubation times in the enflurane (15.2 +/- 4.6 h) and propofol groups. There was also no significant difference among the groups in the amount of MS administered in the ICU (mean +/- SD) (enflurane: MS = 8 +/- 8 mg; fentanyl: MS = 15 +/- 15 mg; propofol: MS = 8 +/- 9 mg; and thiopental: MS = 9 +/- 9 mg). Finally, the number of hours that patients remained in the ICU did not differ among treatment groups (enflurane: 82 +/- 26 h; fentanyl: 66 +/- 22 h; propofol: 66 +/- 23 h; and thiopental: 71 +/- 26 h).
Three patients (one each in the enflurane, fentanyl, and thiopental treatment groups) reported intraoperative recall. Review of the patients' anesthetic records suggests that intraoperative awareness occurred before they received the maintenance study medication (e.g., after receiving fentanyl, 25 micro gram/kg). For example, the enflurane-treated patient recalled hearing conversation after she was paralyzed but denied recall of endotracheal intubation, skin incision, or sternotomy. In this case, enflurane was first administered for hypertension during skin incision. The thiopental-treated patient reported remembering his leg incision, which occurred prior to the initiation of thiopental therapy. The third patient recalled intubation and had been hypertensive immediately after laryngoscopy. Supplemental fentanyl was administered to treat the acute hypertensive response, and he had no other episodes of intraoperative awareness.
There were two patient deaths prior to hospital discharge. One enflurane-treated patient, who suffered a perioperative myocardial infarction, died 48 h postoperatively, and one fentanyl-treated patient died intraoperatively. Five patients had a perioperative myocardial infarction: two enflurane-treated patients and one patient from each of the other three treatment groups had new Q waves on their postoperative ECGs.
Assuming the following drug acquisition costs, we calculated the cost of the four maintenance anesthetic techniques based on actual drug use and waste per case: enflurane $135/50 mL, fentanyl $1.56/20 mL, propofol $9.80/20 mL, and thiopental $10.46 per kit. For the enflurane group, the average use was 13.2 +/- 5.8 mL per patient, for a mean cost of $7.12 +/- 2.9 (with a range of $4.54-$16.48). In the fentanyl group, the mean dose was 6141 +/- 2588 micro gram, for an average cost of $10.2 +/- 4.02 (with a range of $3.12-$15.60). In the propofol group, the mean dose was 884 +/- 466 mg for an average cost of $46.44 +/- 21.96 (with a range of $19.60-$98.00). Finally, the mean dose of thiopental was 2076 +/- 1390 mg for an average cost of $5.62 +/- 2.57 (with a range of $1.64-$13.36).
When shorter-acting drugs (e.g., enflurane and propofol) are administered for maintenance of cardiac anesthesia, an improved recovery profile can be achieved after coronary artery revascularization without compromising intraoperative hemodynamic stability. However, vasoconstrictor, inotropic, and vasodilator drugs may be frequently required to optimize the intraoperative hemodynamic profile, in particular after CPB. This study demonstrated that the use of propofol, as an alternative to high-dose fentanyl and thiopental, shortened the time to patient awakening and extubation but did not affect the duration of time the patient stayed in the ICU.
Propofol patients tended to have fewer hypertensive episodes than patients in the other three groups. Since propofol is both an arterial and a venous vasodilator, hypertensive responses to surgical stimuli can be rapidly controlled using a variable-rate propofol infusion [10-12]. In contrast, it appears to be more difficult to deepen anesthesia quickly with enflurane during cardiac surgery . However, the use of higher total gas flows (>2 L/min) would have facilitated the control of acute hypertensive responses with enflurane.
Intraoperative vasoconstrictor and inotropic drug requirements differed among the groups. The large dose of fentanyl required in the pre-CPB period, to suppress the cardiovascular responses to sternotomy and mediastinal dissection, resulted in significant vasoparesis and an increased need for phenylephrine and norepinephrine during and after CPB . Since all anesthetic drugs were discontinued at the time of aorta cross-clamp removal, enflurane and propofol levels in the blood were low at the time of discontinuation of CPB. Thus, the myocardial depressant effects of these drugs would be minimized. This may account for the decreased requirements for vasopressors and inotropic support in these two groups. The decreased systemic vascular resistance after CPB in the thiopental group may be secondary to the peripheral vascular effects [15,16].
The hemodynamic effects of the anesthetics used in this study are similar to those reported by other investigators over the last decade. Analogous to the findings of Philbin et al. , Wynands et al. , and others, our findings were that high-dose fentanyl did not predictably blunt the hemodynamic responses to surgical stimuli in patients undergoing coronary revascularization [14,17,18] and may result in increased requirements for alpha-agonist drugs . Enflurane compares favorably to opioids and other inhalational anesthetics in its ability to maintain cardiovascular stability during cardiac surgery [13,19-21]. However, Samuelson et al.  reported that enflurane-based techniques required "considerable fine-tuning of the dose of enflurane" to maintain hemodynamic stability compared to sufentanil. We also found that responses to surgical stimuli of differing intensities may be more difficult to control with enflurane.
Propofol causes a 20%-40% decrease in blood pressure when it is used for induction of anesthesia in patients undergoing cardiac surgery [8,22-24]. Although many clinicians are reluctant to use propofol in patients with ischemic heart disease, excellent hemodynamic stability can be achieved when it is administered via a variable-rate continuous infusion for maintenance of anesthesia [9,23,24]. In a study of elective coronary revascularization patients receiving propofol, 2.5 mg/kg IV, for induction of anesthesia, Kaplan et al.  reported a mean decrease of 42% in MAP. Hall et al.  also reported a significant decrease in blood pressure when propofol was used for induction of anesthesia but stable hemodynamics when propofol was used as a continuous infusion for maintenance of anesthesia during coronary revascularization. In a large multicenter study comparing propofol and sufentanil, propofol, when administered using a computer-controlled infusion pump, was not associated with an increased risk of hypotension during induction or maintenance of anesthesia prior to CPB, and sufentanil-treated patients experienced more frequent hypertensive episodes .
Thiopental was included in our study in order to compare propofol to a commonly used IV sedativehypnotic. In a study reporting on a similar dose of thiopental (39.5 mg/kg) in cardiac surgery patients, there was no difference in the incidence of hypotension among patients treated with thiopental and those anesthetized with fentanyl and enflurane during CPB . In contrast to our findings, significantly more thiopental (versus fentanyl/enflurane)-treated patients in the study by Nussmeier et al.  required inotropic support during discontinuation of CPB. Although the use of a midazolam infusion was included in our initial protocol design, this group was discontinued because of difficulty controlling acute hypertensive responses and prolonged times to awakening and extubation compared to the other treatment groups.
During the perioperative period, both patient and nonpatient factors determine when a patient can be tracheally extubated after coronary revascularization. Clearly, a patient must be hemodynamically stable, normothermic, and have adequate hemostasis prior to being considered for extubation. Other factors that affect the timing of tracheal extubation include: 1) the type and amount of perioperative drugs administered; 2) the ICU protocols and standards of care pertaining to postoperative respiratory management; and 3) the personal preferences of the ICU health care professionals responsible for the patient [26,27]. Sanford et al.  commented on the importance of ICU personnel practices on the timing of extubation after cardiac surgery in their study.
While it is difficult to assess the effect of any one drug on postoperative recovery, because of the nonanesthetic factors that determine the timing of extubation, a review of the literature identifies anesthetic techniques that are likely to result in prolonged ventilatory support after cardiac surgery and techniques that may facilitate early extubation Table 5. Analogous to these previous investigations, our findings were that the high-dose fentanyl technique resulted in prolonged tracheal intubation, even though patients were responsive to verbal stimulation within 5 h after the end of the operation. In contrast to our findings, de Lange et al.  reported that 80% of their fentanyl-treated patients were ready for extubation within 4.6 +/- 1.8 h of their cardiac procedures; however, these investigators did not report when the patients were actually extubated. Although their patients may have appeared ready for endotracheal extubation, the subtle factors that affect the timing of extubation were not assessed in this study.
Several investigators have reported on the use of enflurane with moderate doses of opioids, benzodiazepines, and/or nitrous oxide for anesthesia in cardiac surgery [13,19-21]. In these studies, enflurane provides acceptable intraoperative hemodynamic stability and permits extubation within 16 h after cardiac surgery. Recent reports also suggest that patients receiving propofol may have a shortened recovery time after cardiac surgery [3,9]. In the study by Russell et al. , patients were responsive within 1 h after completion of their cardiac surgery when propofol was used in combination with fentanyl, 20-40 micro gram/kg, for induction and maintenance of anesthesia. In contrast, Hall et al.  reported that patients anesthetized with propofol and sufentanil required endotracheal intubation for an average of 22 h after their cardiac operation Table 5. These investigators suggested that the postoperative analgesic/sedative drug administration may have contributed to the unexpectedly prolonged recovery times. Consistent with thiopental's pharmacokinetic profile, thiopental-treated patients had a longer intubation time than patients in the propofol group.
There are several potential limitations in interpreting these data. First, this was an open-label investigation, necessitated by the various modes of drug administration and the differing physical appearances of the anesthetic and analgesic drugs. Importantly, the personnel caring for study patients in the postoperative period were unaware of the anesthetic and analgesic drugs given intraoperatively and thus remained unbiased. Second, uniform criteria were not used to determine the extubation time, since these patients were cared for by different nurses and respiratory therapists. Foster et al.  have reported that rigorously applied protocols for weaning patients from mechanical ventilation could decrease the amount of time a patient was intubated after cardiac surgery. Third, fentanyl (unlike propofol or thiopental) was given by intermittent injection rather than as a continuous, titrated infusion. Nevertheless, several investigators have demonstrated that hemodynamic stability is not improved when fentanyl is administered as a continuous infusion (versus intermittent bolus doses) [17,29]. Finally, the fentanyl, 10 micro gram/kg, administered to all patients after the termination of CPB may have obscured the differences in recovery among the four treatment groups.
The use of the shorter-acting anesthetics enflurane and propofol, compared to large dose fentanyl and thiopental, facilitates recovery after coronary artery revascularization without compromising intraoperative hemodynamic stability. If our patients had been more aggressively weaned from ventilatory support, an even greater difference in recovery time might have been observed among the treatment groups. The pharmacokinetic and dynamic profile of propofol permits a rapid, transient increase in anesthetic depth without prolonging recovery. Although the anesthetic cost in the propofol group was approximately four times that in the other three groups, the overall cost (approximate $46) represents a small fraction of the total hospital cost. Future studies are needed to determine the optimal dosing regimen of these short-acting drugs for different noxious stimuli, in order to improve our ability to ensure hemodynamic stability without over- or underdosing patients during cardiac surgery.
The authors thank the cardiac surgeons, Drs. Richard M. Edie and John L. Mannion, and the ICU nurses at Thomas Jefferson University Hospital for their cooperation, Dr. Richard Epstein for his technical assistance, and Dr. James M. Bailey for reviewing the manuscript.
1. Shapiro BA, Lichtenthal PR. Inhalation-based anesthetic techniques are the key to early extubation of the cardiac surgical patient. J Cardiothorac Vasc Anesth 1993;7:135-6.
2. Higgins TL. Pro: early endotracheal extubation is preferable to late extubation in patients following coronary artery surgery. J Cardiothorac Vasc Anesth 1992;6:488-93.
3. Chong JL, Grebenik C, Sinclair M, et al. The effect of a cardiac surgical recovery area on the timing of extubation. J Cardiothorac Vasc Anesth 1993;7:137-41.
4. Gal TJ. Effects of endotracheal intubation on normal cough performance. Anesthesiology 1980;52:324-9.
5. Foster GH, Conway WA, Pamulkov N, et al. Early extubation after coronary artery bypass: brief report. Crit Care Med 1984;12:994-6.
6. Gall SA, Olsen CO, Reves JG, et al. Beneficial effects of endotracheal extubation on ventricular performance. J Thorac Cardiovasc Surg 1988;95:819-27.
7. Lichtenthal PR, Wade LD, Niemyski PR, Shapiro BA. Respiratory management after cardiac surgery with inhalation anesthesia. Crit Care Med 1983;11:603-5.
8. Sebel PS, Lowdon JD. Propofol: a new intravenous anesthetic. Anesthesiology 1989;71:260-77.
9. Russell GN, Wright EL, Fox MA, et al. Propofol-fentanyl anaesthesia for coronary artery surgery and cardiopulmonary bypass. Anaesthesia 1989;44:205-8.
10. Boer F, Ross P, Bovill JG, et al. Effect of propofol on peripheral vascular resistance during cardiopulmonary bypass. Br J Anaesth 1990;65:184-9.
11. Muzi M, Berens RA, Kampine JP, Ebert TJ. Venodilation contributes to propofol-mediated hypotension in humans. Anesth Analg 1992;74:877-83.
12. Goodchild CS, Serrao JM. Cardiovascular effect of propofol in the anaesthetized dog. Br J Anaesth 1989;63:87-92.
13. Samuelson PN, Reves JG, Kirklin JK, et al. Comparison of sufentanil and enflurane-nitrous oxide anesthesia for myocardial revascularization. Anesth Analg 1986;65:217-26.
14. Wynands JE, Townsend GE, Wong P. Blood pressure response and plasma fentanyl concentrations during high- and very high-dose fentanyl anesthesia for coronary artery surgery. Anesth Analg 1983;62:661-5.
15. Seltzer JL, Gerson JI, Allen FB. Comparison of the cardiovascular effects of bolus vs incremental administration of thiopentone. Br J Anaesth 1980;52:527-30.
16. Reiz S, Balfors E, Friedman A, et al. Effects of thiopentone on cardiac performance, coronary hemodynamics, and myocardial oxygen consumption in chronic ischemic heart disease. Acta Anaesthesiol Scand 1981;25:103-10.
17. Philbin DM, Rosow CE, Schneider RC, et al. Fentanyl and sufentanil anesthesia revisited: how much is enough? Anesthesiology 1990;73:5-11.
18. Waller JL, Hug CC, Nagle DM, Craver JM. Hemodynamic changes during fentanyl-oxygen anesthesia for aortocoronary bypass operation. Anesthesiology 1981;55:212-7.
19. Reves JG, Samuelson PN, Lell WA, et al. Myocardial damage in coronary bypass surgical patients anesthetized with two anesthetic techniques: a random comparison of halothane and enflurane. Can Anaesth Soc J 1980;27:238-46.
20. Nussmeier NA, Arlund C, Slogoff S. Neuropsychiatric complications after cardiopulmonary bypass: cerebral protection by a barbiturate. Anesthesiology 1986;64:165-70.
21. Slogoff S, Keats AS. Randomized trial of primary anesthetic agents on outcome of coronary bypass operations. Anesthesiology 1989;70:179-88.
22. Kaplan JA, Guffin AV, Mikula S, et al. Comparative hemodynamic effects of propofol and thiamylal sodium during anesthetic induction for myocardial revascularization. J Thorac Cardiovasc Anesth 1988;2:297-302.
23. Vermeyen KM, Erpels FA, Janssen LA, et al. Propofol-fentanyl anesthesia for coronary bypass surgery in patients with good left ventricular function. Br J Anaesth 1987;50:1115-20.
24. Hall RI, Murphy JT, Moffitt EA, et al. A comparison of the myocardial metabolic and hemodynamic changes produced by propofol-sufentanil and enflurane-sufentanil anaesthesia for patients having coronary artery bypass graft surgery. Can J Anaesth 1991;38:996-1004.
25. Mora CT, Body S, Bellows W, et al. Propofol and hemodynamic stability: a comparison with sufentanil anesthesia in patients undergoing coronary revascularization. Anesthesiology 1994;81:A139.
26. Quasha AL, Loeber W, Feeley TW, et al. Postoperative respiratory care: a controlled trial of early and late extubation following coronary-artery bypass grafting. Anesthesiology 1980;52:135-41.
27. Sanford TJ, Smith NT, Dec-Siver H, Harrison WK. A comparison of morphine, fentanyl, and sufentanil anesthesia for cardiac surgery: induction, emergence, and extubation. Anesth Analg 1986;65:259-66.
28. de Lange S, Boscoe MJ, Stanley TH, et al. Comparison of sufentanil-O (2
) and fentanyl-O2
for coronary artery surgery. Anesthesiology 1982;56:112-8.
29. Wynands JE, Wong P, Towns GE, et al. Narcotic requirements for intravenous anesthesia. Anesth Analg 1984;63:101-5.
30. Tuman KJ, McCarthy RJ, Spiess BD, Maschke SP. Does choice of anesthetic agent significantly affect outcome after coronary artery surgery? Anesthesiology 1989;70:189-98.
31. Howie MB, McSweeney TD, Lingam RP, et al. A comparison of fentanyl-O (2
) and sufentanil-O2
for cardiac anesthesia. Anesth Analg 1985;64:877-87.
32. Tuman KJ, Keane DM, Silins AI, et al. Effect of high-dose fentanyl on fluid and vasopressor requirements after cardiac surgery. Anesth Analg 1988;67:S236.
33. Howie MB, Smith DF, Reilley TE, et al. Postoperative course after sufentanil or fentanyl anesthesia for coronary artery surgery. J Cardiothorac Vasc Anesth 1991;5:485-9.
34. Murkin JM, Moldenhauer CC, Griesemar RW, et al. Fentanyl vs sufentanil: comparison of hemodynamic and catecholamine responses during coronary artery (CABG) surgery. Anesthesiology 1984;61:A375.