Patients receiving general anaesthesia are usually transferred from the operating room to the postanaesthesia care unit (PACU). Fulfilling some simple discharge criteria offers the chance to transfer patients to a Phase II unit (outpatients) or to the ward . The availability of new anaesthetic agents has facilitated early recovery and has led to an earlier discharge of patients when standardized discharge criteria are applied. When these criteria have already been met in the operating room or during transfer to the PACU, it should be appropriate to consider bypassing the PACU and transferring the patient directly to the ward or step-down unit. This process is known as 'fast tracking' . This procedure offers some potential cost savings. Decreased occupation of the operating room for induction of anaesthesia or until tracheal extubation can improve total operating room capacity. Bypassing the PACU reduces the workload of the nurses working there. Finally, when anaesthetics with a low incidence of side-effects are used, less treatment need be applied postoperatively, e.g. to treat postoperative pain or nausea and vomiting. However, these potential cost savings are more difficult to calculate than the direct costs associated with drug acquisition .
In this prospective, randomized, patient- and observer-blinded study, the total costs of anaesthesia were evaluated in a German university department of anaesthesia. Direct costs, e.g. drug acquisition costs, as well as indirect costs were considered. The latter item was the sum resulting from the administration of additional drugs during anaesthesia or postoperatively to treat side-effects in the PACU and also the salaries for the anaesthetist and nurse during the time the patient was looked after before and after surgery. The period was mainly determined by the time the patient required to recover from anaesthesia until discharge to the ward or a step-down unit.
The study was performed to test the hypothesis that intravenous anaesthesia with propofol-remifentanil compared with a balanced anaesthesia technique using isoflurane-alfentanil improves the speed of recovery and minimizes postoperative side-effects. Cost saving achieved by decreased indirect costs could outweigh higher drug acquisition costs and thus could lead to an improved quality of recovery without increasing total costs.
The local Ethics Committee approved this patient- and observer-blinded, randomized study, which was planned and performed according to the recommendations of the CONSORT Statement (http://www.consort-statement.org). Patients undergoing nasal surgery (rhinoplasty, septorhinoplasty) gave informed written consent to participate in the study. They received oral premedication with dipotassium clorazepate 20 mg on the evening before and the morning of surgery. Antihypertensive medication taken before was continued throughout the study. After arrival in the operating room, patients were covered with a warmed blanket, but no active warming was performed. Using sealed envelopes, which had been prepared according to a computer-generated random table, the patients were allocated to one of the following two groups immediately before the start of anaesthesia:
• Intravenous anaesthesia (IVA group): induction 2 mg kg−1 and maintenance 4-8 mg kg−1 h−1 with propofol and a continuous infusion of remifentanil 10-30 μg kg−1 h−1
• Balanced anaesthesia (BA group): induction with propofol 2 mg kg−1 and alfentanil 20 μg kg−1, maintenance with isoflurane 0.4-1.0 vol.% and repetitive doses of alfentanil 0.5-1 mg up to a maximum dose of 75 μg kg−1.
In all patients, endotracheal intubation was facilitated with attacurium 0.5 mg kg−1. Patients' lungs were ventilated with oxygen-nitrous oxide (FiO2 = 0.35) with a fresh gas flow of 1 L min−1. Patients received a diclofenac suppository (100 mg) and metamizole (dipyrone, noramidopyrine) 1 g intravenously (i.v.) immediately after induction of anaesthesia. The attending ear, nose and throat (ENT) surgeon applied local anaesthesia before starting surgery. Mepivacaine 1% 5-10 mL with epinephrine 1:100 000 was used for this purpose.
All anaesthetists involved in the trial had considerable experience with both techniques. They were allowed to adjust drug dosages as clinically indicated to keep the heart rate and mean arterial pressure within a range of 20% of preoperative baseline values. If relative hypertension persisted - even with maximum allowed anaesthetic concentrations - urapidil 10 mg was used to control blood pressure and was repeated when necessary. At the end of surgery, the aim was to stop administration of anaesthetics to allow fast emergence from anaesthesia. After finishing the surgical dressings (defined as starting time), the times until tracheal extubation and until eligibility for leaving the operating theatre were recorded. These decisions were made by the attending anaesthesiologist according to the usual clinical criteria (e.g. sufficient breathing and/or co-operation). After leaving the operating room and until entering the recovery room, an observer, blinded to the type of anaesthesia given, judged at 1 min intervals whether or not the patient was fit for fast tracking directly to the ward, using a modified Aldrete recovery score. This score is used for postoperative evaluation and consists of seven items, each with a three-point scale (0-2). Thus, the maximum score is 14, indicating full recovery. A minimum of 12 points with no score lower than the medium ranking is required for a patient to be fast tracked. The criteria used in this score are level of consciousness, physical activity, haemodynamic stability, respiratory stability, oxygen saturation, and postoperative assessment of pain and emetic symptoms . Patients not judged eligible for fast tracking were transferred to the PACU. The same tool was used there (every 3 min) to judge when the patient was fit to be discharged to the ward. The times spent until meeting this discharge criterion were also recorded. In the PACU, pain was treated with small doses (2-3 mg) of pirinitramide (piritramide), an opioid analgesic, until pain was assessed as none or mild on a four-point verbal rating scale (none, mild, moderate, severe). PONV was treated with dolasetron 12.5 mg, a 5-HT3 receptor antagonist, if patients suffered moderate or severe nausea for more than 10 min or vomited twice or more. Patients with shivering not subsiding within 10 min received meperidine 25 mg i.v.
In the ward, the patients were visited 6 and 24 h postoperatively by a blinded observer (M.E.). The patients were interviewed for pain intensity, analgesic requirement and occurrence of PONV. During the second postoperative visit, the quality of recovery was measured using a validated German translation  of the quality of recovery (QoR) score . A nine-item postoperative questionnaire carefully developed according to psychological standards , it was validated in a large number of patients . By asking the patients, the question whether they would be happy to receive the same anaesthetic for a future operation the study aimed to assess the global satisfaction with anaesthesia.
Consumption of anaesthetics was recorded by the numbers of used ampoules (propofol 1% 200 mg, remifentanil 1 mg, alfentanil 1 mg). Isoflurane consumption was measured by weighing the vapour before and after anaesthesia on a high-precision scale (LP34000P®; Sartorius AG, Göttingen, Germany).
For all calculations of costs, those of a German university department of anaesthesia were chosen and the actual prices of labour costs  and drugs (for 2002) were taken into account. The following situations and assumptions were the basis of all calculations: a nurse (salary €0.84 min−1) is responsible for the preparation of all drugs and syringes. An anaesthetist (expense €1.57 min−1) induces general anaesthesia together with a nurse. This team is also present after the end of surgery until the patient leaves the operating room. The cost for each 1 min of occupation beginning with the completion of the surgical dressings is €3.36. After leaving the operating room, the patient is either discharged to the ward or, in case of fast tracking not being achievable, transferred to the PACU. In our hospital, the anaesthetist is not available while transferring the patient to the PACU. A nurse is present there and is responsible for three patients on average.
Power estimation and statistic analysis
The underlying standard deviation of times to achieve fast-tracking eligibility was based on an earlier observation of 2 weeks. It had been found to be 15 min. Thus, 60 patients per group offered a 95% chance of detecting a clinically relevant difference between the two groups of 10 min with a type I error of 5%.
The time until fast-tracking eligibility was analysed using the non-parametric Kaplan-Meier method with Mantel-Cox's log-rank test. The QoR score was compared using the U-test. Total costs were calculated considering local conditions (German university hospital) and prices, which were analysed descriptively.
Five of 120 patients (4%) initially randomized were excluded from the analysis. Further details are shown in Figure 1. No patient was excluded because of delayed emergence from anaesthesia or postoperative complications. Thus, no additional intention to treat analysis was performed. Biometric data, the duration and type of surgery were not different between the two groups. Since PONV could have contributed to a delayed fast-track eligibility, the risks for the patients of both groups were calculated using a validated risk score for prediction of PONV [9,10]. Using this approach, a comparable baseline risk for both groups (35 versus 39%; Table 1) could be demonstrated.
The times from end of surgery to eligibility for fast tracking were shorter in the IVA group (P = 0.0082; Mantel-Cox log-rank-test). This result is presented as a cumulative survival plot in Figure 2. Several steps of the emergence from anaesthesia (e.g. at the end of surgery until extubation, the time from extubation to leaving the operating room and, finally, the occurrence of side-effects precluding fast tracking) contributing to this main end point are listed in Table 2. A higher incidence of patients eligible for fast tracking following anaesthesia with propofol-remifentanil was associated with calculated cost savings of €4.20. This calculation was based on results from a survey on resources, staffing costs and the costs of running an operating room involving 267 German hospitals. For a university department, 1 min of idle time for the operating room is €3.36, the salary for the anaesthetist is €1.57 and that for a nurse is €0.84. All steps in this calculation are shown in Table 2.
The costs for anaesthetic drugs were higher in the IVA group (Table 3). Normalized values for the average duration of anaesthesia (135 min) were €16.46 (€0.122 min−1) for the isoflurane-alfentanil group but €34.63 (€0.257 min−1) for propofol-remifentanil. Additional costs were incurred by administration of drugs during anaesthesia (e.g. to treat hypertension) and postoperatively (e.g. to treat pain or PONV). These costs are also shown in Table 3. Predefined indications for the administration of each drug are described above. Additional costs caused by disposals and waste were low in both groups and were calculated at €0.93 for the BA group and €1.71 for the IVA group (Table 3). Thus, the costs of the two anaesthetic regimens (anaesthetic drugs, drug to treat side-effects, waste and disposals) were €20.90 for the BA group and €37.96 for the IVA group.
The incidence of side-effects are shown in Table 4. PONV was slightly less than predicted by Koivuranta's score and no major difference was noted between the two groups (BA group 22%; IVA group 18%). Six patients of the BA group and three in the IVA group received dolasetron as an antiemetic rescue medication. More patients after propofol-remifentanil suffered from pain postoperatively that required treatment. More than twice the patients in the IVA group shivered postoperatively and six compared with one patient in the BA group had to be treated with meperidine because of long-lasting symptoms.
In spite of a high incidence of postoperative side-effects, the quality of recovery was judged to be good by the patients using the QoR score. Scores were similar in the two groups (Table 2; P = 0.71).
When asked whether they would be very happy to receive the same anaesthetic for future surgery, there was no significance in the incidence of patients expressing complete satisfaction between the intravenous technique (93%) and balanced anaesthesia (87%).
It is neither new nor surprising that new anaesthetic drugs are associated with higher direct costs. A review of the literature reveals that several studies compared direct costs of total intravenous anaesthesia (TIVA) using remifentanil with costs associated with balanced anaesthesia. For instance, Heidvall and colleagues showed that sevoflurane in oxygen/nitrous oxide and fentanyl was less expensive (US$15) than TIVA using propofol and either alfentanil (US$18) or remifentanil (US$20) . Since the authors did not observe differences in emergence and postoperative side-effects, they concluded that from a cost-minimization point of view, balanced anaesthesia was the technique of choice.
However, when comparing TIVA with a balanced anaesthesia technique, the simplified approach used in cost-minimization analysis may not be appropriate since most authors found a marked difference in postoperative outcomes. For example, in a multicentre comparison of the costs of anaesthesia, more PONV was observed in patients who received sevoflurane for maintenance. When anaesthesia was induced with sevoflurane, patients were less satisfied with this technique compared with TIVA using propofol . An accompanying editorial stressed the importance of including indirect costs into the calculation of total costs. This is of particular relevance, since expenses for personnel are the most costly item in the surgical setting.
Model-based calculations have postulated that replacing traditional anaesthetic drugs (e.g. thiopental and isoflurane) with newer anaesthetic drugs (e.g. propofol) can help to reduce the personnel costs of anaesthesia [13,14]. However, there are only a limited numbers of studies providing these results in the clinical environment. A recent well-designed study compared the potential costs saving of a propofol-remifentanil-based TIVA technique with balanced anaesthesia using etomidate-isoflurane-fentanyl in geriatric patients undergoing cataract surgery . Similar to our results, direct costs were higher with propofol and remifentanil. Relying on comparable data for personnel costs (€1.31 min−1 for an anaesthetist, €0.89 for an anaesthesia nurse) significantly decreased emergence times from anaesthesia in the geriatric patients resulted in total cost savings of €12.30 per case. Furthermore, significantly more patients in the TIVA group wished to receive the same anaesthetic again (93%) compared with patients anaesthetized with a balanced anaesthesia (65%).
In this trial, administration of anaesthesia was not stopped until full completion of surgery. However, in our trial, the participating anaesthetists were encouraged to adjust anaesthesia at the end of surgery to allow quick recovery in both groups, reflecting clinical routine. Furthermore, geriatric patients (mean age 76-77 yr) were studied in Epple and colleagues'  trial undergoing extracapsular cataract extraction, a type of surgery known for its low intensity of postoperative pain. In these patients, administration of an ultrashort-acting opioid-like remifentanil may be of particular advantage. However, in surgery followed by moderate or severe postoperative pain, remifentanil as a sole opioid without additional analgesics is associated with acute postoperative pain.
Interestingly, in the present study, the major difference in emergence from anaesthesia was not observed in extubation times or the time interval required to leave the operating room. However, times to discharge were more predictable in the IVA group, e.g. 95% of the patients were ready to be discharged to the ward within 30 min after completion of surgery, whereas twice the time was necessary for the patients receiving balanced anaesthesia. This might be an additional advantage for the IVA group when the last surgery of the day is performed and the PACU can be closed earlier in the afternoon. The present clinical trial included the direct cost associated with two different anaesthetic regimens, and also considered the indirect cost arising from the need to treat side-effects and also those costs caused by higher demands for staffing.
For economic calculations, the view of a university department of anaesthesiology was used. No attempts were made to include costs arising from other components of the perioperative care, e.g. nursing efforts following fast tracking the patient back to the ward. However, the incidence of postoperative side-effects on the ward did not differ between the two groups, indicating that it is unlikely that there were major differences with respect to nursing interventions.
Overall, there were only small differences between the two groups with respect to several minor items, e.g. preparation of anaesthesia equipment, induction of anaesthesia, additional drugs used during or after anaesthesia, or costs for material and disposal. However, two major items are responsible for economic advantages of the balanced anaesthesia group: (1) direct costs associated with the acquisition of anaesthetic drugs (€37.96 versus €20.90); and (2) personnel costs associated with a higher proportion of patients not ready to bypass the PACU (€51.02 versus €57.06).
The resulting higher expense of €11.00 for the total anaesthesia technique must be considered against some intangible factors such as the overall quality of recovery or patient satisfaction. However, there are only small non-significant differences with respect to these issues. A difference of only 6% (87 versus 93%) of the patients expressing high satisfaction with their anaesthesia would mean a number needed to treat of 17 for one additional patient who was completely satisfied with general anaesthesia. In monetary terms, this would mean extra expenses of €184.00 for this patient. The same applies for the quality of recovery as measured by the QoR score.
In conclusion, faster recovery and higher fast-tracking eligibility has been demonstrated, as determined by a modified Aldrete score in patients anaesthetized with propofol-remifentanil compared with those receiving traditional balanced anaesthesia (isoflurane-alfentanil). More patients of the IVA group were ready to be discharged to the ward immediately after leaving the operating room (59 versus 80%). Early discharge to the ward was associated with some cost savings. However, these did not outweigh the higher acquisition costs of the newer anaesthetics. Under the assumptions made in this trial, intravenous anaesthesia with propofol and remifentanil offers no monetary advantage over a traditional balanced anaesthesia technique using isoflurane and alfentanil.
1. White PF, Song D. New criteria for fast-tracking after outpatient anesthesia: a comparison with the modified Aldrete's scoring system. Anesth Analg
2. Song D, Chung F. Fast-tracking in ambulatory anesthesia. Can J Anaesth
3. Crozier TA, Kettler D. Cost effectiveness of general anaesthesia: inhalation vs. i.v. Br J Anaesth
4. Eberhatt LHJ, Greiner S, Geldner G, Wulf H. Patienten-beurteilung der postoperativen Erholung. Eine Validierung des QoR-Scores an 577 Patienten. Anaesthesist
5. Myles PS, Hunt JO, Nightingale CE, et al.
Development and psychometric testing of a quality of recovery score after general anesthesia and surgery in adults. Anesth Analg
6. Fung D, Cohen MM. Measuring patient satisfaction with anesthesia care: a review of current methodology. Anesth Analg
7. Myles PS, Reeves MD, Anderson H, Weeks AM. Measurement of quality of recovery in 5672 patients after anaesthesia and surgery. Anaesth Intensive Care
8. Bach A, Bauer M, Geldner G, et al.
Erfassung der IST-Kosten der Anästhesieabteilungen in Deutschland. Anästh Intensivmed
9. Koivuranta M, Läärä E, Snåre L, Alahuhta S. A survey of postoperative nausea and vomiting. Anaesthesia
10. Eberhart LHJ, Högel J, Seeling W, Staack AM, Geldner G, Georgieff M. Evaluation of three risk scores to predict postoperative nausea and vomiting. Acta Anaesthesiol Scand
11. Heidvall M, Hein A, Davidson S, Jakobsson J. Cost comparison between three different general anaesthetic techniques for elective arthroscopy of the knee. Acta Anaesthesiol Scand
12. Smith I, Terhoeve PA, Hennart D, et al.
A multicentre comparison of the costs of anaesthesia with sevoflurane or propofol. Br J Anaesth
13. Marais ML, Maber MW, Wetchler BV, Korttila K, Apfelbaum JL. Reduced demands on recovery room resources with propofol (Diprivan) compared to thiopental-isoflurane. Anesthesiol Rev
14. Suver J, Arikian SR, Doyle JJ, Sweeney SW, Hagan M. Use of anesthesia selection in controlling surgery costs in an HMO hospital. Clin Ther
15. Epple J, Kubitz J, Schmidt H, et al.
Comparative analysis of costs of total intravenous anaesthesia with propofol and remifentanil vs. balanced anaesthesia with isoflurane and fentanyl. Eur J Anaesthesiol