Secondary Logo

Journal Logo

Induced hypotension for tympanoplasty: a comparison of desflurane, isoflurane and sevoflurane

Dal, D.; Çelker, V.ı; Özer, E.; Başgül, E.; Salman, M. A.; Aypar, Ü.

European Journal of Anaesthesiology: November 2004 - Volume 21 - Issue 11 - p 902-906
Original Article

Background and objectives: This prospective, randomized, double-blinded study was designed to compare the effects of desflurane, isoflurane and sevoflurane when combined with remifentanil for induced hypotension on surgical conditions and operative field during tympanoplasty.

Methods: Sixty patients undergoing tympanoplasty were enrolled in the study. The patients were randomized into three groups of 20 each to receive the inhalation anaesthetics desflurane, isoflurane or sevoflurane. Propofol 2 mg kg−1 was administered for induction of anaesthesia in all groups. All patients received a continuous infusion of remifentanil which was titrated between 0.2 and 0.5 μg kg−1 min−1 to achieve a mean blood pressure (BP) of 60-70 mmHg. Nitroglycerine was infused if this BP could not be achieved. Arterial pressures were recorded continuously throughout the operation. Surgical conditions were assessed every 20 min by the blinded surgeon using a six-point category scale (0-5).

Results: One patient in the desflurane group and two patients in isoflurane group required nitroglycerine to maintain desired mean BP. Sustained controlled hypotension was sufficient in all of the groups throughout surgery. Category scale scores were ≤3 throughout the study, except one patient in the sevoflurane group who had a score of 4 at the 60th minute of the operation. No difference was found among groups when haemodynamic parameters and surgical category scale scores were compared. There were no postoperative respiratory and circulatory complications.

Conclusion: Desflurane, sevoflurane or isoflurane combined with remifentanil provided adequate induced hypotension and similar operating conditions and any of them could be safely and equally used in anaesthesia for tympanoplasty.

Hacettepe University, Faculty of Medicine, Department of Anesthesiology and Reanimation, Ankara, Turkey

Correspondence to: Didem Dal, Department of Anesthesiology and Reanimation, Faculty of Medicine, Hacettepe University, Sihhiye TR-06100, Ankara, Turkey. E-mail:; Tel: +90 312 305 1264; Fax: +90 312 310 9600

Accepted for publication August 2004 EJA 1839

Controlled hypotension is commonly used to achieve a bloodless operative field for tympanoplasty [1]. Vasodilators such as sodium nitroprusside, nicardipine or nitroglycerine, β-adrenergic antagonists such as propranolo or esmolol, the α- and β-adrenergic antagonist labetolol, as well as high doses of potent inhaled anaesthetics such as halothane have been used [1-6]. Disadvantages have been reported including long post-anaesthetic recovery for halothane, resistance to vasodilators, tachyphylaxis and cyanide toxicity for nitroprusside, or possibility of myocardial depression for esmolol [1]. The role of remifentanil and propofol combination in optimizing the operative field has been demonstrated [7], but the combinations of volatile agents with remifentanil have not been. Desflurane has been used for ear nose throat surgery in combination with remifentanil due to its beneficial effects on recovery [8]. Bertrand and colleagues have demonstrated that isoflurane reduces bleeding compared to halothane in microsurgery of the middle ear [9]. Preckel and colleagues have suggested that isoflurane has a greater propensity to alter cochlear autoregulation when compared to propofol [10]. Recently it has been demonstrated that sevoflurane has a hypotensive effect without modifying cochlear blood flow compared to propofol [11]. The efficacies of desflurane, isoflurane and sevoflurane have not been compared previously in ear nose and throat surgery with regards to bleeding and quality of the operative field. We hypothesized that these agents may have different effects on operative field bleeding due to possible differences in their vasoconstriction or vasodilator properties.

Accordingly, this prospective study was designed to compare the effects of remifentanil combined with desflurane, sevoflurane or isoflurane on the quality of the operative field and surgical conditions during tympanoplasty.

Back to Top | Article Outline


After obtaining Ethics Committee approval and written informed consent, 60 patients, American Society of Anaesthesiologists grades I and II (ASA I-II), aged from 17 to 50 yr undergoing tympanoplasty were included in this prospective randomized study. Patients were excluded from the trial if they had significant dysrhythmia, uncontrolled hypertension, uncontrolled diseases of the central nervous system or hypersensitivity to opioids.

All patients were admitted on the day before surgery and fasted for at least 8 h before surgery. A standard premedication with diazepam 10 mg orally was given 1 h prior to surgery. On arrival in the operating area, two cannulae were inserted at different sites on the same arm, one for the infusion of remifentanil and the other for administration of fluid and other drugs. A 22-G catheter was inserted into a radial artery for direct determination of arterial pressure and heart rate (HR), which were continuously recorded. First values before induction were defined as baseline values.

A crystalloid solution (5 mL kg−1) was administered at the start of the induction period and 100% oxygen was administered by mask for the first 3 min. Patients were assigned to three equal groups by computer generated random numbers. In all groups, propofol was used as the induction drug. Propofol 2 mg kg−1 was administered by slow intravenous (i.v.) injection (10 mg s−1) until loss of consciousness (defined as loss of response to verbal command) was attained. The remifentanil infusion at 0.5 μg kg−1 min−1 was initiated simultaneously. Vecuronium 0.1 mg kg−1 was administered 2 min after the start of the remifentanil infusion. Tracheal intubation was performed 3 min after neuromuscular blockade. Patients were mechanically ventilated which was adjusted to provide an end-tidal CO2 concentration of 30-35 mmHg and a SPO2 greater than 95% with 50% air in oxygen. General anaesthesia was maintained with one of the inhaled anaesthetics desflurane, isoflurane or sevoflurane according to the group the patient was assigned to. The end-tidal concentrations of the inhalation anaesthetics were standardized to one minimal alveolar concentration (MAC) corrected for the patient's age.

The patients in all of the groups received continuous remifentanil infusion of 0.2-0.5 μg kg−1 min−1 until the mean arterial pressure (MAP) was 60-70 mmHg. The infusion rate was then titrated to maintain the blood pressures (BP) within these limits. A group without remifentanil infusion was not included as control since it was considered unethical not to use any agent for induced hypotension, which is in our routine practice in tympanoplasty operations. Nitroglycerine was infused if these limits could not be achieved by the uppermost dose of remifentanil (0.5 μg kg−1 min−1). In all groups, signs of inadequate anaesthesia such as increases in arterial pressure more than the targeted BP or somatic responses such as movement, tearing or sweating were treated with additional increases in the remifentanil infusion rate. The remifentanil infusion rate was then decreased when the targeted BP was achieved.

The surgical condition for a bloodless operative field was rated every 20 min by the same attending surgeon who was blinded to the pharmacological treatments. A six-point scale was used: 0: no bleeding - a virtually bloodless field; 1: bleeding - so mild that it was not a surgical nuisance; 2: moderate bleeding - a nuisance but without interference with accurate dissection; 3: moderate bleeding - which moderately compromised surgical dissection; 4: bleeding - heavy but controllable, and which significantly interfered with surgical dissection; 5: massive uncontrollable bleeding [1]. The mean surgical field scores obtained throughout the operation were also calculated for each patient and mean scores were compared among groups. Ideal category scale values for optimal surgical conditions are defined as ≤2 which do not interfere with surgical dissection. No local vasoconstrictor was used throughout the surgery in order to provide control of bleeding. Baseline scoring was obtained at the skin incision.

Ten minutes before the anticipated end of surgery, the infusion rate of remifentanil was decreased by 50%. At the end of surgery, any residual neuromuscular block was antagonized with neostigmine and atropine.

Side-effects such as nausea, retching, vomiting, postoperative desaturation (SPO2 < 90% with room air) and additional hypotensive agent (nitroglycerine) requirements, were recorded in each group. Analgesic requirements were also recorded as number of doses. Emergence time was assessed by the time of eye opening of the patient as a response to a verbal command.

Back to Top | Article Outline

Statistical analysis

Distribution of data was evaluated by the Kolmogorov-Smirnov test. Statistical analysis of the data was undertaken by parametric or non-parametric methods as required. One-way analysis of variance (ANOVA) was used for comparison of age, weight, height, body mass index (BMI), weight adjusted infusion dose of remifentanil used, duration of surgery and anaesthesia among groups. The Kruskal-Wallis test was used for comparison of gender, nitroglycerine requirement, category scores and eye opening time among groups. When a significant difference was noted, a Bonferroni's test was performed for post hoc comparisons in data with normal distribution. When a significant difference was noted in data without normal distribution, a U-test was performed for post hoc comparisons with correction of significance values. Repeated measures analysis of variance was used for analysing haemodynamic parameters and operative field scores among groups and within time. Cross-tabulations of category scores were obtained and the ratios of scores ≥3 were compared among groups by using Pearson χ2-tests. SPSS version 10.0 statistical software (SPSS Inc., Chicago, IL, USA) was used for data analysis. Data were expressed as mean ± standard deviation (SD) and P < 0.05 was considered significant.

A prospective power analysis was performed for estimating the sample size based on the assumption that mean surgical field category scores would be 1 with an SD of 0.5. The analysis revealed that a sample size of 16 patients in each of three groups offered a chance of 81.4% to detect a 50% difference in mean category scores with an α error of 0.05. Since our data demonstrated a difference in mean and SD than the assumption we used in the a priori power analysis, a post hoc power analysis was made based on actual data. The second analysis revealed that our study had 84.7% power to detect a 50% difference in mean category scores with a mean score of 0.667 and SD of 0.3598 with an α error of 0.05.

Back to Top | Article Outline


There were 60 adult patients recruited, 20 male and 40 female. All the patients were similar with respect to age, sex, duration of anaesthesia and surgery (Table 1). The weights of patients in the sevoflurane group were found to be lower than the weights of patients in the isoflurane group (P = 0.04). Isoflurane patients weighed statistically more than patients in the sevoflurane and desflurane groups. Isoflurane patients were statistically taller than patients in sevoflurane and desflurane groups. BMI were similar in all groups.

Table 1

Table 1

Total dose of remifentanil administered was similar in all groups. One patient in the desflurane group and two patients in the isoflurane group received nitroglycerine infusions (Table 2). Mean doses of nitroglycerine were not statistically different among groups. Mean time to eye opening was also similar in all groups.

Table 2

Table 2

There were significant decreases in systolic and mean BP in all groups but no difference was observed among groups (P > 0.05) (Fig. 1). The HR also decreased with time in all groups. No statistically significant difference was observed in the HR among groups (P > 0.05).

Figure 1

Figure 1

The mean category scale scores of surgical condition for bloodless operative field decreased with time in all groups but they were not statistically different among groups except the scores obtained at the 40th minute of the surgery (Fig. 1). The mean bleeding scores at the 40th minute of the operation were significantly lower in the sevoflurane group compared to both isoflurane and desflurane groups. Our scores were ≤3 throughout the study except one patient in sevoflurane group that had a score of 4 only in 60th minute of the operation. The surgeons used the bleeding score of 3, five times in the desflurane group, three times in the isoflurane group and three times in the sevoflurane group. The median and range of the category scale scores of desflurane, isoflurane and sevoflurane groups were 1 (0-3), 1 (0-3) and 1 (0-4) respectively. No statistically important significance was found among groups (P = 0.393). Five scores out of 128 in desflurane group, 3 scores out of 137 in isoflurane group and 7 scores out of 130 scores in the sevoflurane group were ≥3, thus demonstrating interference with the surgical conditions.

The category scale scores were not found to be correlated with systolic, diastolic BP and HR. Category scores were weakly correlated with MAP only at the 20th minute of the surgery (P = 0.040, r = 0.266).

There were no postoperative respiratory and circulatory complications. No patient was eliminated from the study due to bradycardia or oxygen desaturation. Only one patient in the sevoflurane group reported emesis after operation. Postoperative analgesic requirements were not different among the groups. There were no other postoperative complications such as nausea, retching or vomiting in any group. None of the patients showed difference in the time of eye opening and all of the patients were discharged at the first postoperative day.

Back to Top | Article Outline


The main finding of the present study was that desflurane, isoflurane and sevoflurane had similar effects when used in combination with remifentanil hydrochloride in providing a bloodless surgical field. All three agents demonstrated similar haemodynamic effects during tympanoplasty. Remifentanil was effective in reaching a MAP of 60-70 mmHg and ensured good surgical conditions without delaying emergence when used together with the volatile agents.

It has become recognized that haemostasis for middle ear surgery presents special problems for the anaesthetist since the magnification of the operating microscope makes even minimal bleeding appear major. Many techniques have been devised to deal with this problem but none seems ideal. If inhaled anaesthetics are used to decrease BP, larger inspired concentrations are used than are required to provide surgical anaesthesia which may result in delayed awakening. In addition, β-blockade may need to be used to control unwanted tachycardia, regardless of whether nitroprusside, isoflurane or sevoflurane is used for hypotension [12,13]. The hypotensive effects of different agents are provided by different mechanisms that may affect the microcirculation especially as autoregulation may lead to vasodilatation in order to maintain blood flow. The use of esmolol decreases cardiac output and sodium nitroprusside decreases vascular smooth muscle tone. Beaussier and colleagues reported that desflurane maintains better haemodynamic stability compared to isoflurane when used in combination with fentanyl during moderate hypotensive anaesthesia [14]. The study was performed in patients undergoing spinal surgery and targeted a systolic BP of 80-100 mmHg. In that study blood loss and surgical conditions rated as bleeding scores were similar with isoflurane and desflurane. In our study, we did not identify any statistical difference among the volatile agents compared which is in accordance with the Beaussier study. Our choice of targeted MAP was between 60 and 70 mmHg. This was moderate controlled hypotension. Mortality due to controlled hypotension and consequent ischaemic organ failure is 0.02-0.06% [15] and we accordingly avoided profound controlled hypotension (MAP = 50 mmHg).

The concentrations of inhaled anaesthetics and the infusion rate of remifentanil used in our study were satisfactory for maintenance of controlled hypotension in our patients and showed efficacy and safety. In our clinic, we routinely use remifentanil for providing hypotension during tympanoplasty in order to benefit both from its analgesic effect in the nitrous oxide-free anaesthesia technique and also from the middle ear blood flow decrease as reported by Degoute and colleagues [1,7]. Although we did not study the mechanisms by which anaesthetics influence the mucosal blood flow of the middle ear, it has been suggested that the baroreflex regulatory mechanisms of inhaled anaesthetics combined with the arteriolar and precapillary vasoconstrictor effects of remifentanil which are presumed to result from unopposed α-adrenergic effects during hypotension may reduce mucosal blood flow of the ear [7].

The most important aspect of tympanoplasty is the graft placement at the end of the operation, which takes place approximately between 60th and 100th minute of the operation that is the last 20-30 min of the operation. Oozing of blood at this stage obscures the surgeon's vision and hence may prevent the correct placement of the graft in connection with the drum remnant. Therefore, a bloodless operative field is fundamentally important for the survival of the tympanic membrane graft. All of our patients had category scale scores 0-2 during the graft placement. These scores indicate good surgical conditions. The same surgeon performed all the operations and did not complain of any surgical bleeding that could interfere with his work causing longer operating times.

We aimed to provide optimal surgical conditions and this might mask minute effects of volatile agents on bleeding. The scoring we used despite being rated by blinded surgeons is a subjective scoring system and therefore might hide small differences among groups. An objective parameter, such as measuring the middle ear blood flow, could improve our results. Albera and colleagues have stated that sevoflurane produces its hypotensive effect without decreasing middle ear blood flow [11]. Unfortunately, we did not have the facilities to measure this parameter in our study.

In conclusion, the present study showed that desflurane, sevoflurane and isoflurane combined with remifentanil provided adequate induced hypotension and that any of these agents may be equally and safely used in anaesthesia for tympanoplasty.

Back to Top | Article Outline


1. Degoute CS, Ray MJ, Manchon M, Dubreil C, Bansillon V. Remifentanil and controlled hypotension; comparison with nitroprusside or esmolol during tympanoplasty. Can J Anesth 2001; 48: 20-27.
2. Saarnivaara L, Brander P. Comparison of three hypotensive anaesthetic methods for middle ear microsurgery. Acta Anaesthesiol Scand 1984; 28: 435-442.
3. Muzzi DA, Black S, Losasso TJ, Cucchiara RF. Labetolol and esmolol in the control of hypertension after intracranial surgery. Anesth Analg 1990; 70: 68-71.
4. Saarnivaara L, Klemola UM, Lindgren L. Labetolol as a hypotensive agent for middle ear microsurgery. Acta Anaesthesiol Scand 1987; 31: 196-201.
5. Jacobi KE, Böhm BE, Rickauer AJ, Jacobi C, Hemmerling TM. Moderate controlled hypotension with sodium nitroprusside does not improve surgical conditions or decrease blood loss in endoscopic sinus surgery. J Clin Anesthesia 2000; 12: 202-207.
6. Boezaart PA. Comparison of sodium nitroprusside and esmolol induced controlled hypotension for functional endoscopic sinus surgery. Can J Anesth 1995; 42: 373-376.
7. Degoute CS, Ray MJ, Gueugniaud PY, Dubreil C. Remifentanil induces consistent and sustained controlled hypotension in children during middle ear surgery. Can J Anesth 2003; 50: 270-276.
8. Loop T, Priebe HJ. Recovery after anesthesia with remifentanil combined with propofol, desflurane, or sevoflurane for otorhinolaryngeal surgery. Anesth Analg 2000; 91: 123-129.
9. Bertrand D, Boivin G, Manel J, Laxenaire MC. Effects of isoflurane on bleeding in microsurgery of the middle ear. Ann Fr Anesth Reanim 1987; 6: 416-418.
10. Preckel MP, Ferber-Viart C, Leftheriotis G, et al. Autoregulation of human inner ear blood flow during middle ear surgery with propofol or isoflurane anesthesia during controlled hypotension. Anesth Analg 1998; 87: 1002-1008.
11. Albera R, Ferrero V, Canale A, De Siena L, Pallavicino F, Poli L. Cochlear blood flow modifications induced by anaesthetic drugs in middle ear surgery: comparison between sevoflurane and propofol. Acta Otoloryngol 2003; 7: 812-816.
12. Hackmann T, Friesen M, Allen S, Precious DS. Clonidine facilitates controlled hypotension in adolescent children. Anesth Analg 2003; 96: 976-981.
13. Tobias JD. Sevoflurane for controlled hypotension during spinal surgery: preliminary experience in five adolescents. Paediatr Anaesth 1998; 8: 167-170.
14. Beaussier M, Paugam C, Deriaz H, et al. Haemodynamic stability during moderate hypotensive anaesthesia for spinal surgery. A comparison between desflurane and isoflurane. Acta Anaesthesiol Scand 2000; 44: 1154-1159.
15. Leigh JM. The history of controlled hypotension. Br J Anaesth 1975; 47: 745-749.

ANAESTHETICS INHALATION, desflurane, isoflurane, sevoflurane; OTOLOGIC SURGICAL PROCEDURES, tympanoplasty; HYPOTENSION, induced

© 2004 European Academy of Anaesthesiology