Secondary Logo

Journal Logo

Postoperative pain management in orthopaedic patients: no differences in pain score, but improved stress control by epidural anaesthesia

Adams, H. A.*; Saatweber, P.; Schmitz, C. S.; Hecker, H.

European Journal of Anaesthesiology: September 2002 - Volume 19 - Issue 9 - p 658-665
Original Article
Free

Background and objective: To investigate the interactions of postoperative pain and endocrine stress response, three groups of 21 patients each with total knee arthroplasty were compared in a randomized, prospective design. For postoperative pain management, a three-in-one block, an epidural catheter analgesia or an intravenous patient-controlled analgesia was used.

Methods: After standardized balanced anaesthesia, the pain intensity was measured by a visual analogue scale (VAS). For detection of epinephrine, norepinephrine, antidiuretic hormone, adrenocorticotropic hormone and cortisol in the plasma, blood samples were taken at six time points before and up to 180 min after the start of pain therapy. In addition, systolic arterial pressure, heart rate, partial arterial oxygen saturation, nausea, vomiting and satisfaction of the patients were recorded.

Results: Within 15 min after the start of pain therapy, VAS in all groups was similarly reduced from >40 mm to a range <10 mm (P < 0.001). Initially, all endocrine stress variables exceeded the normal range. Epidural anaesthesia led to a significant decrease of epinephrine and norepinephrine concentrations, while an increase was observed in the group with patient-controlled analgesia, and the decrease in patients with the three-in-one block was less than in patients receiving epidural anaesthesia (P = 0.001). Differences in antidiuretic hormone, adrenocorticotropic hormone and cortisol were less pronounced. Systolic arterial pressure decreased significantly in all groups, particularly in patients with epidural anaesthesia. Partial arterial oxygen saturation and the incidence of nausea and vomiting were comparable. All patients were satisfied with the methods used.

Conclusions: All methods of pain management led to sufficient analgesia, but they were not accompanied by an adequate reduction in endocrine stress response. Thus, postoperative pain is only a secondary stressor and sufficient analgesia with subjective well-being does not prove a stress-free state. With regard to the reduction of sympathoadrenergic stress response, epidural anaesthesia is superior to the three-in-one block and patient-controlled analgesia. Epidural anaesthesia is recommended particularly for high-risk patients with hypertension, coronary heart disease and diabetes mellitus. In these patients, the reduction of a 'hidden' endocrine stress response in addition to prevention of pain is of special interest.

Medizinische Hochschule Hannover, *Zentrum Anaesthesiologie andInstitut für Biometrie, Hannover;St Elisabeth-Krankenhaus Neuwied, Klinik für Anaesthesie und Intensivmedizin;Abteilung Anaesthesie und Intensivmedizin, Marienkrankenhaus Trier-Ehrang, Germany

Correspondence to: H. A. Adams, Zentrum Anaesthesiologie, Medizinische Hochschule Hannover, Carl-Neuberg-Str. 1, D-30625 Hannover, Germany. E-mail: adams.ha@mh-hannover.de; Tel: +49 511 532 3495; Fax: +49 511 532 8033

Accepted for publication April 2001 EJA 531

Back to Top | Article Outline

Introduction

Adequate postoperative pain management is an essential part of perioperative anaesthesia care. Many centres have organized an acute pain service and offer 24 h coverage. The main methods are 'on-demand' analgesia, patient-controlled analgesia (PCA) and various methods of regional anaesthesia, which differ significantly in their costs [1]. For validation of therapeutic success, the visual analogue scale (VAS) is in widespread use. In contrast to costs and analgesic potency, the interactions of postoperative pain management and endocrine stress response are more difficult to evaluate. Study results are inconsistent, and a correlation between the degree of analgesia and the endocrine stress response [2] as well as to the contrary [3,4] have been demonstrated.

The aim of this prospective, randomized study was to examine subjective pain intensity, endocrine stress response, cardiovascular and clinical variables in a homogeneous patient population with different methods for postoperative pain management, using either a three-in-one block, epidural catheter analgesia (EDA) or PCA.

Back to Top | Article Outline

Methods

Patients and general management

After approval by the Ethics Committee and informed consent, three groups of 21 patients (ASA I-III) were examined undergoing unilateral total knee arthroplasty with balanced anaesthesia. Exclusion criteria were: age below 18 yr, endocrine disorders except diabetes mellitus, glucocorticoid therapy, use of etomidate up to 48 h before the beginning of the study, and allergy towards one of the test substances. Patients requesting another method than the randomly intended postoperative pain management received their requested method and were excluded from the study.

Premedication and anaesthesia were standardized in all groups. About 60 min before operation, patients received midazolam 0.1 mg kg−1 body weight orally. For induction of balanced anaesthesia and orotracheal intubation, 2.5 μg kg−1 fentanyl, 5 mg kg−1 thiopental and 0.1 mg kg−1 vecuronium were injected intravenously (i.v.), and fentanyl was supplemented up to 5 μg kg−1 before the skin incision. For maintenance of anaesthesia and controlled ventilation of the lungs, nitrous oxide in oxygen in a ratio of 2:1 was used together with enflurane up to 1 MAC according to clinical requirements. If necessary, repeated doses of vecuronium 0.025 mg kg−1 were administered.

Back to Top | Article Outline

Postoperative pain management

The assigned postoperative pain management scheme (three-in-one, EDA or PCA) started immediately with the first expression of pain in the postanaesthesia care unit. The three-in-one blocks were performed with bupivacaine 40 mL 0.375% solution after detection of the nerve using a nerve stimulator. The EDA catheter was inserted immediately before operation and tested with bupivacaine 3 mL 0.25% solution. After surgery, bupivacaine 0.375% (1 mL per 10 cm body height, maximum 15 mL) was administered. Patients in the PCA group initially received repetitive small i.v. boluses injections of pirinitramide (piritramide) - up to 0.1-0.2 mg kg−1 - until a clear decrease of pain was reached. In addition, diclofenac 100 mg was administered rectally. After this initial treatment, patients could request single doses of pirinitramide - about 0.025 mg kg−1 (2 mg at 80 kg) - from a PCA pump with a 'lock-out' time of 10 min.

Back to Top | Article Outline

Evaluation

Variables. Variables under investigation were biometric data; pre-existing diseases and medical treatment; the dosage of bupivacaine, pirinitramide and diclofenac; the assessment of pain intensity by the patients with VAS of 0-100 mm; epinephrine, norepinephrine, antidiuretic hormone (ADH), adrenocorticotropic hormone (ACTH) and cortisol in the plasma (endocrine stress variables); systolic arterial pressure (SAP, by oscillometric measurement), heart rate (HR, continuous electrocardiogram) and partial arterial oxygen saturation (SPO2, pulse oximeter) with recording of SPO2 reductions of ≤90%; concentration of bupivacaine in the plasma (only for the three-in-one blocks); nausea and vomiting; and satisfaction with the method of postoperative pain management. Measurements were undertaken at the first expression of pain before pain therapy and at 15, 30, 60, 120 and 180 min after application of local anaesthetics or analgesics, respectively. The incidence of nausea and vomiting as well as satisfaction was recorded at interviews on the first postoperative morning.

Laboratory methods. Blood samples were taken from a central venous catheter, stored in iced water, separated and frozen at −25°C within 2 h. Epinephrine and norepinephrine concentrations were determined by high-pressure liquid chromatography with electrochemical detection (HPLC-ECD). Normal ranges for awake subjects were 30-85 pg mL−1 for epinephrine and 185-275 pg mL−1 for norepinephrine. The detection limit for plasma catecholamines was 3-4 pg mL−1, recovery 70-90%, and the intra-assay variation coefficient was 5.0% for norepinephrine and 7.2% for epinephrine, respectively. ADH, ACTH and cortisol were determined by radioimmunoassay. The normal range for ADH was 0-6.7 pg mL−1 with a sensitivity of about 1.25 pg mL−1, an intra-assay variation coefficient of 2.1-15.3% and a recovery of 91-109%. The normal range for the ACTH concentration was 9-52 pg mL−1, for sensitivity 1 pg mL−1, for the intra-assay variation coefficient about 3.0% and recovery of 89-109%. The normal range for cortisol concentration in the morning was 20-250 ng mL−1 with a detection limit of 2 ng mL−1 and an intra-assay variation coefficient of 3.0-5.1%. Bupivacaine concentrations were determined by HPLC and UV detection [5].

Statistical analysis. For descriptive statistics, group means, standard deviations (SD) and ranges were determined. Owing to non-normal distribution, values of endocrine variables were log-transformed; the results are given as retransformed geometric means and range. For evaluation of group variables, univariate analysis of variance (Ryan-Einot-Gabriel-Welsch test) or bivariate analysis of variance (Wilks-Lambda test) with measuring repetition on the factor 'time' were used. Differences at single measurement points were determined by the Ryan-Einot-Gabriel-Welsch test. α ≤ 0.05 was considered as significant.

Back to Top | Article Outline

Results

With regard to age, height and gender, the three groups were statistically comparable (Table 1). The body weight of patients receiving the three-in-one blocks was significantly lower than in the EDA group (P = 0.03). No differences in pre-existing diseases and medication became obvious.

Table 1

Table 1

All patients who received three-in-one blocks were treated with the intended dosage of bupivacaine, 40 mL 0.375% solution. The mean dosage in the EDA group was 14.3 ± 1.8 mL bupivacaine 0.375%. For initial reduction of pain, patients of the PCA group required 12.1 ± 3.7 mg pirinitramide with 100 mg diclofenac additionally.

Back to Top | Article Outline

Assessment of pain intensity

Assessment of pain intensity by the patients (Fig. 1 and Table 2) resulted in all groups in a significant and comparable decrease within 15 min after starting the scheme of postoperative pain management (P < 0.001). Pain levels decreased from >40 to <10 mm and remained in this range during the observation period of 180 min. No significant differences between group levels or for the time course were found.

Figure 1

Figure 1

Table 2

Table 2

Back to Top | Article Outline

Endocrine stress variables (Tables 2, 3).

Table 3

Table 3

Plasma catecholamine concentrations(Table 2). Initial plasma concentrations of epinephrine (Fig. 2) were comparable and above the normal range in all groups. A significant decrease in time course was found in all groups (P < 0.001). After the start of the scheme of postoperative pain management, concentrations declined immediately in the three-in-one patients and in the EDA patients. In the PCA patients, the decline was significantly delayed (P < 0.001). The group mean in the PCA group was higher than in the three-in-one group and the EDA patients (P = 0.004). At 15, 30, 60 and 120 min after the start of the scheme of postoperative pain management, concentrations in the PCA group were significantly higher than in three-in-one patients and in the EDA patients; after 180 min, the difference remained significant between the PCA and the three-in-one groups. Initial plasma concentrations of norepinephrine (Fig. 3) were comparable in all groups, means of about 600 pg mL−1 were above the normal range. After the start of the postoperative pain management, concentrations in the three-in-one group decreased slightly. A significant decrease occurred in the EDA group immediately after application of the local anaesthetic; 180 min later, the initial value was not regained. In contrast, the levels for the PCA patients increased initially; at the end of the observation period, the initial values were almost regained. These differences in time course were significant between the groups (P = 0.001); as well as the differences between the EDA and the PCA groups 30 and 60 min after the start of postoperative pain management (P < 0.05). After 120 min, the concentrations were comparable again.

Figure 2

Figure 2

Figure 3

Figure 3

ADH, ACTH, cortisol(Table 3). In all groups, initial plasma concentrations of ADH were comparatively above the normal range. Values declined significantly thereafter (P < 0.001). Fifteen minutes after the start of postoperative pain management, concentrations in the EDA group were higher than in the three-in-one patients; after 120 and 180 min, concentrations in the EDA patients were higher than in both other groups (P = 0.001). In addition, the initial concentrations of ACTH and cortisol were comparatively above the normal range in all groups. Over time, the concentrations of ACTH decreased significantly in all groups (P < 0.001) and approached the normal range or reached it in the EDA group, respectively. No differences between the groups were found. Concentrations of cortisol showed significant changes over time within the groups (P = 0.007), but altogether the concentrations remained constantly above the normal range, and no group differences occurred.

Back to Top | Article Outline

Circulatory variables (Table 4)

Table 4

Table 4

The systolic arterial pressure significantly decreased in all groups over time (P = 0.001). This decrease was significantly pronounced in the EDA patients (P < 0.001), where values after 15, 30, 60 and 120 min were lower than for the three-in-one and the PCA patients. For HR, only minor changes occurred. In the PCA group, values were significantly higher 30 and 60 min after the start of postoperative pain management. Means of SPO2 were comparable at all measurements; and no fall ≤90% was observed.

Back to Top | Article Outline

Further variables

Nausea occurred in six patients with the three-in-one block, in 12 patients with the EDA and in eight patients with the PCA system; vomiting occurred in three patients with the three-in-one block, in five patients with the EDA method and in six patients with the PCA system. These differences did not reach significance. All patients were satisfied with their individual scheme of postoperative pain management. The plasma concentrations of bupivacaine for the patients receiving the three-in-one block are given in Table 4.

Back to Top | Article Outline

Discussion

Selye [6,7] defined stress as a 'general adaptation syndrome' evolving in three consecutive stages and emphasized the role of cortisol. The dual stress concept of Henry [8] distinguished two basic reactions without imperative chronological succession [9]. The primary and active reaction is determined by epinephrine and norepinephrine according to Cannon and de la Paz's 'fight or flight' [10]. The release of catecholamines allows the fast mobilization of cardiovascular, respiratory and metabolic reserves of the threatened organism, while intestinal circulation and function are reduced. The simultaneous release of ADH protects the fluid reserves of the organism; beyond this, ADH serves as a reserve circulation hormone [11-13]. Interactions between the central sympathetic system, ADH and release of ACTH (and subsequently cortisol) are evident, but have not yet been sufficiently examined [14]. A connection to the secondary stage of stress is possible. This second stress stage is dominated by the activated pituitary-adrenal axis. According to Henry [8], this primary slower response indicates tolerance, adaptation and loss of control in the reaction to the stressor. With similar intention, Selye [15] characterized cortisol as a 'syntoxic hormone' for maintenance of life under stress, meaning a 'coexistence' between the organism and stressor without attacking the noxes.

During the intraoperative period, the aim of the different methods of general or regional anaesthesia is the moderation of the endocrine stress response. Moderation means avoidance of an overwhelming stress response as well as total suppression of this principally physiological and necessary response [9]. After the operation, this role is taken over by the method of management of postoperative pain.

In this study, all patients scheduled for total knee arthroplasty received standardized premedication and balanced anaesthesia to establish comparable conditions for the assessment of postoperative pain management. Allocation to the procedure for postoperative pain management (three-in-one, EDA or PCA) followed a randomized, prospective design, and postoperative pain management started immediately with the first expression of pain. The groups were comparable; differences in body weight were insignificant for the methods of regional anaesthesia used and were compensated by weight-dependent doses in the PCA group.

The analgesic potency of the three methods was comparable. Within 15 min, VAS were reduced from >40 to <10 mm and were stabilized there for 180 min. This result led to resultant satisfaction scores in all groups of patients. VAS in all was significantly lower than in comparable studies in orthopaedic patients [16,17]. However, the limited observation of 180 min has to be considered [18].

In contrast to the equivalent efficacy of these methods regarding pain scores, differences in the effects on the endocrine stress response were found. All groups showed a considerable endocrine stress response before the start of postoperative pain management, and the concentrations of the stress hormones examined were far above normal ranges at this time. The intraoperative stress response was not investigated in this study. Nevertheless and corresponding to other studies [9], an acceptable intraoperative stress reduction is assumed, and the significant increases in stress response are assigned to the recovery phase from general anaesthesia.

Under these conditions, EDA led to a fast and marked reduction especially of the sympathoadrenergic stress response. Significant analgesia together with segment-dependent sympatholysis induced a considerable decrease in norepinephrine concentrations and was accompanied by a reduction of epinephrine concentrations. The small initial increase in ADH concentrations can be interpreted as compensation for impaired sympathoadrenergic function [11-13]. A corresponding temporary reduction of systolic arterial pressure without significant hypotension occurred, but compensatory increases in HR were missing. In addition, ACTH concentrations were influenced by EDA. In contrast to the other methods, normal ranges were regained at the end of the observation period. There was no clear influence on the cortisol level, but concomitant reactions were to be expected after a certain delay. A reduction of endocrine stress response was also found in the three-in-one group. The extent was smaller than in EDA patients, especially with regard to norepinephrine concentrations. Despite comparable analgesia, the endocrine stress variables in the PCA group showed a consistently different course. The sympathoadrenergic stress response was attenuated only after a considerable delay, and an initial increase of catecholamines occurred. ADH remained on a clearly higher level than in the EDA and the three-in-one patients. However, the stress-lowering effect on the pituitary-adrenal axis corresponded to the other methods.

A distinct postoperative endocrine stress response, despite subjectively acceptable analgesia, is well documented in the literature [4,19-23]. Hjortsø and colleagues [3] even found no differences in endocrine reactions despite a differing quality of analgesia, but the methodology of catecholamine detection in the present study is outdated. Other authors [2,24] could at least verify a partial association of pain intensity and stress response; however, this may also be interpreted as a sign of insufficient analgesia.

No significant differences in side-effects were found. Arterial oxygen saturation was comparable in all groups, and the critical threshold of 90% was not reached. The incidence of nausea and vomiting was statistically comparable too, and the use of pirinitramide had no special negative consequences in the PCA group. The initial reduction in arterial pressure may have contributed to the surprisingly high incidence of nausea in the EDA group. Altogether the incidence of nausea and vomiting was high and requires further preventative efforts. Nevertheless, satisfaction expressed by the patients was excellent. The plasma concentrations of bupivacaine in patients with the three-in-one block, which have not yet been studied as extensively as concentrations during EDA, etc., stayed far below toxicologically relevant limits [25].

The study verifies that postoperative pain management with either a three-in-one block, EDA or PCA admirably eliminates postoperative pain in patients after total knee arthroplasty. In contrast to these comparable clinical results, the effects on endocrine stress response were distinctly different. In this respect, EDA was superior, especially in the reduction of sympatho-adrenergic stress response. Stress reduction in PCA patients was significantly poorer, while the three-in-one block was intermediate. Despite considerable reduction, the EDA group still showed a significant endocrine stress response. This suggests that pain is not the main stressor in the immediate postoperative period, and sufficient analgesia with subjective well-being does not indicate a stress-free state. In this early postoperative period, other stressors than pain predominate. These humoral and neural stressors have their origin in tissue trauma with release of mediators such as cytokines, which is not prevented by adequate peripheral or central sensory blockade. In this context, pain becomes a secondary stressor. Although a reduction of morbidity by adequate postoperative pain management is not evident [26], there are at least indications for the reduction of myocardial ischaemia in high-risk patients [27]. The superior stress-moderating effect of EDA shown here, therefore, would be suitable for patients with hypertension, coronary heart disease and diabetes mellitus, where the 'hidden' stress reduction beyond freedom of pain is of significance. Postoperative pain management with EDA may contribute to avoidance of sympathoadrenergic circulatory disorders, e.g. hypertension and myocardial ischaemia, but undesired hypotension has to be prevented as well. In addition, EDA counteracts stress-induced metabolic effects. Altogether, the stressor 'pain' is only one of numerous postoperative stressors, and a considerable stress response cannot be avoided, but only be reduced.

Back to Top | Article Outline

Conclusions

Despite excellent pain control by postoperative pain management, endocrine stress variables remained elevated after total knee arthroplasty, thus characterizing postoperative pain as a secondary stressor. When compared with a three-in-one block and PCA, epidural anaesthesia was superior in reduction of sympatho-adrenergic stress response.

Back to Top | Article Outline

References

1. Choinière M, Rittenhouse BE, Perreault S, et al. Efficacy and costs of patient-controlled analgesia versus regularly administered intramuscular opioid therapy. Anesthesiology 1998; 89: 1377-1388.
2. Rutberg H, Håkanson E, Anderberg B, Jorfeldt L, Martensson J, Schildt B. Effects of the extradural administration of morphine, or bupivacaine, on the endocrine response to upper abdominal surgery. Br J Anaesth 1984; 56: 233-238.
3. Hjortsø N-C, Christensen NJ, Andersen T, Kehlet, H. Effects of the extradural administration of local anaesthetic agents and morphine on the urinary excretion of cortisol, catecholamines and nitrogen following abdominal surgery. Br J Anaesth 1985; 57: 400-406.
4. Scheinin B, Scheinin M, Asantila R, Lindberg R, Viinamäki O. Sympatho-adrenal and pituitary hormone responses during and immediately after thoracic surgery - modulation by four different pain treatments. Acta Anaesthesiol Scand 1987; 31: 762-767.
5. Adams HA, Biscoping J, Ludolf K, Borgmann A, Bachmann-M B, Hempleman G. Die quantitative Analyse von Amid-Lokalanaesthetika mittels Hochdruck Flüssigkeits-Chromatographie und UV-Detektion (HPLC-UV). Reg Anaesth 1989; 12: 53-57.
6. Selye H. A syndrome produced by diverse nocuous agents. Nature 1936; 138: 32.
7. Selye H. The general adaptation syndrome and the diseases of adaptation. J Clin Endocrinol 1946; 6: 117-130.
8. Henry JP. Present concept of stress theory. In: Usdin E, Kvetnansky R, Kopin IJ, eds. Catecholamines and Stress: Recent Advances. New York, USA: Elsevier-North Holland, 1980: 557-571.
9. Adams HA, Hempelmann G. Die endokrine Streßreaktion in Anästhesie und Chirurgie-Ursprung und Bedeutung. Anästhesiol Intensivmed Notfallmed Schmerzther 1991; 26: 294-305.
10. Cannon WB, de la Paz D. Emotional stimulation of adrenal secretion. Am J Physiol 1911; 27: 64-70.
11. Möhring J, Glänzer K, Maciel JA Jr, et al. Greatly enhanced pressor response to antidiuretic hormone in patients with impaired cardiovascular reflexes due to idiopathic orthostatic hypotension. J Cardiovasc Pharmacol 1980; 2: 367-376.
12. Peters J, Schlaghecke R, Thouet H, Arndt JO. Endogenous vasopressin supports blood pressure and prevents severe hypotension during epidural anesthesia in conscious dogs. Anesthesiology 1990; 73: 694-702.
13. Carp H, Vadhera R, Jayaram A, Garvey D. Endogenous vasopressin and renin-angiotensin systems support blood pressure after epidural block in humans. Anesthesiology 1994; 80: 1000-1007.
14. Schürmeyer T, Wagner TOF. Grundlagen der Endokrinologie. In: Schedlowski M, Tewes U, eds. Psychoneuroimmunologie. Heidelberg, Germany: Spektrum Akademischer Verlag, 1996: 69-106.
15. Selye H. Hormone und Widerstandsfähigkeit. Münch Med Wschr 1970; 112: 1401-1407.
16. Rundshagen I, Kochs E, Standl T, Schnabel K, Schulte am Esch J. Subarachnoid and intravenous PCA versus bolus administration for postoperative pain relief in orthopaedic patients. Acta Anaesthesiol Scand 1998; 42: 1215-1221.
17. Singelyn FJ, Deyaert M, Joris D, Pendeville E, Gouverneur JM. Effects of intravenous patient-controlled analgesia with morphine, continuous epidural analgesia, and continuous three-in-one block on postoperative pain and knee rehabilitation after unilateral total knee arthroplasty. Anesth Analg 1998; 87: 88-92.
18. Silvasti M, Pitkänen M. Continuous epidural analgesia with bupivacaine-fentanyl versus patient-controlled analgesia with i.v. morphine for postoperative pain relief after knee ligament surgery. Acta Anaesthesiol Scand 2000; 44: 37-42.
19. Scott NB, Mogensen T, Bigler D, Kehlet H. Comparison of the effects of continuous intrapleural vs epidural administration of 0.5% bupivacaine on pain, metabolic response and pulmonary function following cholecystectomy. Acta Anaesthesiol Scand 1989; 33: 535-539.
20. Scott NB, Mogensen T, Bigler D, Lund C, Kehlet H. Continuous thoracic extradural 0.5% bupivacaine with or without morphine: effect on quality of blockade, lung function and the surgical stress response. Br J Anaesth 1989; 62: 253-257.
21. Salomäki TE, Leppäluoto J, Laitinen JO, Vuolteenaho O, Nuutinen LS. Epidural versus intravenous fentanyl for reducing hormonal, metabolic, and physiologic responses after thoracotomy. Anesthesiology 1993; 79: 672-679.
22. Eriksson-Mjöberg M, Kristiansson M, Carlström K, Eklund J, Gustafsson LL, Olund A. Preoperative infiltration of bupivacaine - effects on pain relief and trauma response (cortisol and interleukin-6). Acta Anaesthesiol Scand 1997; 41: 466-472.
23. Klasen JA, Opitz SA, Melzer C, Thiel A, Hempelmann G. Intraarticular, epidural, and intravenous analgesia after total knee arthroplasty. Acta Anaesthesiol Scand 1999; 43: 1021-1026.
24. Møller IW, Dinesen K, Søndergård S, Knigge U, Kehlet H. Effect of patient-controlled analgesia on plasma catecholamine, cortisol and glucose concentrations after cholecystectomy. Br J Anaesth 1988; 61: 160-164.
25. Fitzgibbons DC, Moore DC, Balfour RI. Convulsant blood levels of bupivacaine. Anesthesiology 1981; 55: 476.
26. Seeling W, Bothner U, Eifert B, et al. Patientenkontrollierte Analgesie versus Epiduralanalgesie mit Bupivacain oder Morphin nach großen abdominellen Eingriffen. Kein Unterschied in der postoperativen Morbidität. Anaesthesist 1991; 40: 614-623.
27. Beattie WS, Buckley DN, Forrest JB. Epidural morphine reduces the risk of postoperative myocardial ischaemia in patients with cardiac risk factors. Can J Anaesth 1993; 40: 532-541.
Keywords:

ANAESTHESIA, anaesthesia conduction, anaesthesia epidural, anaesthesia nerve block; CATECHOLAMINES, EPINEPHRINE, NOREPINEPHRINE; ENDOCRINE SYSTEM; PAIN, postoperative; STRESS

© 2002 European Academy of Anaesthesiology