BJA Advance Access originally published online on March 10, 2006
British Journal of Anaesthesia 2006 96(5):633-639; doi:10.1093/bja/ael049
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Measuring the quality of continuous epidural block for abdominal surgery
University Department of Anaesthesia and Ninewells Hospital and Medical School Dundee DD1 9SY, UK
*Corresponding author. E-mail: g.a.mcleod{at}dundee.ac.uk
Accepted for publication January 30, 2006.
 |
Abstract |
|---|
 Top Abstract IntroductionMethodsResultsDiscussionReferences |
|---|
Background. In view of the wide variation in pain experience between patients, a clinical standard—the time from the end of surgery to the first experience of pain—was applied to 1359 consecutive patients in order to investigate whether the initial quality of epidural block has an effect on the overall quality of postoperative pain relief.
Methods. Clinical data were recorded in 58 118 out of 72 412 h in 1359 patients, and transferred to a database. Data collected included pain scores on a four-point verbal rating scale; nausea and vomiting; motor block; sedation scores; systolic blood pressure <100 and <90 mm Hg; ventilatory frequency <10 and <8 bpm; and hourly epidural infusion rate.
Results. As the time to first experience of pain increased from nil to >24 hours, the time from the first to last experience of pain shortened from 34 (19–50) h to 3 (1–12) h (p<0.001) and the proportion of patients receiving an epidural bolus decreased from 53 to 8% (p<0.001). Increases in the initial pain free time increased the proportion of patients with systolic BP<100 mmHg from 59 to 77%, (p<0.001) and increased the proportion of patients with respiratory rate <10 bpm from 13 to 26%, (p<0.001).
Conclusion. Extending pain relief for more than 12 h beyond the end of abdominal surgery significantly improves the overall quality of postoperative pain relief, but is associated with an increase in side-effects.
Keywords: anaesthetic techniques, epidural, thoracic; analgia, preventive; pain, postoperative
|
Introduction |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
Thoracic epidural block has the potential to provide postoperative pain relief of such quality that, after abdominal surgery, patients may be mobilized quickly, recovery accelerated and function restored.1 However, increasing evidence suggests that, even with thoracic epidural block, postoperative pain is still not being managed well. Two recent studies2 3 have shown that between 20 and 38% of patients experience moderate to severe pain on movement, contrasting markedly with the Audit Commission (UK) proposal that by the year 2002, less than 5% of patients should have suffered pain after surgery.3 Thus, clinical expectations are not being met, and a need exists to determine what makes thoracic epidural block work in everyday practice.
However, before measuring the quality of thoracic epidural block, a standard of care needs to be defined. A previous attempt to measure the quality of epidural pain services2 found that the experience of pain and pain relief during epidural analgesia is not as narrow as has been hitherto assumed, but is, in fact, very broad. Most patients are kept pain free, but for variable lengths of time; the remainder experience little pain relief and have frank episodes of continuous pain, even when bilateral sensory block has been identified.
Consideration of the ‘broad theory of pre-emptive analgesia’ or ‘preventive analgesia’4 helps define a standard for thoracic epidural block. It proposes that the emphasis of acute pain management is placed on providing intense, prolonged, uninterrupted pain relief such that, when the block wears off, pain and the biochemical changes in the spinal cord5 are less than they would have been if the block had been shorter, or had been interrupted. Because this requires patients to be pain free when awakening from surgery, and for as long as possible, the most appropriate measure of the quality of thoracic epidural block would appear to be the time from the end of surgery to the first experience of pain. Therefore, the aim of this study was to measure the time to first experience of pain in 1359 consecutive patients after abdominal surgery, and investigate its influence on the need for epidural bolus injection, epidural infusion rate and side-effects.
|
Methods |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
Data for all patients receiving combined epidural–general anaesthesia and postoperative epidural infusion after laparotomy in Ninewells Hospital between 1993 and 2001 were collected on the standard epidural pain service chart. Patient characteristics such as age, gender and ASA status, and technical details of the epidural, such as vertebral level of insertion, and composition of the intraoperative bolus and postoperative infusion, were entered by the anaesthetist responsible for epidural block. During the 9 yr period of assessment, the choice of epidural solution changed. Initially, in 1993, bupivacaine 0.6 mg ml–1 and fentanyl 4 µg ml–1 (n=225) was used, then changed to bupivacaine 0.9 mg ml–1 and fentanyl 2 µg ml–1 (n=77) in 1995, because of fears of respiratory depression. However, a perception of inadequate pain relief from the latter solution accelerated a change to bupivacaine 1 mg ml–1 and diamorphine 25 µg ml–1 (n=1057) in 1996.
Details recorded during epidural infusion
Recording of epidural data and vital signs took place within the recovery room every 15 min for the first postoperative hour, every 30 min in the following hour and hourly thereafter for the duration of the epidural. Hourly data (Table 1) included pain scores, blood pressure, ventilatory frequency, sedation score, motor block, nausea and vomiting, epidural infusion rate and dermatomal level of block measured using ice. Pain was measured as a verbal rating score (VRS), a four-point scale of ranked ordinal data (0–3) in which 0=no pain on movement; 1=no pain at rest, but mild pain on movement; 2=moderate pain on movement; and 3=severe pain on movement. Inability to record pain because the patient was sleeping was recorded as ‘S’. Both nausea/vomiting and sedation were recorded on a four-point VRS in which 0=nil; 1=mild; 2=moderate; 3=severe. Transfer from the recovery room to the high dependency unit (HDU) was at the recovery nurse's discretion, and dependent on the clinical condition of the patient.
|
Care in the HDU was based on the same strict bedside monitoring, nursing protocols and algorithms used in the recovery room. Local implementation in 1996 of standards pertaining to the use of clinical infusion devices made hourly recording of all clinical data in HDU compulsory. Furthermore, since the inception of the HDU epidural service, all medical analgesic interventions have been written on the epidural chart describing the reasons for epidural failure and the nature of the clinical response.
Data collection
The epidural charts of 1359 continuous patients were collected by the senior author, and every item of data manually transferred, irrespective of frequency, to an electronic database (Excel 98, Microsoft, Seattle, WA, USA). In order to minimize error, manual transfer of data and cross checking of randomly chosen charts were performed by the authors. Once every item of data was entered, the worst pain score, highest infusion rate, highest sedation score, lowest ventilatory frequency, lowest systolic blood pressure and the presence of motor block and nausea corresponding to each hour of epidural infusion were recorded. For analysis, we defined pain relief as a VRS=0, nausea and vomiting as a VRS 
1; motor block as any difficulty in straight leg raising on either side; sedation as a VRS 2; hypotension as the presence of a systolic blood pressure of either <100 mm Hg or <90 mm Hg and low ventilatory frequency was as either <10 or <8 bpm.
Because clinical observation of patients suggested the existence of patterns of pain and pain relief during thoracic epidural block, definitions of clinical end points were made. The time from the end of surgery (T0) to the first experience of pain score 1, 2 or 3 (Tf) was defined as the duration in hours (h) of initial or preventive pain relief (Tf–T0). The time (h) between the first (Tf) and last (Tl) experience of pain was characterized as the period of time (Tl–Tf) in which all episodes of pain, all increases in epidural infusion rate and all epidural bolus injections were made. In addition, the time (h) from the last experience of pain (Tl) to the end of the epidural (Te) was defined as the terminal pain free time (Te–Tl).
Statistical analysis
Patient and epidural characteristics were assessed using descriptive statistics. The Shapiro–Wilk test was applied to continuous data to determine the distribution of data. If variables were normally distributed, and independent groups had similar sample size and variance, repeated measures ANOVA and ANOVA were used, expressed as an F-ratio with Tukey–Kramer multiple comparison post-hoc test. For cumulative epidural infusion volume over time, group slopes were tested for parallelism, then compared using ANCOVA. For data not normally distributed, such as (Tf–T0), (Tl–Tf) and (Te–Tl), the non-parametric Kruskal–Wallis test was used, followed by the Kruskal–Wallis Z multiple comparison post-hoc test. In order to describe the frequency of nominal variables with respect to group, 2x6 contingency tables were used and analysed with the 
2-test. For all data P<0.05 was considered significant. Statistical analysis was performed using Number Cruncher Statistical Systems, UT, USA.
|
Results |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
Patient characteristics
The mean age (range) of patients was 66.5 (22–97) yr and the proportion of ASA I/II to III/IV patients was 50:50. The number of epidural catheters inserted per year increased from 126 in 1993 to a maximum of 302 in 2001. The proportion of females to males was 46:54.
Most epidural catheters (n,%) were inserted for bowel surgery (n=751, 55%). The remainder were inserted for oesophago-gastric (n=206, 15%), hepato-pancreatic (n=131, 10%), vascular (n=170, 13%), renal (n=31, 3%), umbilical hernia (n=20, 2%), pelvic (n=9, <1%) and splenic surgery (n=5,<1%). The operative details of 36 (2%) were not recorded. All patients undergoing vascular surgery had a transverse surgical incision.
Five hundred and forty-seven epidural catheters (40%) were inserted at interspaces T 8/9 or T 9/10 and 421 (31%) were inserted at T 10/11 or T 11/12. Only 120 (9%) epidural catheters were placed at T 7/8 and above, despite 337 patients undergoing oesophago-gastric or hepato-pancreatic surgery. One hundred and nineteen (9%) anaesthetists did not document the insertion level, and 152 (11%) of catheters were inserted at lumbar interspaces, although the proportion of the latter reduced from a peak of 35% in 1993 to a trough of 4% in 2000. All patients receiving a lumbar epidural catheter were included in the analyses.
Epidural administration of an opioid bolus during anaesthesia increased from 20 to 52% of patients between 1993 and 2001. The opioid of choice changed from fentanyl to diamorphine in 1996. Technical problems occurred in 142 patients (10%). Median [interquartile range (IQR)] time to technical failure was 27 h (14–46 h). Technical problems were classified as: (i) catheter problems (n=92): leaking (n=47), fell out (n=17), occluded (n=15), pulled out (n=9), severed (n=3), and migration (n=1); (ii) catheter malposition (epidural never worked) n=31; (iii) ineffective epidural despite bilateral sensory block (n=19). Twenty-five epidural catheters were resited.
Pain scores
Bedside recording of pain scores was undertaken for 58 118 out of 72 412 h, representing 80% of time during which patients received epidural analgesia. Pain scores 0, 1, 2 or 3 and S were recorded for 35 264, 7944, 2378 and 12 532 h respectively. On discontinuation of the epidural 1024 (75%) were pain free and 325 (25%) patients were in pain, 119 of these having a pain score of 2 or 3.
Patterns of pain
Graphical representation of the end points, Tf, Tl and Te is shown in Figure 1 as Kaplan–Meier survival plots. Data from patients not achieving any end point (event did not occur) was treated as right censored. The elapsed time between end of surgery (T0) and (Tf) is shown as (Tf–T0) and represents the initial pain free time. The period of time during which all episodes of pain, epidural boluses and increases in infusion rate occurred is represented by (Tl–Tf). The time (Te–Tl) represents the terminal pain free period from the last experience of pain until the end of the epidural.
|
Time to first pain score 1, 2 or 3 (Tf–T0)
More than one-third of patients (n=489, 36%) had pain immediately after operation, and another 328 (24%) had pain relief for less than 6 h after surgery. In contrast, 542 patients (40%) were pain free for >6 h, of which 196 were pain free for at least 24 h. One hundred and twenty patients had no pain.
Time to first pain score 1, 2 or 3 and subsequent pain
Increase in the time to first experience pain from 0 to >24 h (Tf–T0, Fig. 1) increased the median (IQR) duration of epidural analgesia from 44 (32–65) h to 51 (42–72) h (2 28.2, P<0.001) (Table 2); reduced the median (IQR) duration of intermittent pain (Tl–Tf, Fig. 1) from 34 (19–50) h to 3 (1–12) h (2 144, P<0.001); and increased median (IQR) terminal pain free time (Te–Tl, Fig. 1) from 6 (0–19) h to 15 (6–29) h (2 50.2, P<0.001).
|
Time to first pain score 1, 2 or 3 and pain at the end of epidural block
The duration of the initial pain free time also had a significant influence on the proportion of patients free from pain when the epidural was stopped. Table 3 shows that, as the initial pain free period increased from 0 to >24 h, the proportion of patients not experiencing pain rose from 0 to 45% (
2 398, P<0.001) and fewer patients had pain at the end of the epidural block, 31 to 11% (2 21.8, P<0.001).
|
Time to first pain score 1, 2 or 3 and epidural interventions
During epidural block, 1032 patients (76%) had 2708 increases in infusion rate and 543 patients (40%) had 1045 epidural boluses for pain relief. Extending the initial pain free time from 0 to >24 h had a significant effect on the proportion of patients requiring increases in epidural infusion rate and the proportion of patients receiving an epidural bolus (Table 4). As the initial pain free period was increased, the proportion of patients receiving an increase in epidural infusion rate reduced from 80 to 49% (
2 111, P<0.001) and the proportion of patients receiving an epidural bolus reduced from 53 to 8% (2 148, P<0.001).
|
Time to first pain score 1, 2 or 3 and epidural infusion rate
Overall, patients pain free >24 h received significantly less mean (SD) epidural drug compared with all other groups as assessed by repeated measures ANOVA [10.1 (0.5) ml h–1, F-ratio 9.9, P<0.001] (Fig. 2). Throughout the 60 h period, patients with pain relief between 12 and 18 h had a mean (SD) infusion rate of 12.4 (0.7) ml h–1 which was significantly greater than patients with <6 h pain relief 14.1 (0.4) ml h–1 and patients waking up in pain, 14.2 (0.4) ml h–1
|
When compared at 6 hourly intervals, using ANOVA, significant differences also emerged between groups. Up to 12 h after surgery, patients with between 6 and 12 h pain relief used significantly less drug compared with patients with <6 h or no pain relief. Thereafter infusion rates became indistinguishable in all three patient groups. In contrast, at this juncture, infusion rates in the three groups with >12 h pain relief became consistently lower than patients in the three groups with <12 h pain relief (at 12 h, F-ratio 40.7, P<0.001). Patients with >24 h pain relief required significantly less drug compared with all other groups (at 24 h, F-ratio 36.7, P<0.001).
Cumulative drug use over time is plotted in Figure 3. Because the slopes of cumulative drug volume were parallel (Fig. 3) (F-ratio 15.5, P<0.001), ANCOVA was used to compare groups. The cumulative volume of local anaesthetic was significantly less in patients with >12 h of freedom from pain (F-ratio 77.0, P<0.001) compared with those with <12 h pain relief.
|
Time to first pain score 1, 2 or 3 and side-effects
As the initial pain free period increased from 0 to >24 h, the proportion of patients with hypotension, low ventilatory frequency and sedation score
2 also increased (Table 5).
|
The proportion of patients with systolic BP <100 mm Hg increased from 59 to 77% (
2 30, P<0.001) and the proportion of patients with systolic BP <90 mm Hg increased from 32 to 43% (2 13.1, P<0.001). Increases in the initial pain free time increased the proportion of patients with ventilatory frequency <10 bpm from 13 to 26% (2 20.7, P<0.001) and the proportion of patients with sedation score 2 from 66 to 78% (2 15.5, P<0.008). The proportion of patients with ventilatory frequency <8 bpm increased from 3 to 7%, but was not statistically significant (2 5.3, P=0.38). Increasing the initial pain free period reduced the proportion of patients with motor block from 24 to 11% (2 15.4, P<0.008). No differences were seen with regard to nausea and vomiting (2 4.8, P=0.44). |
Discussion |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
This study has shown that the longer pain is prevented after laparotomy, the easier it is to manage that pain. After prolonged, uninterrupted pain relief, the proportion of patients receiving an increase in epidural infusion rate or an epidural bolus became smaller, episodes of pain became shorter and subsequent pain relief became longer, extending the duration of epidural block. Thus, our results confirm that time to first experience of pain is an important marker of the quality of thoracic epidural analgesia.
The strength of the study is that every piece of information recorded on the epidural charts of 1359 consecutive patients receiving epidural analgesia in the HDU was assessed. This comprised 80% of all potential hourly recordings over 72 412 patient hours. Recording of every pain score, increase in infusion rate, epidural bolus and side-effect provided the means of determining the distribution of pain scores and side-effects, and would not have been possible if nurse management of thoracic epidural analgesia had been conducted in the surgical wards.
We used the time to first experience of pain as our qualitative marker of pain relief because it incorporated the elements of depth, continuity and duration of pain relief integral to Kissin's hypothesis of preventive analgesia.4 We found that 489 patients woke in pain after surgery, and that 870 patients were pain free, but for a variable length of time, and that 120 patients had no pain at all (Fig. 1). Therefore, the experience of patients with thoracic epidural block was very wide, but within a narrow 2- to 3-fold range of epidural infusion rate. This is in marked contrast to PCA where patients have a 30-fold range of drug use.
Previous attempts to measure the quality of postoperative pain6 7 have focused on average patient experience over time, without taking the dynamics of postoperative pain into account. Awareness of the dynamics of pain helped us to reveal three distinct patient experiences (Fig. 1) during thoracic epidural block: a pain free period after surgery (Tf–T0); a period of intermittent pain, when epidural boluses are given and infusion rates are increased (Tl–Tf); and another pain free period (Te–Tl) until the end of the epidural. Identifying patterns of experience allowed us to determine how long patients remained within the period of intermittent pain (Tl–Tf), and the end pain free period (Te–Tl), and allowed us to determine any association with our quantitative marker, time to first experience of pain (Tf–T0).
We showed that when time to first pain was extended beyond 12 h, benefits started to accrue; pain became shorter, epidural boluses or increases in infusion rate were less frequent, local anaesthetic infusion rates were less likely to increase, and cumulative local anaesthetic consumption was reduced. Evidence from a variety of animal models confirms our findings, and suggests that profound blockade (both in intensity and duration) of noxious input to the spinal cord can provide benefits beyond the duration of the block. In a rat model using intraplantar carrageenan,8 sciatic and saphenous nerve block with tonicaine (lasting 12–16 h) prevented primary and secondary hyperalgesia, whereas, lidocaine, a short acting local anaesthetic, provided no benefit. In another rat model, but using paw pressure and hot plate withdrawal time,9 the combination of resiniferatoxin, a new ultrapotent vanilloid, and bupivacaine prevented hyperalgesia for up to 8 days compared with bupivacaine alone. Similarly, after transection of the saphenous nerve,10 both N-butyl tetracaine and lidocaine prevented early hyperalgesia for 3 h, but only N-butyl tetracaine prevented hyperalgesia for almost a week. One study on human volunteers has also shown similar benefits from prolonged nerve blockade. Pedersen and colleagues11 demonstrated that a prolonged (8–9 h) saphenous nerve block administered before thermal skin injury reduced post burn secondary hyperalgesia beyond the duration of the block.
However, our results have also shown that the longer the duration of pain relief, the greater the incidence of low ventilatory frequency and hypotension. Although a recent study12 has stated that Acute Pain Services should expect an incidence of respiratory depression, as defined by a low ventilatory frequency, of <1%, and an incidence of hypotension of <5%, the time course of side-effects during epidural block was not considered. In contrast, two studies using peripheral nerve block have shown changes in respiratory function after complete postoperative pain relief. In the first,13 the combination of femoral block and morphine was associated with airway obstruction and desaturation compared with femoral block alone, and, in the second,14 pain relief during axillary brachial plexus block was associated with reductions in CO2 sensitivity in patients with upper limb injuries.
Although we have shown an increase in hypotension with prolonged pain relief, this highlights the inadequacy of management of low blood pressure when i.v. fluids are preferred to vasoactive drugs.15 Greater consideration should be given to the treatment of perioperative hypotension in patients with epidural block by evaluating a wide range of haemodynamic parameters.
A recent editorial16 has stated that the quality of epidural anaesthesia needs to be defined before conclusions are made about the impact of thoracic epidural block on perioperative outcome.17 18 The results of our study suggest that the quality marker of thoracic epidural block should be preventive pain relief for at least 12 h, and we suggest that this should be used in studies when comparing thoracic epidural block with other modes of analgesia. But how can we routinely attain high quality in practice when only 415 (30%) patients achieved the standard in our study? The authors' own practice is to place an epidural catheter in the appropriate thoracic interspace, inject high concentration, low volume local anaesthetic before surgery, ensure that the patient awakens free from pain and take ongoing responsibility for the block in the days after surgery. However, sometimes it is not possible to attain long lasting pain relief, even with bilateral block, and a need exists to determine which independent factors predict prolonged pain relief and side-effects. Variables to be considered as independent predictors of the quality of epidural block may include age, sex, site of needle insertion, duration of surgery, experience of anaesthetist and surgeon, composition and mode of administration of the epidural solution, and prior chronic pain, depression or anxiety. Although the aim of our study was to provide a measure of the quality of epidural analgesia after laparotomy and not to identify factors which may account for large variability in pain scores, drug requirement and side-effects, it is our future intention to determine which independent variables predict time to first experience of pain, pain relief and side-effects by using statistical modelling techniques such as Cox's Proportionality method.
In conclusion, the time from surgery until the first experience of any pain using a VRS can be regarded as a marker of the quality of postoperative thoracic epidural analgesia. Application of this standard in 1359 consecutive patients showed that, as pain relief was extended, the proportion of patients needing an increase in epidural infusion rate or an epidural bolus became smaller, episodes of pain were shorter and subsequent pain relief became longer, extending the duration of epidural block. In our patient population, benefits started to accrue after 12 h of freedom from pain.
|
Acknowledgments |
|---|
We wish to acknowledge the contribution of Dr S. Manimaran FRCA in the collection of data.
|
Footnotes |
|---|
1Present address: John Radcliffe Hospital, Oxford, UK.
2Present address: Royal Hospital for Sick Children, Edinburgh, UK
|
References |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
1 Wilmore DW and Kehlet H. Management of patients in fast track surgery. Br Med J 2001; 322:473–6
2 McLeod G, Davies H, Munnoch N, Bannister J, MacRae W. Postoperative pain relief using thoracic epidural analgesia: outstanding success and disappointing failures. Anaesthesia 2001; 56:75–81[CrossRef][Web of Science][Medline]
3 Dolin SJ, Cashman JN, Bland JM. Effectiveness of acute postoperative pain management: I. Evidence from published data. Br J Anaesth 2002; 89:409–23
4 Kissin I. Preemptive analgesia. Anesthesiology 2000; 93:1138–43[CrossRef][Web of Science][Medline]
5 Eisenach JC. Preemptive hyperalgesia, not analgesia? Anesthesiology 2000; 92:308–9[CrossRef][Web of Science][Medline]
6 Barden J, Edwards JE, Mason L, McQuay HJ, Moore RA. Outcomes in acute pain trials: systematic review of what was reported? Pain 2004; 109:351–6[CrossRef][Web of Science][Medline]
7 Ready LB. Acute pain: lessons learned from 25,000 patients. Reg Anesth Pain Med 1999; 24:499–505[CrossRef][Web of Science][Medline]
8 Kissin I, Lee SS, Bradley EL Jr. Effect of prolonged nerve block on inflammatory hyperalgesia in rats: prevention of late hyperalgesia. Anesthesiology 1998; 88:224–32[CrossRef][Web of Science][Medline]
9 Kissin I, Bright CA, Bradley EL Jr. Selective and long-lasting neural blockade with resiniferatoxin prevents inflammatory pain hypersensitivity. Anesth Analg 2002; 94:1253–8
10 Kissin I, Lee SS, Bradley EL Jr. Hyperalgesia caused by nerve transection: long-lasting block prevents early hyperalgesia in the receptive field of the surviving nerve. Anesth Analg 1999; 89:1475–81
11 Pedersen JL, Crawford ME, Dahl JB, Brennum J, Davies H, Kehlet H. Effect of preemptive nerve block on inflammation and hyperalgesia after human thermal injury. Anesthesiology 1996; 84:1020–6[Web of Science][Medline]
12 Cashman JN and Dolin SJ. Respiratory and haemodynamic effects of acute postoperative pain management: evidence from published data. Br J Anaesth 2004; 93:212–23
13 Combes X, Cerf C, Bouleau D, Duvaldestin P, Dhonneur G. The effects of residual pain on oxygenation and breathing pattern during morphine analgesia. Anesth Analg 2000; 90:156–60
14 Bourke DL. Respiratory effects of regional anaesthesia during acute pain. Reg Anesth 1993; 18:361–5[Web of Science][Medline]
15 Gould TH, Grace K, Thorne G, Thomas. M. Effect of thoracic epidural anaesthesia on colonic blood flow. Br J Anaesth 2002; 89:446–51
16 de Leon-Casasola OA. When it comes to outcome, we need to define what a perioperative epidural technique is. Anesth Analg 2003; 96:315–18
17 Park WY, Thompson JS, Lee KK. Effect of epidural anaesthesia and analgesia on perioperative outcome: a randomized, controlled Veterans Affairs cooperative study. Ann Surg 2001; 234:560–9[CrossRef][Web of Science][Medline]
18 Rigg JR, Jamrozik K, Davies H, Myles PS, et al. Epidural anaesthesia and analgesia and outcome of major surgery: a randomised trial. Lancet 2002; 359:1276–82[CrossRef][Web of Science][Medline]
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  T. M. Cook, D. Counsell, J. A. W. Wildsmith, and on behalf of The Royal College of Anaesthetists Th Major complications of central neuraxial block: report on the Third National Audit Project of the Royal College of Anaesthetists Br. J. Anaesth., February 1, 2009; 102(2): 179 - 190. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. A. J. van Zundert, G. Stultiens, J. J. Jakimowicz, D. Peek, W. G. J. M. van der Ham, H. H. M. Korsten, and J. A. W. Wildsmith Laparoscopic cholecystectomy under segmental thoracic spinal anaesthesia: a feasibility study Br. J. Anaesth., May 1, 2007; 98(5): 682 - 686. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||