Postoperative pain relief using epidural analgesia has been used successfully for over two decades. Several trials and meta-analyses have shown better pain relief than intravenous opioids.1–4 The use of an epidural local anaesthetic has been shown to have beneficial effects on gastrointestinal mobility,5,6 the incidence of postoperative myocardial infarction,7 the incidence of pulmonary complications8,9 and postoperative morbidity and mortality in general.10 Epidural anaesthesia attenuates the stress response to surgery, facilitating the immune response.6,11 These benefits, however, have to be weighed against potential side effects and the risk of severe complications, such as epidural haematoma3 or epidural abscess.12 Currently, the place of epidural analgesia is being questioned due to newly available regional techniques and the lack of large published audits of this topic.13
Continuous epidural infusion (CEI) has the major disadvantage of not allowing for individual variation in the management of pain. Consequently, its use has given way in the last decade to patient-controlled epidural analgesia (PCEA). An increasing number of studies have shown significant benefit of PCEA over CEI, with a reduction in drug requirements for equivalent analgesia.14–16 Combining PCEA with a background infusion may provide the benefits of good pain relief together with high patient satisfaction.17–20
In our University Hospital in Sweden, we used CEI for the management of pain following major surgery as a routine procedure for more than 15 years, and the experience of both patients and staff has been satisfactory. However, the morphine-containing epidural solution required postanaesthesia care unit (PACU) observation overnight. That, and the inflexible nature of the administration system, gradually lessened its popularity. In the beginning of 2004 we implemented a new standardised protocol for intraoperative and postoperative epidural analgesia. We present here our hospital experience with a combined PCEA/CEI-based regimen in the management of postoperative pain.
Materials and methods
Formal approval from the Local Ethics Committee was not requested, as the study has the form of a paper audit, in which all data were collected from an anonymised quality registry. However, informal approval was given by the chairman.
A new regimen of postoperative epidural analgesia was introduced in our department on 1st March 2004. Prior to the introduction, endeavour was made to educate all anaesthetists, nursing and ancillary staff in the care of patients receiving PCEA. A bedside PCEA chart was introduced, wherein all patient-specific data were recorded (Appendix 1, http://links.lww.com/EJA/A49). The chart included a follow-up section to be filled in by the attending nurse during an interview with the patient at termination of the treatment. A short guide on how to treat observed side effects accompanied the chart, with the intention of facilitating prompt action by the ward staff (Appendix 2, http://links.lww.com/EJA/A49).
All patients scheduled for major surgery from March 2004 until February 2011 requiring epidural analgesia for postoperative pain relief were included in the survey. The five major surgical groups represented were elective upper abdominal surgery (proximal to the ligament of Treitz), elective lower abdominal surgery (distal to the ligament of Treitz), urologic surgery (nephrectomy and cystectomy), gynaecological cancer surgery and unspecified emergency abdominal surgery. A small number of other patients included were mainly scheduled for orthopaedic or plastic surgery. Obstetric patients were not included. The exclusion criteria were bleeding disorders, pre-existing neurological conditions, cardiovascular instability, infection at the site of epidural insertion and patient refusal. Low molecular weight heparin (LMWH) was given as thrombosis prophylaxis on the day before elective surgery. Emergency patients received LMWH postoperatively.
Prior to induction of anaesthesia, an epidural catheter (Portex Epidural Minipack System 2; Smiths Medical, St. Paul, Minnesota, USA) was placed using a standard 16-gauge Tuohy needle. The protocol gave firm guidance to the anaesthetist, instructing insertion of the catheter at the level most appropriate for the type of surgery. The following levels were recommended: T6-T7 for thoraco-abdominal surgery; T8-T10 for upper abdominal surgery; T9-T11 for lower abdominal surgery; and T10-T12 for gynaecological surgery. During surgery a CEI of bupivacaine 2.4 mg ml−1, fentanyl 1.8 μg ml−1 and adrenaline 2.4 μg ml−1 was administered by syringe pump, following a bolus of epidural fentanyl (25 to 100 μg). The intraoperative infusion rate was left to the discretion of the attending anaesthetist. In the event of severe cardiovascular instability, the intraoperative epidural infusion was abandoned.
After surgery, the patient was transferred to the PACU. There the intraoperative epidural solution was exchanged for a commercially produced (APL Pharma Specials, Stockholm Sweden) solution of bupivacaine 1 mg ml−1, fentanyl 2 μg ml−1 and adrenalin 2 μg ml−1, which was given via a PCEA pump (CADD-Legacy PCA Pump Model 6300; Smiths Medical, St. Paul, Minnesota, USA). The background infusion rate was titrated by the attending PACU anaesthetist within the range of 3 to 10 ml h−1. The aim was to achieve a numerical rating scale (NRS) pain level less than 4 without causing hypotension or other side effects. Bolus doses were set in a standardised manner (bolus dose 2 ml, refractory time 10 min, maximum four bolus doses per hour). In accordance with national recommendations, all patients were monitored in the PACU for a minimum of 6 h after starting the intraoperative fentanyl infusion. For the first 6 h, vital signs were recorded once every hour. These included blood pressure, heart rate, respiratory rate, pain level (NRS 0 to 10) at rest and on coughing, the degree of motor block (Bromage score 0 to 3), sedation score (0 to 3), and occurrence of nausea and pruritus (yes or no).
All discharge criteria had to be fulfilled before transfer from the PACU to the general surgical ward. These included circulatory and respiratory stability, an awake and alert state, ability to communicate, NRS less than 4, no motor block and satisfactory progress with mobilisation, for example, standing unassisted. In addition each patient was taught how to self-administer bolus doses on the PCEA pump. After the first 6 h the recordings of vital signs were reduced to every 4 h. This interval was maintained throughout the remaining treatment period. Renewal of the infusion system was recommended if the treatment exceeded 4 days. To prevent continuation of the infusion beyond that required, we recommended catheter removal within 5 days. Bolus doses requested and given were recorded on a daily basis by the ward staff. Paracetamol was given routinely throughout the postoperative period, whereas NSAIDs were given as indicated. In general, we recommended that wards refrain from giving supplementary intravenous (i.v.) opioids or sedatives during the epidural infusion.
An acute pain service team consisting of two intensive care nurses made their rounds on the surgical wards twice every week and on-demand during daytime hours. The anaesthetist on-call was responsible for answering questions and solving acute problems related to epidural analgesia. After termination of the treatment, the PCEA chart was collected by the acute pain service APS team. The variables of interest were then manually transferred from the chart to a PCEA quality registry, in which all patients were anonymised.
The following data were processed over the entire 7-year period 2004 to 2011: the level of insertion of the epidural catheter; the duration of treatment (days), the infusion rate at PACU discharge (ml h−1); the reason for termination of treatment (1 = elective, 2 = inadequate analgesia, 3 = suspected infection, defined as inflammation around the puncture site, 4 = other); the satisfaction score (1 to 10, where 1 = very poor and 10 = excellent analgesia) at termination of the treatment; and the incidence of reinsertion of an epidural catheter due to inadequate analgesia (%).
During the 6-year period 2005 to 2011, data were collected to determine the incidence of transition to i.v. PCA with morphine as a rescue method when the epidural treatment had failed.
During the 3-year period 2004 to 2006, the following data were processed: the degree of motor block (Bromage 0 to 3) at 12 h postoperatively; the degree of sedation 12 h postoperatively (sedation score 0 = awake; 1 = drowsy; 2 = sleeping, easy to wake up; 3 = sleeping, difficult to wake up); the occurrence of nausea (yes or no); and pruritus (yes or no). For each patient, the worst recording for motor block and sedation was processed. As for nausea and pruritus, a single recording of ‘yes’ during the whole treatment would label that patient positive for the respective side effect.
During the 2-year period 2007 to 2008, data were collected to calculate to what extent bolus doses contributed to the total infused volume in each patient from readings from the PCEA pump. The total amount of epidural solution administered to each patient was recorded, and the number of standardised bolus doses given was available from the PCEA pump.
Statistical analysis was primarily descriptive. Median and interquartile range was used for the level of insertion. χ2 test was used for categorical data, such as the requirement for a new epidural catheter, conversion to PCA, satisfaction score, reason for termination, motor blockade, sedation, nausea and pruritus. Student's t-test was used for continuous data, such as infusion rate, duration and bolus dose volume ratio, which are expressed as mean ± SD. Analysis of variance (ANOVA) with Sidak post hoc test was used for comparison of more than two groups for continuous data. When comparing women with men, a risk ratio with 95% confidence interval (95% CI) was calculated. A P value of less than 0.05 was considered significant.
A total of 4912 epidurals were inserted in 4663 patients over the 7-year period. General characteristics of the patients are presented in Table 1. Four thousand eight hundred and thirty-one (98.4%) received the standard postoperative solution. In the remaining 81, analgesia failed despite adequate dermatomal spread, and the standard solution was usually replaced with one that contained a combination of ropivacaine 3 mg ml−1, fentanyl 5 μg ml−1 and adrenaline 2 μg ml−1. However, to be complete and to avoid bias, all patients were included in the survey.
The levels of catheter insertion are presented in Table 2 and are subdivided according to the type of surgery performed.
Infusion rates at discharge from the PACU subdivided according to type of surgery are presented in Table 2. The infusion rate in women was significantly lower than in men, 5.1 ± 1.6 vs. 5.9 ± 1.8 ml h−1 (P < 0.001). Downward adjustment of the mean infusion rate with increasing age corresponded to 0.3 ml h−1 per 10 years (P < 0.001), and is presented in Fig. 1.
The duration of the epidural infusions in days is shown in Table 2, after subdivision into the types of surgery. The duration is also presented unselectively in Fig. 2.
The contribution of bolus dose to the total infused volume (compliance 98%) is presented in Table 2. These ratios did not vary significantly between different types of surgery or between men and women.
The satisfaction scores at the end of the infusion (compliance 64%) are presented in Table 2. Scores of 8 to 10 were considered satisfied, whereas scores of 1 to 3 were considered dissatisfied.
The incidence of reinsertion of an epidural catheter, the incidence of converting to i.v. PCA with morphine and the reason for termination of the treatment are presented in Table 2. Other causes for termination are specified in Table 3. The risk ratio for terminating the treatment due to motor blockade was 6.3 (2.8–13.9, 95% CI) for patients who received a lumbar epidural catheter. No patient developed signs of epidural haematoma. One patient had an epidural abscess diagnosed by computed tomographic (CT) scan and received antibiotic treatment but no surgical intervention was necessary.
The incidences of motor blockade (n = 1603) and sedation (n = 1631) are subdivided into the different types of surgery and presented in Table 4. Only one patient developed a Bromage score of 3, seven a Bromage score of 2 and 20 a score of 1. The remainder had no motor block. One patient recorded a sedation score of 3 on the ward on a single occasion, but no medical intervention was required. Eight had a score of 2, and 25 scored 1. The remainder scored 0. No gender differences were seen for motor blockade and sedation.
The incidences of nausea and pruritus with gender differences for these side effects are presented in Table 4. There was no significant difference between the incidence of nausea in the PACU following gynaecological and nongynaecological surgery in women (77/457, 16.8% vs. 90/581, 15.5%).
This survey reflects 7 years of use of our PCEA protocol. Except for introducing the administration of a stronger solution to a minority of patients in the last years, we have remained faithful to the original concept. The standard epidural solution we use is well known for its efficacy and safety profile.21–24 We believe that choosing a commercially available solution like this overcomes the sterility concerns attached to administration of a home-made preparation.
It is well recognised that inserting the epidural catheter at the spinal level most appropriate for the surgery in question will reduce the amount of drug required for effective pain relief. With reduced dose, the incidence of side effects, such as hypotension, motor blockade and sedation is also reduced.25 Only 0.6% of our patients had hypotension as the reason for terminating their treatment. This low incidence is due to our emphasis on placing the epidural catheter at the appropriate level together with thorough individual titration of the background infusion prior to discharge from the PACU.
The vast majority of patients (94%) had their epidural treatment for a maximum of 5 days. Longer duration than this has been associated with an increased risk of infection.26,27 Only a small proportion of our treatments exceeded 5 days, and we attribute our low incidence of suspected infection (0.7%) to this. Gynaecological patients had a significantly shorter length of treatment in our survey. As most gynaecological operations do not involve the intestine, postoperative ileus is a rare problem,28 and this may have contributed to the shorter need for epidural analgesia.
We confirm the relationship between age and epidural dose requirements.25 Older patients need less volume to produce the same response in an almost linear correlation. The 14% lower infusion rate in women is likely to be related to BMI, which agrees with recent findings in which a 10% dose reduction in women was seen.29
Inadequate analgesia led to premature termination of epidural analgesia in 10.9%. In most instances, the patient was offered either a new epidural catheter (5.3%) or i.v. morphine PCA as rescue treatment (5.0%). Premature termination due to inadequate analgesia was significantly more common following upper abdominal surgery than the rest (15.0%, P = 0.006). This group were also more likely to receive rescue i.v. PCA (7.9%, P < 0.001) and showed a tendency towards an increased risk of receiving a new epidural catheter (6.8%, P = 0.11). This increased failure rate might be related to the extensive nature of the surgery, which included oesophageal resections, liver resections and pancreatic surgery.30 Overall failure rates of between 7.5 and 25% have been reported for epidural analgesia,31–33 placing our results at a mid-range level. In our survey, dislodgement of the catheter and leakage of the solution from the puncture site were the most common catheter-related causes for termination of the treatment (5.4%). Other surveys present comparable results ranging between 7 and 12%.3,31,34,35
The incidence of motor block was generally low (1.7%), and only 1% of the overall treatments were terminated due to this complication. There was an expected higher frequency of motor block in the minority of patients who received a lumbar epidural rather than a thoracic epidural (4.0 vs. 0.6%), stressing the importance of the appropriate insertion level of the catheter in abdominal surgery.
No patient in our survey developed symptoms of epidural haematoma. The quoted incidence of epidural haematoma varies in different studies. One large retrospective analysis from Sweden reports an incidence of 1 : 18 000 patients,36 whilst in three more recent prospective, single-centre studies the incidence ranged between 1 : 2700 and 1 : 4761.3,37,38 The incidence of epidural abscess in the same studies varies between 1 : 1350 and 1 : 7112, which puts our own result of 1 : 4912 in mid-range.
Six months into the study, we did an interim analysis of the NRS scores (263 patients). There were no significant differences between patient gender or the types of surgery, and the results showed a great variance. As a result of this, we stopped transferring the NRS data into the database for the remainder of the study. Instead, we took patient satisfaction score as the method for assessing the analgesic effect of the epidural treatment. There is no accepted standard for evaluating patient satisfaction after postoperative epidural analgesia. Many confounding factors may contribute to the degree of satisfaction. However, we asked the patients to score the epidural analgesia specifically at termination of the treatment. We found the 10-point score we used to be simple to perform and easy to understand. Some 87% scored 8 to 10, indicating a high satisfaction level. The recording compliance was only 64%, but given the high sample size (n = 3123), we consider the results valid. Our protocol permitted the use of analgesic adjuncts, which calls for a cautious interpretation of the quality of analgesia from the epidural alone.
Respiratory depression is a well known risk of opioids. We chose to measure the degree of sedation rather than slow respiratory rate, as sedation is more sensitive and always accompanies the latter. There are several sedation scores and various published definitions of sedation, making it difficult to compare the sedative effects of different epidural regimens. The range in several large studies varies enormously from 0.2 to 31%.1,19,31,34,35,39 We regarded any observation of mild sedation as indicating a sedated patient. In that perspective, our overall incidence of 2.1% seems low. Lower abdominal surgery patients experienced significantly more sedation than the others (3.3%, P = 0.025) despite receiving similar volumes of epidural solution. We speculate that this is due to a more frequent use of supplemental opioids in a group in which surgery sometimes involves the perineum, lying below the band of epidural analgesia.
The reported incidence of nausea and vomiting also varies greatly. A review of pooled data compared the incidence of all opioid-related side effects in epidural analgesia, i.v. PCA and intramuscular (i.m.) analgesia.40 The overall mean incidence (95% CI) of nausea in 23 782 patients was 25.2% (19.3 to 32.1). The incidence in the epidural group was 18.8% (14.0 to 24.8) and in the i.v. PCA group 32.0% (26.8 to 37.6). Another large study confirms the lower incidence in PCEA than in i.v. PCA (13.7 vs. 19.7%).3 The incidence of nausea in our survey was 12.8% in the PACU, gradually decreasing over the treatment period. We confirm the well established gender difference with a risk ratio of 2.3 (1.6 to 3.2, 95% CI) for women compared with men. Our relatively low incidence of nausea may be attributed to the low concentration of fentanyl (2 μg ml−1) in the epidural solution.
Pruritus is a very common side effect of opioid-based epidural analgesia. In a large study comprising 21 461 patients, the mean incidence was 16.1% (95% CI 12.8 to 20.0), which was significantly higher than for other opioid-based pain relief methods.40 In our survey, the mean incidence on day 1 after surgery was 18.8%. Female patients were significantly more prone to develop pruritus than males (P < 0.001). We cannot explain why pruritus scores are higher on day 1 than in the PACU. We hypothesise that most patients in the PACU have to cope with more important issues during the first hours after surgery, such as recovering from anaesthesia and regaining circulatory and respiratory stability. Pruritus is often a mild side effect, and in most cases, it will suffice to inform the patient about its cause to gain acceptance. Only five patients of 4135 had to terminate their treatment due to unacceptable pruritus.
Using a combined PCEA/CEI regimen may have the advantage of improving pain relief compared to PCEA alone.17–19 Whether the combined method increases the total administered drug amount or not remains controversial, but the incidence of serious side effects does not seem to increase.17,19 In one study, the authors detected better pain relief during night-time, better sleep and less pain on coughing when using a background infusion in addition to PCEA.20 In this survey, we found that the PCEA contribution accounted for only 16% of the total infused drug amount, actually making our regimen predominantly CEI-based. We believe the PCEA demands are used mainly prior to mobilisation, as this is emphasised when teaching the patients how to use the PCEA system.
The main limitation of this survey is the lack of a long-term follow up regarding potential neurologic complications. However, no cases of neurologic sequelae have come to our attention during the investigation.
In conclusion, we have introduced a ward-based PCEA service for postoperative pain relief at a Swedish University hospital and conducted a prospective survey covering the first 7 years. Despite limited acute pain service team resources, we have found this regimen to be both effective and relatively well tolerated.
Acknowledgements relating to this article
Assistance with the study: the author would like to thank Mona Lindblad, Annki Segerstéen and Annika Johansson, APS nurses for their excellent work on the wards and their persistence in transferring all data into the database. Thanks also to Mats Fredrikson, LARC, for his expertise with the statistical analysis.
Financial support and sponsorship: none.
Conflicts of interest: none.
1. Wu CL, Cohen SR, Richman JM, et al. Efficacy of postoperative patient-controlled and continuous infusion epidural analgesia versus intravenous patient-controlled analgesia with opioids: a meta-analysis. Anesthesiology
2. Werawatganon T, Charuluxanun S. Patient controlled intravenous opioid analgesia versus continuous epidural analgesia for pain after intra-abdominal surgery. Cochrane Database Syst Rev
3. Popping DM, Zahn PK, Van Aken HK, et al. Effectiveness and safety of postoperative pain management: a survey of 18 925 consecutive patients between 1998 and 2006 (2nd revision): a database analysis of prospectively raised data. Br J Anaesth
4. Block BM, Liu SS, Rowlingson AJ, et al. Efficacy of postoperative epidural analgesia: a meta-analysis. JAMA
5. Jorgensen H, Wetterslev J, Moiniche S, et al. Epidural local anaesthetics versus opioid-based analgesic regimens on postoperative gastrointestinal paralysis, PONV and pain after abdominal surgery. Cochrane Database Syst Rev
6. Waurick R, Van Aken H. Update in thoracic epidural anaesthesia. Best Pract Res Clin Anaesthesiol
7. Beattie WS, Badner NH, Choi P. Epidural analgesia reduces postoperative myocardial infarction: a meta-analysis. Anesth Analg
8. Ballantyne JC, Carr DB, deFerranti S, et al. The comparative effects of postoperative analgesic therapies on pulmonary outcome: cumulative meta-analyses of randomized, controlled trials. Anesth Analg
9. Groeben H. Epidural anesthesia and pulmonary function. J Anesth
10. Rodgers A, Walker N, Schug S, et al. Reduction of postoperative mortality and morbidity with epidural or spinal anaesthesia: results from overview of randomised trials. BMJ
11. Ahlers O, Nachtigall I, Lenze J, et al. Intraoperative thoracic epidural anaesthesia attenuates stress-induced immunosuppression in patients undergoing major abdominal surgery. Br J Anaesth
12. Grewal S, Hocking G, Wildsmith JA. Epidural abscesses. Br J Anaesth
13. Rawal N. Epidural technique for postoperative pain: gold standard no more? Reg Anesth Pain Med
14. van der Vyver M, Halpern S, Joseph G. Patient-controlled epidural analgesia versus continuous infusion for labour analgesia: a meta-analysis. Br J Anaesth
15. Standl T, Burmeister MA, Ohnesorge H, et al. Patient-controlled epidural analgesia reduces analgesic requirements compared to continuous epidural infusion after major abdominal surgery. Can J Anaesth
16. Nightingale JJ, Knight MV, Higgins B, et al. Randomized, double-blind comparison of patient-controlled epidural infusion vs nurse-administered epidural infusion for postoperative analgesia in patients undergoing colonic resection. Br J Anaesth
17. Bremerich DH, Waibel HJ, Mierdl S, et al. Comparison of continuous background infusion plus demand dose and demand-only parturient-controlled epidural analgesia (PCEA) using ropivacaine combined with sufentanil for labor and delivery. Int J Obstet Anesth
18. Halpern SH, Carvalho B. Patient-controlled epidural analgesia for labor. Anesth Analg
19. Komatsu H, Matsumoto S, Mitsuhata H, et al. Comparison of patient-controlled epidural analgesia with and without background infusion after gastrectomy. Anesth Analg
20. Komatsu H, Matsumoto S, Mitsuhata H. Comparison of patient-controlled epidural analgesia with and without night-time infusion following gastrectomy. Br J Anaesth
21. Niemi G, Breivik H. The minimally effective concentration of adrenaline in a low-concentration thoracic epidural analgesic infusion of bupivacaine, fentanyl and adrenaline after major surgery. A randomized, double-blind, dose-finding study. Acta Anaesthesiol Scand
22. Niemi G, Breivik H. Adrenaline markedly improves thoracic epidural analgesia produced by a low-dose infusion of bupivacaine, fentanyl and adrenaline after major surgery. A randomised, double-blind, cross-over study with and without adrenaline. Acta Anaesthesiol Scand
23. Niemi G, Breivik H. Epidural fentanyl markedly improves thoracic epidural analgesia in a low-dose infusion of bupivacaine, adrenaline and fentanyl. A randomized, double-blind crossover study with and without fentanyl. Acta Anaesthesiol Scand
24. Kjonniksen I, Brustugun J, Niemi G, et al. Stability of an epidural analgesic solution containing adrenaline, bupivacaine and fentanyl. Acta Anaesthesiol Scand
25. Visser WA, Lee RA, Gielen MJ. Factors affecting the distribution of neural blockade by local anesthetics in epidural anesthesia and a comparison of lumbar versus thoracic epidural anesthesia. Anesth Analg
26. Ranasinghe JS, Lee AJ, Birnbach DJ. Infection associated with central venous or epidural catheters: how to reduce it? Curr Opin Anaesthesiol
27. Holt HM, Andersen SS, Andersen O, et al. Infections following epidural catheterization. J Hosp Infect
28. Fanning J, Hojat R. Safety and efficacy of immediate postoperative feeding and bowel stimulation to prevent ileus after major gynecologic surgical procedures. J Am Osteopath Assoc
29. Schnabel A, Poepping DM, Gerss J, et al. Sex-related differences of patient-controlled epidural analgesia for postoperative pain. Pain
30. Revie EJ, Massie LJ, McNally SJ, et al. Effectiveness of epidural analgesia following open liver resection. HPB (Oxford)
31. Tan T, Wilson D, Walsh A, et al. Audit of a ward-based patient-controlled epidural analgesia service in Ireland. Ir J Med Sci
32. Motamed C, Farhat F, Remerand F, et al. An analysis of postoperative epidural analgesia failure by computed tomography epidurography. Anesth Analg
33. Pan PH, Bogard TD, Owen MD. Incidence and characteristics of failures in obstetric neuraxial analgesia and anesthesia: a retrospective analysis of 19,259 deliveries. Int J Obstet Anesth
34. Wigfull J, Welchew E. Survey of 1057 patients receiving postoperative patient-controlled epidural analgesia. Anaesthesia
35. Liu SS, Allen HW, Olsson GL. Patient-controlled epidural analgesia with bupivacaine and fentanyl on hospital wards: prospective experience with 1,030 surgical patients. Anesthesiology
36. Moen V, Dahlgren N, Irestedt L. Severe neurological complications after central neuraxial blockades in Sweden 1990–1999. Anesthesiology
37. Christie IW, McCabe S. Major complications of epidural analgesia after surgery: results of a six-year survey. Anaesthesia
38. Cameron CM, Scott DA, McDonald WM, et al. A review of neuraxial epidural morbidity: experience of more than 8,000 cases at a single teaching hospital. Anesthesiology
39. Scott DA, Beilby DS, McClymont C. Postoperative analgesia using epidural infusions of fentanyl with bupivacaine. A prospective analysis of 1,014 patients. Anesthesiology
40. Dolin SJ, Cashman JN. Tolerability of acute postoperative pain management: nausea, vomiting, sedation, pruritus, and urinary retention. Evidence from published data. Br J Anaesth