Thoracic epidural analgesia (TEA) is considered the gold standard for analgesia following abdominal surgery. Alternatives to epidural analgesia such as wound infusions, transversus abdominis plane (TAP) block and paravertebral block have been a topic of interest in recent years.1 Although TEA is commonly practised, there is a subset of patients in whom use of TEA poses increased risk. Studies evaluating the alternatives to TEA as continuous techniques are sparse.
Although the incidence of inadequate analgesia with TEA ranges between 28 and 32%,2,3 it is thought to provide better analgesia than techniques such as TAP block due to sparing of visceral innervation with the latter technique. A majority of comparative trials of TAP block and TEA have employed single-injection TAP blocks and reported better analgesia with TEA,4 but to date there have been only a few studies evaluating the analgesic efficacy of continuous TAP block.5–7 These studies have documented increased opioid consumption with TAP block, which could be attributed to unpredictable spread of local anaesthetic in the TAP, resulting in sparing of T7 to T8 or L1.8,9 We recently described a lateral-to-medial approach to perform the TAP (LM-TAP) blocks in 16 cadavers and noted that subcostal injection combined with inferior subumbilical injection covered T7 to L1 dermatomes reliably.10 This technique permits preoperative initiation of continuous TAP blocks. We wanted to evaluate the clinical feasibility of initiating continuous LM-TAP blocks preoperatively and compare the analgesic efficacy with that of the conventional TEA practised in our hospital. The primary outcome was pain score on coughing at 24 h in this noninferiority open-label study. The open-label design was chosen to minimise undue risk to the participants from blinding to the procedures, including monitoring and troubleshooting.
Materials and methods
The study was conducted on an intention-to-treat analysis basis between July 2008 and August 2012 after obtaining institutional ethics board approval (HSREB/IRB number: 100981, Western University, London, Ontario; Dr Joseph Gilbert, Chair, Health Sciences REB) and written informed consent from the patients participating in the study.
A total of 159 patients were approached and 50 patients were included in this prospective, randomised, open-label study (Fig. 1). The patients were allocated using a simple random sequence generated by a random draw of numbers, and the numbers were concealed using a sealed envelope technique. The patients were informed about group allocation only after they consented to take part in the study in the preadmission clinic. Investigator 1 (S.G.) generated the random allocation sequence and investigators 3 and 6 (M.T. and L.S.) enrolled patients for the trial. Investigators 4 and 5 (J.B. and S.D.A.) assigned participants to the intervention. All the study procedures were performed at the University Hospital, London Health Sciences Centre, London, Ontario, Canada. Patients were educated about the 11-point verbal rating score (VRS) for pain (0 to 10) where 0 is no pain and 10 is the worst pain imaginable. Inclusion and exclusion criteria are summarised in Table 1. All patients received preoperative multimodal analgesia using oral naproxen 500 mg, gabapentin 600 mg and paracetamol 975 mg. These were continued postoperatively for 4 days (naproxen 500 mg twice daily, gabapentin 300 mg twice daily and paracetamol 650 mg every 6 h).
All procedures were performed in the block area using standard monitors (NIBP, pulse oximetry) and sedation (fentanyl and midazolam) was titrated to effect. Group TEA received catheter-congruent TEA in the sitting position using sterile precautions at T7 to T8 or T8 to T9 level. After identifying the epidural space with loss of resistance, 5 cm of the catheter was inserted into the epidural space. After a 3-ml test dose of lignocaine 2% with adrenaline 5 μg ml−1 to rule out intravascular or intrathecal placement of the catheter, a deliberate attempt was made to obtain an initial block extent between T6 and T12 dermatomes. To obtain an ideal block between T6 and T12, an initial bolus of 0.25% bupivacaine 5 ml was administered followed by pinprick testing at 15 min and 5-min intervals subsequently until 30 min after the start of the initial injection. Additional 3-ml doses were administered every 5 min if the block was not established between T6 and T12 at each of the assessment periods. An infusion of bupivacaine 0.1% with hydromorphone 10 μg ml−1 was started preoperatively at a rate of 8 ml h−1 and continued for 72 h.
Group LM-TAP had bilateral LM-TAP catheters inserted using a linear ultrasound probe (7 to 13 MHz; Sonosite M-turbo, Bothel, Washington, USA). Patients were in the supine position with a wedge of blankets placed under the back on one side and a preprocedural scan was performed to identify the landmarks for performance of the block. The skin of the abdominal wall was prepared with an antiseptic solution and the probe was positioned anterior to the midaxillary line in an oblique subcostal angle. After anaesthetising the skin with lignocaine 1%, the catheter insertion needle (17G Tuohy needle with cannula assembly) connected to a syringe containing ropivacaine 0.5% was inserted under ultrasound guidance to hydro-dissect the TAP in the subcostal plane until the needle tip was lateral to the edge of the rectus abdominis muscle. Having distended the space with 20 ml of ropivacaine 0.5%, a multiorifice catheter was inserted through the needle/cannula. Following this, the probe was rotated to remain parallel to the iliac crest and the Tuohy needle was inserted from the midaxillary line in the transversus plane until the needle tip was lateral to the edge of the rectus. The inferior TAP catheter was inserted through the Tuohy needle after distending the transversus plane with 10 ml of ropivacaine 0.5% through the needle. Both the catheters were tunnelled subcutaneously and the entry site was covered with skin glue and an occlusive dressing (see supplementary video, http://links.lww.com/EJA/A77). A similar procedure was repeated to insert the block catheters on the opposite side. The two catheters on each side were connected to a single elastomeric pump via a Y-connector to deliver ropivacaine 0.35% at a rate of 4 to 5 ml h−1 for 72 h. In the first eight patients, the I-Flow or Solace insertion system was used with a peelable cannula to facilitate insertion of the multiorifice catheter, but this was abandoned due to technical difficulties discussed later. In the remainder of the patients, we used the Pajunk system (InfiltraLongSono kit; Pajunk Medizintechnologie GmbH, Geisingen Germany). Because we used the maximum permissible dosage of local anaesthetic to initiate the block in the LM-TAP group, we could not inject any further local anaesthetic through the catheters in the event of inadequate block spread.
The extent of the preoperative sensory block in both groups was determined as the blockade of pinprick sensation bilaterally 30 min after the initial injection or prior to moving the patient to the operating room. No opioids were added to the local anaesthetic solution used for the LM-TAP blocks and no patient-controlled bolus option was available for the LM-TAP infusions. Once the effectiveness and level of the block were confirmed, the patients received general anaesthesia with propofol, fentanyl, rocuronium and an inhalational agent for surgery. Anaesthesiologists providing intraoperative care had a choice of using either intravenous (i.v.) fentanyl or hydromorphone during the surgery, guided by haemodynamics.
Postoperatively, patients in the TEA group were allowed to self-administer 3-ml boluses every 20 min, while patients in the LM-TAP group were provided with rescue i.v. patient-controlled analgesia (i.v.-PCA) using hydromorphone 0.2 mg with a lockout time of 6 min and no basal infusion. The dose of PCA was increased to 0.3 mg if pain control was poor (VRS >5/10) consistently at any two assessment periods. The second rescue analgesia for both group of patients in the postanaesthesia care unit (PACU) consisted of nurse-administered i.v. hydromorphone 0.4 mg boluses every 5 min as necessary if the pain was severe (VRS >6/10), irrespective of their use of the first rescue analgesic modality. If a patient in either group had a primary catheter failure as defined by lack of sensory block on preoperative testing at 30 min or a secondary block failure after discharge from PACU, i.v.-PCA was given for rescue analgesia. If secondary block failure was noted in the PACU, the catheters were resited prior to discharge from PACU.
Postoperatively, patients were evaluated for pain during rest and coughing immediately after surgery, hourly for the first 4 h and 6-hourly thereafter until 72 h. Nausea and pruritus were documented using a categorical scale (none, mild, moderate or severe) every 12 h. Time to onset of bowel sounds and time to first bowel movement were also recorded, with time 0 being arrival in PACU. Intraoperative fentanyl and hydromorphone consumption were reported separately and the intraoperative and postoperative hydromorphone consumptions were reported as morphine equivalents. Haemodynamic events were documented as the incidence of significant hypotension (defined as SBP less than 90 mmHg) even after corrective measures of volume loading and cessation of the analgesic infusion. Patient satisfaction was documented on a visual analogue scale where 0 was totally dissatisfied and 100 was completely satisfied. Monitoring for adverse events included secondary block failure rates, postoperative confusion and neurological deficits in the postoperative period. Postoperative confusion was defined for study purposes as the subjective reporting of confusion as assessed by the floor nurses.
Power analysis was performed for the primary variable, which was pain scores on coughing at 24 h postoperatively. Mean ± SD dynamic pain scores with the use of TEA in the first 24 h ranged from 1.2 ± 0.7 to 3.9 ± 2.4, as reported in previous studies comparing TEA with i.v.-PCA.11 Assuming an intermediate SD of 1.7 in TEA, we needed a total of 46 patients (23 patients per group) to be 90% sure that the lower limit of a two sided 90% confidence interval would be above the noninferiority limit of -1.5 if there was truly no difference between the standard (TEA) and the experimental (LM-TAP) treatments. We included 25 patients per group to account for a 10% possible attrition of patients. If the noninferiority between the groups was disproved, we planned to analyse for superiority between the two groups.
All data were analysed using SPSS version 17 for windows (SPSS Inc, IBM Corp, Chicago, Illinois, USA). All variables were analysed for normality of distribution using the one-way Kolmogorov–Smirnov test. Categorical variables were analysed using the χ2 test and continuous variables were analysed using the independent sample t-test. Pain scores were tested for noninferiority using the difference in 95% confidence interval of the means at various time intervals. Continuous variables were analysed using repeated measures analysis of variance (ANOVA) after testing for sphericity. A P value less than 0.05 was considered statistically significant.
All 50 patients enrolled successfully completed the study and no patient had a primary block failure. Patient characteristics and duration of surgery were comparable between the groups. The surgical indications (Table 2) were comparable.
The pain scores at rest (Fig. 2a) and on coughing (Fig. 2b) were noninferior between the two groups at 24 h postoperatively. At other time intervals, the pain scores at rest and coughing were noninferior except for the dynamic pain scores (VRScough) at 4, 12 and 18 h (Fig. 2). The maximum pain scores at rest in the first 24 h were noted in the first 4 h postoperatively and on movement at 24 h in the TEA group and at 18 h in the LM-TAP group. The pain scores during rest and coughing in the LM-TAP group showed a lower variance than in the TEA group at all time intervals. In both groups, the pain scores at rest were significantly lower at 24 to 72 h postoperatively than the first 24 h (repeated measures ANOVA, P = 0.001) whereas the dynamic pain scores did not show a similar decline over time (first 24 h vs. the subsequent 24 to 72 h time intervals, P = 0.020).
The procedure time and extent of block are summarised in Table 2. The upper limit of the block in TEA was as high as T3 in four patients and as low as T8 in one patient. In the LM-TAP group, the upper limit of the block was no higher than T5 to T6 in any patient and was as low as T9 in two patients. The lower limit of the block in TEA was as high as T10 (three patients) and as low as L3 (one patient). Similarly, in the LM-TAP group, the lower limit of the block was as high as T10 (one patient) and as low as L1 (six patients).
The secondary study parameters were comparable between the two groups except for hypotension, and for opioid requirements during operation and between 24 and 48 h (Table 3); there were no significant differences in respiratory rate and oxygen saturation between groups at any time. Five of the 24 patients in the TEA group had significant hypotension necessitating cessation of the epidural infusion. Three of these patients needed assessment for ICU care. None of the patients in the LM-TAP group had significant hypotension requiring cessation of infusions. There were seven patients (29.2%) with inadequate pain relief (VRSrest >5/10) at some point within the first 24 h postoperatively in the TEA group compared with four patients (15.4%) in the LM-TAP group. Of the seven patients with inadequate analgesia in the TEA group, two patients had epidural catheter displacement, one of which was recognised and resited in the PACU; the other was on the first postoperative day and i.v.-PCA was substituted. Of the remaining five patients, three had increased pain scores in the first 4 to 8 h postoperatively and the other two patients had VRS pain scores more than 5 at multiple time intervals. One of the four patients who had inadequate pain relief in the LM-TAP group had leakage from both the right and left catheters on postoperative day 1, which resulted in inadequate pain relief. Of the three other patients with higher pain scores at various time intervals, two had an anastomotic leak.
We evaluated the feasibility of initiating preoperative continuous LM-TAP blocks and assessing the efficacy and safety in comparison with TEA for abdominal surgery. Our study shows that the LM-TAP block can be initiated safely prior to a variety of abdominal operations. The LM-TAP blocks underlying a multimodal analgesic regimen resulted in noninferior analgesia to TEA without any significant differences between the two groups in terms of secondary outcomes. Most of the techniques of performing TAP block necessitate a medial-to-lateral insertion of the needle, which positions the puncture point close to the surgical area, precluding preoperative initiation. We described a technique of performing TAP blocks from the mid-axillary line in a lateral-to-medial approach in a recent cadaveric study.10 We wanted to evaluate the feasibility and clinical performance of this technique in a variety of abdominal operations when initiated preoperatively. Earlier techniques were often initiated after induction of anaesthesia or at the end of surgery. The advantages of our technique include better identification of tissue planes due to lack of postsurgical tissue oedema, the puncture point being away from the surgical site, the ability to test the block preoperatively and the suitability of the technique for both upper and lower abdominal surgery.
The TAP block is intended to provide analgesia to the parietal peritoneum, the skin and abdominal musculature. Described as recently as 2001,12 the target for this field block is the neurovascular plane transmitting the nerve supply of the anterior abdominal wall (T6 to L1) between the internal oblique and transversus abdominis muscles wherein they inter-communicate with neighbouring segments.13 An ultrasound-guided technique14 has become popular in the last decade as evidenced by the increase in the number of publications.15 Although TAP blocks have been shown to provide superior analgesia to i.v.-PCA and multimodal analgesia alone,16 comparison of TAP block as a continuous technique with TEA is lacking. Only three studies have compared TAP block catheters with TEA.5–7 In a retrospective case-matched cohort study, Kadam and Moran6 employed both mid-axillary and subcostal catheters7 and found comparable analgesia in both groups except on arrival at PACU. In their study, the failure was higher in the epidural group (27%) than in the TAP group (unilateral block in two patients), which is similar to the findings of our study. In a prospective trial comparing subcostal TAP catheters and intermittent boluses with epidural infusions for hepatobiliary and renal surgery, Niraj et al.5 found no significant difference in pain scores between 8 and 72 h postoperatively except for a higher rescue analgesic consumption in the TAP group. Although the failure rate of TEA (22%) was comparable with that of the TAP group (30%), eight out of 29 cases in the TAP group had sources of pain (drains or incisions) not affected by subcostal TAP blocks. LM-TAP catheters may obviate missed dermatomal segments due to blockade of both upper (T6 to T7) and lower (T12-L1) dermatomes. In the most recent comparative trial of continuous TAP blocks with epidural analgesia following laparoscopic colorectal surgery, TAP blocks were found to provide comparable analgesia to that of epidural analgesia.7 Their patients received either epidural analgesia or bilateral four-quadrant TAP blocks, which were intended to cover both upper and lower abdominal dermatomes, following which the lateral TAP catheters were inserted at the end of surgery. The postoperative analgesic regimen was comparable with that in our study except for the use of gabapentin.
The variability in cutaneous sensory blockade with TAP block (Table 2) seen in our study could be explained by either the presence of a larger intra and inter-individual variation in cutaneous sensory distribution17 or by the variability of local anaesthetic spread. The upper dermatomes (T7 to T8) are usually spared with midaxillary approaches, whereas the lower dermatomes (T12-L1) are spared with subcostal injections.9,10 In a cadaveric study,8 injection of 10 ml of dye in the subcostal and midaxillary TAP plane consistently blocked T8 to T11 segments (100%) and also affected T7 (14%), T12 (71%) and L1 segments (43%), which is similar to our findings. In an MRI study of injectate spread following single-injection TAP blocks,9 posterior injections of the same volume (30 ml) were insufficient to provide analgesia of the hemi-abdomen (T6 to T12), especially in the subcostal territory, whereas combined subcostal and posterior injections (of 15 ml each) covered a greater surface area and were sufficient for both upper and lower abdominal analgesia. Our technique of LM-TAP block is based on the same hypothesis as that of bilateral dual TAP blocks as described by Børglum et al.,9 but instead of the needle being advanced in a medial-to-lateral direction in bilateral dual TAP blocks, the needle is advanced from a lateral-to-medial direction starting from the midaxillary line in LM-TAP blocks. The procedural time was significantly longer in group LM-TAP (Table 1). Initiation of TAP blocks in the operating room adds to the surgical time and affects resource utilisation. The advantage of our technique is that these blocks can be initiated in a designated block room ahead of surgical time.
The secondary outcomes in our study were comparable between the TEA and LM-TAP groups except for the incidence of hypotension. The differences in opioid consumption can be explained by their route of administration, but this did not result in any increase opioid-related adverse effects in either group. If we were to compare the opioid consumption converted to i.v. equivalents, it would show a significantly higher opioid consumption in the TEA group (1 : 10 conversion ratio for epidural to i.v. opioids). One has to keep in mind that the amount of opioid delivered closer to the neuraxis may have greater implications for central nervous system adverse effects. We chose not to convert epidural to i.v. opioid equivalents due to the reported inconsistencies in opioid conversion ratios.18–20 In our experience, bilateral LM-TAP block would be unsuitable or difficult in patients with major abdominal scars, poor musculature or reduced reserve for local anaesthetic binding such as hepatic insufficiency or malnutrition. The catheter location precludes their use in subcostal incisions such as for hepatic surgery. Morbidly obese patients may pose a challenge with insertion of catheters due to the thickness of adipose tissue and the needle length needed. The postoperative period is another challenging situation for inserting TAP catheters due to tissue oedema and loss of delineation of the plane.
The noninferiority analgesia observed in our study could be due to multiple factors. A high failure rate of epidural analgesia in group TEA might have contributed to the results in our study. Another factor was the absence of systemic opioid rescue analgesia in group TEA if there was a functioning epidural after discharge from PACU, whereas the patients in group LM-TAP had systemic opioids in the form of i.v.-PCA for rescue analgesia. We wanted to evaluate the performance of the continuous LM-TAP block and TEA in a clinically representative manner in this feasibility study and the results of the trial might have been different if both groups had received systemic opioid rescue analgesia. The incidence of analgesic failure with TEA has been quoted to range between 27 and 50%.2,3,7,21 In a review of 25 000 patients seen by the acute pain service, Ready2 reported the failure of TEA to be around 32%. Most often, the reason for failure was unknown (58%), but the known causes included unilateral block (7%), leakage (7%) and dislodged (17%) or misplaced (11%) catheters.
In a systematic review of safety and efficacy of epidural analgesia, Wheatley et al.22 opined that the high failure rates with epidural analgesia may represent the ‘real-world’ performance of the technique. The evidence regarding epidural failure rates either from previous studies or from our data indicates that between one-third and half, the number of patients receiving an epidural are at risk of having no analgesic benefit. This is hard to justify for such a procedure with risks of potentially devastating complications. Apart from the lack of analgesic benefit, the incidence of hypotension is a concern among surgical colleagues. The incidence of hypotension noted in studies of TEA range between 3 and 6.8%,23,24 which is much lower than our observation of 20.8%. This may be related in part to differences in the definition of hypotension or the use of different epidural solutions in different studies, but we believe that the lack of hypotension in group LM-TAP is of interest. Although the current study was not powered to examine this outcome, lack of hypotension in the LM-TAP group may offer a major advantage over TEA and should be explored further.
We changed the TAP catheter system in the LM-TAP group after the initial eight cases because of shortcomings in the previous system such as the length of the insertion needle and the fused catheter connectors, which prevented tunnelling. Although these did not result in catheter dislodgement, the use of the catheter through an echogenic Tuohy needle (Pajunk kits) was found to be much easier. In our experience, the use of elastomeric pumps enclosed inside a bag hanging from each shoulder did not impede mobilisation.
There are some limitations to our study. The open-label nature of the study was designed to establish a proof of principle of initiating the LM-TAP blocks preoperatively. Although blinding in randomised trials ensures minimisation of observer and performance bias,25 performing two nonzero-risk procedures on each patient for the purposes of blinding might have had cost and ethical implications.26–28 We included a variety of abdominal surgical procedures in the study because we wished to compare the applicability of this technique in a variety of situations in which TEA would have been otherwise indicated. Although we did not measure plasma ropivacaine concentrations in the LM-TAP group, none of the patients had symptoms of local anaesthetic systemic toxicity (e.g. seizures, cardiovascular collapse, metallic taste, tinnitus) either immediately following blocks or in the postoperative period, probably due to the peak local anaesthetic concentration occurring around 35 min after the block,10,29 coinciding with general anaesthesia. In studies evaluating plasma ropivacaine concentration following continuous TAP infusions, 7 ml of ropivacaine 0.2% bilaterally resulted in total plasma concentrations as high as 2.2 μg ml−1 in 20% of the individuals,30,31 but the unbound concentrations were always less than 0.15 μg ml−1. The final clinical effect is probably determined by the interplay between the rate of rise of systemic concentration of local anaesthetic, the state of hepatic metabolism (CYP4501A activity) and the increase in the concentration of α1-acid glycoprotein postoperatively.
In conclusion, continuous LM-TAP blocks can be a viable alternative to TEA in a variety of abdominal operations and can be initiated preoperatively. Further randomised, double-blind studies are required to determine the role of this technique in pain management following abdominal surgery.
Acknowledgements relating to this article
Assistance with the study: we would like to thank Dr B Taylor, Dr C Rajgopal, Dr W Wall, Dr P Colquhoun, Dr W Davies and Dr V McAlister who allowed access to their patients to take part in the study.
Funding and sponsorship: none.
Conflict of interest: none.
1. Rawal N. Epidural technique for postoperative pain: gold standard no more? Reg Anesth Pain Med
2. Ready LB. Acute pain: lessons learned from 25,000 patients. Reg Anesth Pain Med
3. Varadhan KK, Neal KR, Dejong CHC, et al. The enhanced recovery after surgery (ERAS) pathway for patients undergoing major elective open colorectal surgery: a meta-analysis of randomized controlled trials. Clin Nutr
4. Wu Y, Liu F, Tang H, et al. The analgesic efficacy of subcostal transverses abdominis plane block compared with thoracic epidural analgesia and intravenous opioid analgesia after radical gastrectomy. Anesth Analg
5. Niraj G, Kelkar A, Jeyapalan I, et al. Comparison of analgesic efficacy of subcostal transverses abdominis plane blocks with epidural analgesia following upper abdominal surgery. Anaesthesia
6. Kadam VR, Moran JL. Epidural infusions versus transverses abdominis plane (TAP) block infusions: retrospective study. J Anesth
7. Niraj G1, Kelkar A, Hart E, et al. Comparison of analgesic efficacy of four-quadrant transversus abdominis plane (TAP) block and continuous posterior TAP analgesia with epidural analgesia in patients undergoing laparoscopic colorectal surgery: an open-label, randomised, noninferiority trial. Anaesthesia
8. Murouchi T, Yamauchi M, Gi E, et al. Ultrasound-guided subcostal and mid-axillary transversus abdominis plane block: a cadaveric study of the spread of injectate. Masui
9. Børglum J, Jensen K, Christensen AF, et al. Distribution patterns, dermatomal anesthesia, and ropivacaine serum concentrations after bilateral dual transversus abdominis plane block. Reg Anesth Pain Med
10. Sondekoppam RV, Brookes J, Morris L, et al. Injectate spread following ultrasound-guided lateral to medial approach for dual transversus abdominis plane blocks. Acta Anaesthesiol Scand
11. Werawatganon T, Charuluxanun S. Patient controlled intravenous opioid analgesia versus continuous epidural analgesia for pain after intra-abdominal surgery. Cochrane Database Syst Rev
12. Rafi A. Abdominal field block: a new approach via the lumbar triangle. Anaesthesia
13. Rozen WM, Tran TM, Ashton MW, et al. Refining the course of the thoracolumbar nerves: a new understanding of the innervation of the anterior abdominal wall. Clin Anat
14. Hebbard P, Fujiwara Y, Shibata Y, Royse C. Ultrasound-guided transversus abdominis plane (TAP) block. Anaesth Intensive Care
15. Johns N, O’Neill S, Ventham NT, et al. Clinical effectiveness of transversus abdominis plane (TAP) block in abdominal surgery: a systematic review and meta-analysis. Colorectal Dis
16. Abdallah FW, Chan VW, Brull R. Transversus abdominis plane block: a systematic review. Reg Anesth Pain Med
17. Stoeving KB, Rothe C, Rosenstock C, Aasvang EK, Lundstroem LH, Lange KH. Cutaneous sensory block area after ultrasound guided transversus abdominis plane block in volunteers [Abstract number A3011]. American Society of Anesthesiologists’ annual meeting 12-16th October 2013; San Francisco, USA
18. Syrmis W, Good P, Wootton J, Spurling G. Opioid conversion ratios used in palliative care: is there an Australian consensus? Intern Med J
19. Haffey F, Brady RR, Maxwell S. A comparison of the reliability of smartphone apps for opioid conversion. Drug Saf
20. Mercadante S, Caraceni A. Conversion ratios for opioid switching in the treatment of cancer pain: a systematic review. Palliat Med
21. Andersen G, Rasmussen H, Rosenstock C, et al. Postoperative pain control by epidural analgesia after transabdominal surgery. Efficacy and problems encountered in daily routine. Acta Anaesthesiol Scand
22. Wheatley RG, Schug SA, Watson D. Safety and efficacy of postoperative epidural analgesia. Br J Anaesth
23. Tsui SL, Irwin MG, Wong CM, et al. An audit of the safety of an acute pain service. Anaesthesia
24. 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
25. Hróbjartsson A, Boutron I. Blinding in randomized clinical trials: imposed impartiality. Clin Pharmacol Ther
26. Rothman KJ, Michels KB. The continuing unethical use of placebo controls. N Engl J Med
27. Temple RT, Ellenberg SS. Placebo-controlled trials and active-control trials in the evaluation of new treatments. Part 1: Ethical and scientific issues. Ann Intern Med
28. Simon R. Are placebo-controlled trials ethical or needed when alternative treatment exists? Ann Intern Med
29. Griffiths JD, Le NV, Grant S, et al. Symptomatic local anaesthetic toxicity and plasma ropivacaine concentrations after transversus abdominis plane block for Caesarean section. Br J Anaesth
30. Hessian EC, Evans BE, Woods JA, et al. Plasma ropivacaine concentrations during bilateral transversus abdominis plane infusions. Br J Anaesth
31. Knudsen K, Beckman Suurkula M, Blomberg S, et al. Central nervous and cardiovascular effects of i.v. infusions of ropivacaine, bupivacaine and placebo in volunteers. Br J Anaesth