Australian Safety and Efficacy Register of New Interventional Procedures-Surgical ASERNIP-S REPORT NO. 47. January 2006

Transcription

1 ASERNIP S Australian Safety and Efficacy Register of New Interventional Procedures-Surgical Paravertebral Blocks for Anaesthesia and Analgesia: A Systematic Review ASERNIP-S REPORT NO. 47 January 2006 Australian Safety & Efficacy Register of New Interventional Procedures Surgical The Royal Australasian College of Surgeons

2

3 Paravertebral blocks for anaesthesia and analgesia: a systematic review ISBN Published January 2006 This report should be cited in the following manner: Thavaneswaran P, et al. Paravertebral Blocks for Anaesthesia and Analgesia: a systematic review. ASERNIP-S Report No. 47. Adelaide, South Australia: ASERNIP-S, January Copies of these reports can be obtained from: ASERNIP-S PO Box 553, Stepney, SA 5069 AUSTRALIA Ph: Fax: college.asernip@surgeons.org

4 The Safety and Efficacy Classification for the Systematic Review of Paravertebral Blocks for Anaesthesia and Analgesia was ratified by: The ASERNIP-S Management Committee in December 2005 The Council of the Royal Australasian College of Surgeons on January

5 Table of Contents Executive Summary...iii The ASERNIP-S Classification System...vii The ASERNIP-S Review Group...ix 1. Introduction... 1 Objective...1 Background...1 The paravertebral space...1 Classical Placement of Paravertebral Blocks...2 Other placement techniques...3 Injection of local anaesthetic...3 Spread of anaesthesia...3 Indications...4 Advantages of Paravertebral Blocks...4 Summary Methods... 5 Literature Search Protocol...5 Inclusion Criteria...5 Literature Searches Strategies...7 Databases Searched and Search Terms Used...7 Methods of the Review...7 Literature Database...7 Data Extraction...8 Data Analysis...8 Ongoing and unpublished trials Studies Included in the Review... 9 Designation of Levels of Evidence and Critical Appraisal...9 Description of studies...9 Anaesthesia...9 Analgesia Results i

6 Anaesthesia...22 Safety...22 Efficacy...31 Analgesia...43 Safety...43 Efficacy...57 Cost considerations Discussion...75 Conclusions...77 Classification and Recommendations...77 Evidence rating...77 Safety...77 Efficacy...77 Clinical and Research Recommendations...77 Acknowledgments...78 References...79 Appendix A Hierarchy Of Evidence...86 Appendix B Excluded Studies...88 Appendix C Methodological Assessment and Data Extraction Tables...93 ii

7 Executive Summary Objective The objective of this review was to make recommendations on the safety and efficacy of thoracic and lumbar paravertebral blocks on the basis of a systematic assessment of the peer-reviewed literature. Paravertebral blocks for surgical anaesthesia were compared to general anaesthesia or other regional anaesthetic techniques, while the postoperative analgesia provided by paravertebral blocks was compared to regional blocks or analgesic drugs. Methods Search strategy Studies were identified by searching MEDLINE, EMBASE, The Cochrane Library, Science Citation Index and Current Contents from inception to December The Clinical Trials Database (US), NHS Centre for Research and Dissemination, NHS Health Technology Assessment (UK), National Research Register (UK), National Institute of Health (US) and Meta Register of Controlled Trials were also searched in December Study selection Randomised controlled trials, historical and/or non-randomised comparative studies, case series and case reports in humans of at least 18 years of age were included for review. Included comparative studies concerned the comparative interventions, defined as general anaesthesia or any other method of analgesia (eg. epidural, opioids). Efficacy outcomes included surgical anaesthesia, pain scores and length of hospitalisation. Safety outcomes included complications such as pneumothorax, nausea and vomiting, local anaesthetic toxicity and urinary retention. Data collection and analysis Data from the included studies were extracted by the ASERNIP-S researcher using standardised data extraction tables developed a priori and checked by a second researcher. Relative risks with 95% confidence intervals were calculated for some outcomes in individual RCTs where it helped the interpretation of results. Results A relatively large evidence base of reasonable quality (57 studies including 15 RCTs and describing over one thousand PVB procedures) was available for this systematic review of paravertebral block. However the ability to draw firm conclusions was limited by the high number of indications (two for anaesthesia and nine for analgesia), and the diversity of outcomes and how they were measured. In addition, although nine RCTs of analgesia using PVB were located, the comparators were thoracic epidural block (3 trials), no additional intraoperative analgesia (2 trials), iii

8 morphine (1 trial), interpleural local anaesthetic (1 trial), nerve block (1 trial) and one trial compared bolus and continuous PVB. For anaesthesia, PVB seems to be a safe procedure which substantially reduces nausea and vomiting in comparison to GA, although there is a small risk of pleural and vascular punctures and epidural spread with PVB. While clearly any form of regional block will have more failures than GA, the PVB failure rate was no higher than 20% and patients were more satisfied with PVB than with GA. There was some indication that PVB could achieve shorter hospital stays than GA, but this was poorly reported in the studies. For analgesia, PVB appears to result in about the same degree of effective block as other forms of regional analgesia. The results for pain relief and nausea and vomiting were not as clear, due to limited evidence, however PVB appeared to be as effective and safe as the comparators. As for anaesthesia, there is a small risk of punctures and epidural spread, which would increase if multiple PVB procedures were required (for example, in treating chronic pain). There was little information about the technical difficulty or learning curve for PVB, and no information was available which compared the costs of PVB with GA for anaesthesia, or PVB with local analgesia. However a very small amount of cost data from two studies indicated that between about A$ 500 to 1,000 could be saved by avoiding an overnight stay after PVB. Classification and Recommendations On the basis of the evidence presented in this systematic review, The ASERNIP-S Review Group agreed on the following classifications and recommendations concerning the safety and efficacy of paravertebral block for anaesthesia and analgesia: Classifications Evidence rating The evidence-base in this review is rated as average. Safety Paravertebral blocks at the level of the thoracic and lumbar vertebrae are at least as safe as (1) general anaesthesia and other regional anaesthetic techniques for anaesthesia during surgery, and (2) analgesic drugs and other regional blocks for analgesia postoperatively. iv

9 Efficacy Paravertebral blocks at the level of the thoracic and lumbar vertebrae are at least as effective as (1) general anaesthesia and other regional anaesthetic techniques for anaesthesia during surgery, and (2) analgesic drugs and other regional blocks for analgesia postoperatively. Clinical and Research Recommendations PVB is an advanced technique, which requires a degree of expertise and competence. It is recommended that anaesthetists wishing to use the PVB technique undergo appropriate training and supervised instruction until competent, and there should be ongoing audit of their performance. Additional high quality, prospective randomised controlled trials would strengthen the evidence base for the PVB technique. Cost-effectiveness studies, taking into consideration the Australian healthcare context should also be considered. v

10 Important note The information contained in this report is a distillation of the best available evidence located at the time the searches were completed as stated in the protocol. Please consult with your medical practitioner if you have further questions relating to the information provided, as the clinical context may vary from patient to patient. vi

11 The ASERNIP-S Classification System Evidence Rating The evidence for ASERNIP-S systematic reviews is classified as Good, Average or Poor, based on the quality and availability of this evidence. High quality evidence is defined here as having a low risk of bias and no other significant flaws. While high quality randomised controlled trials are regarded as the best kind of evidence for comparing interventions, it may not be practical or ethical to undertake them for some surgical procedures, or the relevant randomised controlled trials may not yet have been carried out. This means that it may not be possible for the evidence on some procedures to be classified as good. Good Most of the evidence is from a high quality systematic review of all relevant randomised trials or from at least one high quality randomised controlled trial of sufficient power. The component studies should show consistent results, the differences between the interventions being compared should be large enough to be important, and the results should be precise with minimal uncertainty. Average Most of the evidence is from high quality quasi-randomised controlled trials, or from non-randomised comparative studies without significant flaws, such as large losses to follow-up and obvious baseline differences between the comparison groups. There is a greater risk of bias, confounding and chance relationships compared to high-quality randomised controlled trials, but there is still a moderate probability that the relationships are causal. An inconclusive systematic review based on small randomised controlled trials that lack the power to detect a difference between interventions and randomized controlled trials of moderate or uncertain quality may attract a rating of average. Poor Most of the evidence is from case series, or studies of the above designs with significant flaws or a high risk of bias. A poor rating may also be given if there is insufficient evidence. vii

12 Safety and Efficacy Classification Safety At least as safe compared to comparator * procedure(s) This grading is based on the systematic review showing that the new intervention is at least as safe as the comparator. Safety cannot be determined This grading is given if the evidence is insufficient to determine the safety of the new intervention. Less safe compared to comparator * procedure(s) This grading is based on the systematic review showing that the new intervention is not as safe as the comparator. Efficacy At least as efficacious compared to comparator * procedure(s) This grading is based on the systematic review showing that the new intervention is at least as efficacious as the comparator. Efficacy cannot be determined This grading is given if the evidence is insufficient to determine the efficacy of the new intervention. Less efficacious compared to comparator * procedure(s) This grading is based on the systematic review showing that the new intervention is not as efficacious as the comparator. Research Recommendations It may be recommended that an audit or a controlled (ideally randomised) clinical trial be undertaken in order to strengthen the evidence base. Clinical Recommendations Additional recommendations for use of the new intervention in clinical practice may be provided to ensure appropriate use of the procedure by sufficiently qualified/ experienced centres and on specific patient types (where appropriate). * A comparator may be the current gold standard procedure, and alternative procedure, a nonsurgical procedure or no treatment (natural history) viii

13 The ASERNIP-S Review Group ASERNIP-S Director Professor Guy Maddern ASERNIP-S Royal Australasian College of Surgeons Stepney SA 5069 Advisory Surgeon Mr Rodney Cooter 1 st Floor, Waverly House 360 South Terrace Adelaide SA 5000 Advisory Anaesthetists Professor Donald Moyes Queen Elizabeth Hospital Woodville Road Woodville SA 5011 Dr Glenda Rudkin 18 North Terrace Adelaide SA 5000 ASERNIP-S Researcher Ms Prema Thavaneswaran Ms Philippa Middleton ASERNIP-S Royal Australasian College of Surgeons Stepney SA 5069 ix

14

15 1. Introduction Objective The objective of this review was to make recommendations on the safety and efficacy of thoracic and lumbar paravertebral blocks on the basis of a systematic assessment of the peer-reviewed literature. Paravertebral blocks for surgical anaesthesia were compared to general anaesthesia or other regional anaesthetic techniques, while the postoperative analgesia provided by paravertebral blocks was compared to regional blocks or analgesic drugs. Background The paravertebral space was defined in the early 20th century as the anatomical area lateral to the vertebral column (Lonnqvist 2001). Injection of local anaesthetic solutions into this space was used initially to map the pain pathways of the thoracic and abdominal viscera. Later paravertebral blocks were recognised as a method of producing unilateral analgesia of the trunk without major physiological derangements, and were frequently used in the 1920s and 1930s. After this time paravertebral blocks were not commonly used, until a publication in 1979 describing the benefits of the technique brought renewed interest in the method (Eason and Wyatt 1979). Paravertebral blocks can be used either alone for intraoperative anaesthesia with sedation, or combined with general anaesthesia where its use is primarily for postoperative analgesia. The block may have particular application in a day surgery setting as an alternative to general anaesthesia. The blocks may be classified as cervical, thoracic or lumbar, depending on the vertebral region where the block is performed. This review will be restricted to thoracic and lumbar (to L2) paravertebral blocks. The paravertebral space The thoracic paravertebral space is a wedge shaped area which is bounded posteriorly by the superior costotransverse ligament, anteriorly by the parietal pleura, medially by the vertebrae, intervertebral disc and intervertebral foramen, and laterally the apex extends into the intercostal space. The thoracic paravertebral space is continuous: laterally with the intercostal space; medially with the epidural space via the intervertebral foramen; and with the ipsilateral paravertebral space above and below (Karmakar 2001) 1

16 The contents of the paravertebral space consist of fatty tissue containing the spinal nerve, the dorsal ramus, the rami communicantes, and the sympathetic chain (anteriorly). Within the space the spinal nerves are segmented into small bundles lying freely among the fat and without a fascial sheath, making them highly sensitive to local anaesthesia (Nunn and Slavin 1980). Classical Placement of Paravertebral Blocks The paravertebral block is performed with the patient in a sitting, lateral or prone position. The positioning of the needle for a paravertebral block is illustrated in Figure 1. Figure 1 Placement of the needle for a thoracic paravertebral block (Karmakar 2001) The classical technique for needle placement involves a loss of resistance (Karmakar 2001). A needle is inserted 2.5 cm lateral to the most cephalad aspect of the spinous process, and then advanced perpendicularly to the skin in all planes, to contact the transverse process of the vertebra below. The depth will vary from two to four cm, depending on the build of the patient. If bone is not encountered the needle tip may be lying between adjacent transverse processes, and the needle should be withdrawn in the subcutaneous plane and redirected cephalad or caudad to the same depth until 2 SECTION 1 INTRODUCTION

17 bone is encountered. If this is not done, inadvertently deep insertion with pleural puncture may occur. After contacting the transverse process, the needle is walked above the process and gradually advanced until a loss of resistance to air or saline or a subtle pop is felt as the needle tip transverses the thin superior costotransverse ligament. This is usually within one to 1.5 cm from the superior edge of the transverse process. The syringe should then be gently aspirated before the local anaesthetic is injected, either in aliquots or via a catheter. Other placement techniques The classical technique for needle placement may be difficult for some operators, as the perception of loss of resistance when the needle is in the correct place may be subtle. Alternative methods include: advancement of the needle a fixed predetermined distance once the needle is walked off the transverse process (Karmakar 2001); fluoroscopy, contrast chest radiography or ultrasound to confirm the correct placement of the needle; a pressure measurement technique using the sudden lowering of pressure and expiratory pressure exceeding inspiratory pressure in the paravertebral space (Richardson 1996); nerve stimulation (Naja 2003); and a medial approach, with the needle inserted one cm from the midline and advance perpendicularly to contact the lamina, rather than the transverse process, followed by lateral redirection to slip off the lamina and into the thoracic paravertebral space (Karmakar 2001). Injection of local anaesthetic The local anaesthetic can be injected either using a single shot technique (15 to 20 ml) or multiple injections (5 to 7 ml) at each dermatome level. Spread of anaesthesia At least four intercostal spaces may be covered by a single 15 ml injection of 0.375% bupivacaine (Eason and Wyatt 1979). In a separate study a volume of 15 ml bupivacaine (0.5%) injected into the thoracic paravertebral space produced unilateral somatic nerve (voluntary) block over five (range one to nine) dermatomes, and sympathetic nerve block over eight (range six to ten) dermatomes (Cheema 1995). 3

18 Indications Regional anaesthesia resulting from thoracic paravertebral blocks is suitable for breast and abdominal surgery, and may also be used for surgery involving the thoracic cavity. As a method of postoperative analgesia, the blocks may be useful following thoracic surgery, breast surgery, cholecystectomy, renal and ureteric surgery, herniorrhaphy, appendicectomy, and video assisted or minimally invasive cardiac surgery (Karmakar 2001). Advantages of Paravertebral Blocks The resurgence of interest in paravertebral blocks may be related to the following factors: 1. Technical simplicity compared to a thoracic epidural technique (Karmakar 2001). 2. Paravertebral blocks result in a higher intensity of sensory block versus epidural analgesia (Dahl 1990; Richardson 1998). 3. The limited number of intensive care beds available in most developed countries and the need for anaesthetic techniques reducing the need for these beds (Lonnqvist 2001). 4. Fewer side effects than patient controlled opioids or epidural analgesia (Lonnqvist 2001). 5. An increase in operations performed as day surgery, with the need for anaesthetic techniques to provide postoperative pain relief and reduced nausea and vomiting (Lonnqvist 2001). 6. Recent descriptions of a single paravertebral injection providing cover over a number of segments, and blockade for up to five postoperative days with the use of a catheter (Richardson 1994). Summary Paravertebral nerve blocks are being used increasingly for operative anaesthesia and postoperative analgesia. While these blocks are relatively simple to learn and, applied properly, may have fewer side effects than comparators, including general anaesthesia, they may result in potentially serious complications, such as pneumothorax or epidural spread. Furthermore the efficacy of paravertebral blocks in producing surgical anaesthesia and/or analgesia does not seem to have been systematically evaluated. 4 SECTION 1 INTRODUCTION

19 2. Methods Literature Search Protocol Inclusion Criteria Papers were selected for inclusion in this systematic review on the basis of the following criteria: Participants Studies included human individuals of at least 18 years of age. New Intervention Included studies concerned thoracic or lumbar paravertebral blocks. Lumbar blocks only up to L2 were included; blocks lower than L2 overlap with lumbar plexus and psoas compartment blocks and involve different techniques. The paravertebral blocks were performed primarily for either anaesthesia or analgesia. Studies describing any form of acceptable placement of the block were included, excepting studies describing intrathoracic placement which were excluded. Studies describing paravertebral blocks performed in the cervical region were also excluded. Comparative Intervention Included comparative studies concerned the comparative interventions, defined as general anaesthesia or any other method of analgesia (eg. epidural, opioids). Outcomes The studies included contained information on at least one of the following outcomes of the new or comparative intervention. Safety outcomes, including: mortality bleeding pneumothorax nausea and vomiting spinal anaesthesia hypotension urinary retention respiratory depression local anaesthesia toxicity any other adverse events SECTION 2 METHODS 5

20 Efficacy outcomes, including: surgical anaesthesia number of dermatomes blocked pain scores length of surgery length of hospitalisation any other relevant efficacy outcomes Types of studies Randomised controlled trials, historical and/or non-randomised comparative studies, case series and case reports were included for review. Other study types were included if they were considered relevant and if valid reasons were given in the protocol. Additional relevant published material in the form of letters, conference material, commentary, editorials and abstracts were included as background information where appropriate. Language Restriction Searches were conducted without language restriction. Foreign language papers were subsequently excluded only if the findings supported those reported in well designed studies published in the English language. Exclusions The ASERNIP-S Researchers excluded references that did not meet the inclusion criteria. 6 SECTION 2 METHODS

21 Literature Searches Strategies Databases Searched and Search Terms Used We searched MEDLINE, EMBASE, The Cochrane Library, Science Citation Index and Current Contents from inception to December The Clinical Trials Database (US), NHS Centre for Research and Dissemination, NHS Health Technology Assessment (UK), National Research Register (UK), National Institute of Health (US) and Meta Register of Controlled Trials were also searched in December Search Terms The following search terms were used: 1. (paravertebral block) OR (paravertebral analgesia) OR (paravertebral an*esthesia) 2. Anesthetics, Local (MESH) or Nerve Block (MESH) 3. paravertebral 4. 2 AND OR 4 Note: * is a truncation character that retrieves all possible suffix variations of the root word e.g. surg* retrieves surgery, surgical, surgeon, etc. In Cochrane the truncation character is *; in Current Contents, Embase and Medline (Ovid) it is $. # is a wildcard symbol that substitutes for one required character in Current Contents, Embase and Medline (Ovid). Methods of the Review Literature Database Articles were retrieved when they were judged to possibly meet the selection criteria. Two reviewers then independently applied the selection criteria to these retrieved papers (or abstracts). Any differences were resolved by discussion with another reviewer. The number of articles retrieved for each search category is listed in Table 1. In some cases, when the full text of the article was retrieved, closer examination revealed that it did not meet the inclusion criteria specified by the review protocol. Consequently these papers were not used to formulate the evidence base for the systematic review (Appendix B). However, relevant information contained in these excluded papers was used to inform and expand the review discussion. The SECTION 2 METHODS 7

22 bibliographies of all publications retrieved were manually searched for relevant references that may have been missed in the database search (pearling). Data Extraction Data from all included studies were extracted by the ASERNIP-S researcher using standardised data extraction tables that were developed a priori. Data were only reported if it were stated in the text, tables, graphs or figures of the article, or could be accurately extrapolated from the data presented. If no data were reported for a particular outcome, in particular adverse outcomes, then no value was tabulated. This was done to avoid the bias caused by incorrectly assigning a value of zero to an outcome measurement on the basis of an unverified assumption by this reviewer. All results are tabulated in Appendices C.3 to C.6. Data Analysis Formal statistical pooling (meta-analysis) was only considered for RCTs. Other results were presented in tabular form and described. A meta-analysis was only performed if two or more RCTs addressed the same comparison, with data available for comparable outcomes. A test for statistical heterogeneity was then performed, with p<0.10 chosen to indicate the presence of statistical heterogeneity. For RCTs, relative risks (random effects model) with 95% confidence intervals were calculated for dichotomous outcomes. Ongoing and unpublished trials Crawford-Sykes (2004) reported that a randomised controlled trial is currently underway in the West Indies, which will allow a more detailed comparison between paravertebral block and general anaesthesia, however the authors were unable to obtain details of this trial from searches of Clinical Trials Database and the National Research Register. 8 SECTION 2 METHODS

23 3. Studies Included in the Review Designation of Levels of Evidence and Critical Appraisal The evidence presented in the included studies was classified according to the National Health and Medical Research Council (NHMRC) Hierarchy of Evidence (See Appendix A). Study quality was assessed according to the methods given in Chapter 6 of the Cochrane Reviewers Handbook (Alderson et al. 2004) on a number of parameters such as the quality of the reporting of study methodology, methods of randomisation and allocation concealment (for RCTs), blinding of patients or outcomes assessors, attempts made to minimise bias, sample sizes and their ability to measure true effect, applicability of results outside of the study sample as well as examining the statistical methods used to describe and evaluate the study data. The included studies are shown in Tables 1 and 2 and the study profiles are given in Appendix C.1. Several authors and/or centres have published numerous reports on their experience with paravertebral blocks for anaesthesia. As a result, there are some studies published by the same group where there are very likely to be common pools of patients. When the studies have reported the same outcome, the most recent study, or the study with the largest sample size has been used. Description of studies A total of 57 studies were included - 19 addressing PVB for anaesthesia and 38 addressing PVB for analgesia (usually for postoperative pain after general anaesthesia). Anaesthesia Breast Surgery Randomised Controlled Trials Five randomised controlled trials (Level II evidence) which utilised PVB for anaesthesia during breast surgery were included in this review (Klein 2000, Naja 2003, Terheggen 2002, Pusch 1999, El Nasr 2002) (Table 1). All five RCTs compared PVB with GA (Table 1). In relation to the important criteria of method of randomization, allocation concealment and blinding, three of the five studies provided adequate detail. Randomisation of patients was achieved using numbered sealed envelopes in three studies (Klein 2000, Naja 2003, Terheggen 2002). Two studies did not report the method of randomization (Pusch 1999, El Nasr 2002). SECTION 3 INCLUDED STUDIES 9

24 Five studies did not state whether or not allocation concealment was used (Naja 2003, El Nasr 2002, Klein 2000, Pusch 1999, Terheggen 2002). Outcome assessors were blinded in three of the studies (Klein 2000, Pusch 1999, Terheggen 2002) and in the two remaining studies blinding status was not stated (El Nasr 2002, Naja 2003). One study used power calculations to determine sample size stating to detect a 25% reduction in the primary end-point of the study (reduction of pain by VAS score) with α and β values of 0.05 and 0.10, respectively, a sample size of 30 patients in each group was necessary (Naja 2003). Another study was less clear about its use of power calculations stating that for the purposes of power determination, the two groups were taken to be of equal size, n=30 per group (Klein 2000). The remaining three studies did not state whether power calculations were used (El Nasr 2002, Pusch 1999, Terheggen 2002). All five RCTs provided adequate information regarding inclusion and exclusion criteria (El Nasr 2002, Klein 2000, Naja 2003, Pusch 1999, Terheggen 2002). The majority of studies reported no losses to followup (El Nasr 2002, Naja 2003, Pusch 1999, Terheggen 2002), and in the remaining study (Klein 2000) the loss to follow-up was minimal, 1.7% (1/60 patients). In general, sample sizes were small (15 to 44 patients in each group) and the follow-up period was short (12 hours to 5 days). Comparative studies Two non-randomised comparative studies (Level III-2 evidence), which utilised PVB for anaesthesia during breast surgery, were included in this review (Coveney 1998, Najarian 2003) (Table 1). Both studies compared PVB with GA, and had adequate reporting of methodological detail. The method of patient recruitment to each group was clearly described in both studies. Coveney (1998) stated that the anaesthetic technique used in an individual patient reflected a number of factors, including a trend for surgeon preference for paravertebral blocks during the 2 year study period, the availability of anaesthesiologists to perform the block as increasing numbers learnt the technique, as well as patient preference. Similarly, Najarian (2003) stated that the decision to perform a paravertebral block or general anaesthesia was based on surgeon and patient preferences and expertise of the anaesthesiologist performing the regional anaesthesia. Both studies clearly outlined the inclusion criteria, however only one study stated the exclusion criteria (Coveney 1998). The patients in each of the studies appeared to be well matched for age, sex and ASA classification. In general, sample sizes in both studies were large (100 to 145 patients in each group) and the follow-up period was short (length of hospital stay). Coveney (1998) reported no losses to follow-up, however Najarian (2003) did not state whether any patients were lost to follow-up. 10 SECTION 3 INCLUDED STUDIES

25 Case Series and Reports Seven Level IV studies, which utilised PVB for anaesthesia during breast surgery, were included in this review, of which four were case series and three were case reports (Buckenmaier 2003, Buckenmaier 2002, Crawford-Sykes 2004, D Ercole 1999, Greengrass 1996, Pusch 2000, Stamatiou 2004) (Table 1). A detailed analysis of methodological quality of these studies was not undertaken, although the method of patient selection was not stated in any of the seven studies. One study did not report the period of follow-up (Crawford-Sykes 2004), while the follow-up period reported in the remaining five studies ranged from the length of the hospital stay (D Ercole 1999) to 4 weeks after discharge from hospital (Buckenmaier 2002). Only three studies reported on losses to follow-up, with two studies reporting no losses (D Ercole 1999l, Buckenmaier 2003) and one study reporting a loss of 16% (4 out of 25 patients) (Greengrass 1996). Herniorrhaphy Randomised Controlled Trials One randomised controlled trial (Level II evidence), which utilised PVB for anaesthesia during herniorrhaphy was included in this review (Wassef 1998) (Table 1). This RCT compared PVB with field block (Table 1). The reporting of methodology in this RCT was inadequate. The methods of randomisation were not stated, nor were any details on allocation concealment or blinding. There were no details about sample size calculations or the length of the follow-up period. The sample size was small (15 patients in each group), however no patients were lost to follow-up. The inclusion criteria were patients scheduled for inguinal herniorrhaphy, however the exclusion criteria were not stated. Comparative studies One non-randomised comparative study, which utilised PVB for anaesthesia during herniorrhaphy, was included in this review (Naja 2002) (Table 1). This study compared PVB with GA, and had adequate reporting of methodological detail. The method of patient recruitment to each group was clearly described, with patients given a choice of either PVB or GA during their pre-anaesthesia visit. Sample size was relatively small (30 patients in each group) and was calculated based on a mean reduction of length of hospital stay from 5 to 2.5 days. The first 30 patients having ventral hernia repair and choosing either technique were included, however exclusion criteria were not stated. The patients appeared to be well matched for age, sex, weight, BMI and ASA classification. The follow-up period was short (48 hours) and there were no reported losses to follow-up. SECTION 3 INCLUDED STUDIES 11

26 Case Series and Reports Three Level IV studies, which utilised PVB for anaesthesia during herniorrhaphy, were included in this review, of which 2 were case series and one was a case report (Evans 1981, Klein 1998, Weltz 2003) (Table 1). The method of patient selection was not stated in two of the studies (Evans 1981, Klein 1998), however the remaining study reported that patients were consecutive, and were given the option of having surgery with paravertebral block anaesthesia (Weltz 2003). The follow-up period reported in the three studies ranged from 48 hours (Klein 1998) to 2 years post surgery (Evan 1981). Only one study reported on losses to follow-up, reporting a loss of 3.4% (1 out of 29 patients) at 2 hours, 6.9% (2 out of 29 patients) at 6 hours, 27.6% (8 out of 29 patients) at 12 hours and 65.5% (19 out of 29 patients) at 18 hours (Weltz 2003). 12 SECTION 3 INCLUDED STUDIES

27 Table 1: Paravertebral blocks for anaesthesia: a summary of included studies Study Study type Level Intervention N Breast Surgery El-Nasr 2002 RCT II PVB GA Klein 2000 RCT II PVB GA Naja 2003 RCT II PVB GA Pusch 1999 RCT II PVB GA Terheggen 2002 RCT II PVB GA Coveney 1998 Comparative Study III-2 PVB GA 145* 100 Najarian 2003 Comparative Study III-2 PVB GA 128** 125 Buckenmaier 2003 Case Series IV PVB 2 Crawford-Sykes 2004 Case Series IV PVB 21 Greengrass 1996 Case Series IV PVB 25 Pusch 2000 Case Series IV PVB 22 $ Buckenmaier 2002 Case Report IV PVB 1 D Ercole 1999 Case Report IV PVB 1 Stamatiou 2004 Case Report IV PVB 1 Herniorrhaphy Wassef 1998 RCT II PVB FB Naja 2002 Comparative Study III-2 PVB GA Klein 1998 Case Series IV PVB 22 Weltz 2003 Case Series IV PVB 29 # Evans 1981 Case Report IV PVB 1 PVB: paravertebral block; GA: general anaesthetic; FB: field block *156 procedures **PVB: 125 procedures GA: 100 procedures #30 procedures $ 21 procedures SECTION 3 INCLUDED STUDIES 13

28 Analgesia Breast Surgery Randomised Controlled Trials Two randomised controlled trials (Level II evidence), which utilised PVB for analgesia following breast surgery, were included in this review (Buggy and Kerin 2004, Kairaluoma 2004) (Table 2). One RCT compared PVB with intravenous morphine, while the other RCT compared a PVB injection of bupivacaine with a PVB injection of saline (Table 2). Only one of the studies provided adequate detail in their reports to determine whether minimum standards had been met for randomisation, allocation concealment and blinding. In one study, randomisation of patients was achieved by allocating patients to one of two groups using blocked randomisation from a table of random numbers (Buggy and Kerin 2004), while the other study did not report the method of randomization (Kairaluoma 2004). Allocation was concealed in one study by keeping the assigned groups in sealed, sequentially numbered envelopes (Buggy and Kerin 2004), while the other study did not state whether or not allocation concealment was used (Kairaluoma 2004). In one study the primary outcome measure was postoperative tissue oxygen tension, and the surgeon who inserted the tissue oxygen tension probe at the end of surgery was unaware of patient s group allocation (Buggy and Kerin 2004), while in the other study a staff anaesthesiologist who did not participate in the study assessment performed the blocks behind a drape curtain so that the patient, her attending anaesthesiologist and the nursing staff were blinded to the study group (Kairaluoma 2004). One study used power calculations to determine sample size stating that in order to detect a 25 mmhg increase in tissue oxygen with type I and type II error risks of 0.05 and 0.10, respectively, a sample size of 9 patients in each group was necessary (Buggy and Kerin 2004). The other study stated that a sample size estimate indicated 24 patients in each group would give a power of 80% at a level of 0.05 for detecting a difference of at least 30% in oxycodone consumption (Kairaluoma 2004). Both RCTs provided adequate information regarding inclusion and exclusion criteria and reported no losses to follow up (Buggy and Kerin 2004, Kairaluoma 2004). In general, sample sizes were small (10 to 30 patients in each group) and the follow-up period was short (20-24 hours). Cardiovascular Surgery Randomised Controlled Trials Two randomised controlled trials (Level II evidence), which utilised PVB for analgesia following cardiovascular surgery, were included in this review (Kumar and Rajendran 2003, Dhole 2001) (Table 2). Both RCTs compared PVB with thoracic epidural block (Table 2). One study stated that patients were allocated to groups by picking up the lots prior to surgery (Kumar and Rajendran 2003), while the other 14 SECTION 3 INCLUDED STUDIES

29 study did not report the method of randomisation (Dhole 2001). Neither study stated whether or not allocation concealment was used. Outcome assessors were blinded in one study (Dhole 2001), while in the other study blinding status was not stated (Kumar and Rajendran). Both studies failed to state whether power calculations were used to determine the sample sizes required to detect statistically significant differences. Adequate information regarding inclusion and exclusion criteria was provided in both studies and neither study reported any losses to followup. In general, sample sizes were small (20 to 25 patients in each group) and the follow-up period was short (up to 12 hours). Case Series and Reports Four Level IV studies, which utilised PVB for analgesia following cardiovascular surgery, were included in this review, of which two were case series and two were case reports (Canto 2003, Richardson 1995, Ganapathy 1999, Shine 2004) (Table 2). The method of patient selection was not stated in any of the four studies. The follow-up period reported ranged from 54 hours (Ganapathy 1999) to 10 days, which was the length of hospital stay (Richardson 1995). Only one study reported on losses to follow-up, reporting that no patients were lost to follow up (Canto 2003). Cholecystectomy Randomised Controlled Trials Two randomised controlled trials (Level II evidence), which utilised PVB for analgesia following cholecystectomy, were included in this review (Bigler 1989, Naja 2004) (Table 2). One RCT compared PVB in combination with GA with GA alone, and the other study compared PVB with thoracic epidural block (Table 2). In both studies, patients were randomly allocated to either group using sealed envelopes. One study did not state whether or not allocation concealment was used (Bigler 1989), while in the other study surgeons did not enter the operating room until the patient had received a general anaesthetic alone or the combined technique (Naja 2004). Outcome assessors were blinded in both of the studies. One study used power calculations to determine sample size stating to detect a 25% reduction in the primary end-point of the study (VAS score on activity 6 hours postoperatively) with α and β values of 0.05 and 0.10, respectively, a sample size of 30 patients in each group was necessary (Naja 2004), while the other study did not state whether power calculations were used (Bigler 1989). One RCT provided adequate information regarding inclusion and exclusion criteria (Bigler 1989) and the other study only stated the inclusion criteria (Naja 2004). No patients were lost to follow-up in either study. In general, sample sizes were small (10 to 30 patients in each group) and the follow-up period was short (8 hours to 3 days). SECTION 3 INCLUDED STUDIES 15

30 Hepatectomy Case Series and Reports One Level IV study, which utilised PVB for analgesia following hepatectomy, was included in this review (Ho 2004) (Table 2). The method of patient selection was not stated. The follow-up period was 9 days for one patient, and 8 days for the other patient, which was the length of hospital stay for both. Herniorrhaphy Randomised Controlled Trials One randomised controlled trial (Level II evidence), which utilised PVB for analgesia following herniorrhaphy, was included in this review (Klein 2002) (Table 2). This RCT compared PVB with a peripheral neural block (PNB) of the surgical wound (Table 2). This study provided adequate detail about randomisation of patients which was achieved using computer-generated schedules contained in sealed envelopes, however it was not stated whether or not allocation concealment was used. The study had a single blind design, and after surgery a research nurse who was unaware of the treatment group collected postoperative data. A power calculation demonstrated that in order to show a 2-hour difference in time until first pain between groups (α=0.05, power=0.8) a total of 430 patients would be necessary. Adequate information regarding both inclusion and exclusion criteria were provided in this study and no patients were lost to follow up. The sample size was small (24 and 22 patients in the PVB and PNB group respectively) and the follow-up period was short (72 hours). Comparative studies One non-randomised comparative study (Level III-2 evidence), which utilised PVB for analgesia following herniorrhaphy, was included in this review (Wood 1981) (Table 2). This study compared PVB with cryoanalgesia and oral analgesia, and had inadequate reporting of methodological detail. The method of patient recruitment to each group was described, however exclusion criteria were not stated. The patients appeared to be well matched for age, however ASA classification was not stated. In general, sample sizes were small (10 patients in each group). The follow-up period was 2 months, and 2 patients were lost to follow-up. Labour Case Series and Reports Two Level IV studies, which utilised PVB for analgesia for first stage labour pain, were included in this review, of which one was a case series and the other a case report (Nair and Henry 2001, Okutomi 2002) (Table 2). The method of patient selection was not stated in either study. The length of follow-up period and losses to 16 SECTION 3 INCLUDED STUDIES

31 follow-up were not stated in one study (Okutomi 2002), while the other study reported a follow-up period of one day, with no losses to follow-up (Nair and Henry 2001). Mixed Comparative studies One non-randomised comparative study (Level III-2 evidence), which utilised PVB for analgesia following a number of different surgical procedures, was included in this review (Catala 1993) (Table 2). This study compared thoracic PVB with endovenous morphine, and had inadequate reporting of methodological detail. The method of patient recruitment to each group was not clearly described. The patients appeared to be well matched for age and ASA classification, and in general, sample sizes were small (10 patients in each group). The length of follow-up was not stated, nor was any losses to follow-up. Case Series and Reports Four Level IV studies, which utilised PVB for analgesia following a number of different surgical procedures, were included in this review, all of which were case series (Naja 2004a, Kirvela and Antila 1992, Lonnqvist 1995, Richardson 1996) (Table 2). The method of patient selection was not stated in three studies, however one study described patient selection (Richardson 1996). Three studies failed to report the period of follow-up (Naja 2004a, Lonnqvist 1995, Richardson 1996), while the follow-up period reported in the remaining study ranged from 30 to 60 minutes (Kirvela and Antila 1992). Only one study reported on losses to follow-up, with no patients lost to follow-up (Naja 2004a). Post-traumatic pain Case Series and Reports Ten Level IV studies, which utilised PVB for analgesia for chronic post-traumatic pain, were included in this review, of which three were case series and seven were case SECTION 3 INCLUDED STUDIES 17

32 reports (Cheema 2003, Cheema 1995, Karmakar 2003, Hall and Leach 1999, Karmakar 2000, Karmakar 2001, Karmakar 2001a, Lall and Sharma 1971, Paniagua 2000, Williamson and Kumar 1997, Antila and Kirvela 1998, Richardson 1998a) (Table 2). The method of patient selection was reported in three of the ten studies, reporting that the patients were consecutive (Karmakar 2003, Antila and Kirvela 1998, Richardson 1998a). Seven studies failed to report the period of follow-up, while the follow-up period reported in the remaining five studies ranged from 2 days postoperatively (Lall and Sharma 1971) to up to 8 weeks after discharge from hospital (Cheema 2003), with none of the studies reporting on losses to follow-up. Thoracic surgery Case Series and Reports One Level IV study, which utilised PVB for analgesia following thoracic surgery, was included in this review (Richardson and Sabanathan 1995) (Table 2). The method of patient selection was not stated in this study. The length of the follow-up period was 2 months, however losses to follow-up were not stated. Thoracotomy Randomised Controlled Trials Two randomised controlled trials (Level II evidence), which utilised PVB for analgesia following thoracotomy, were included in this review (Richardson 1998, Catala 1996) (Table 2). One RCT compared a PVB employing a continuous infusion of local anaesthetic with a PVB employing a bolus injection of local anaesthetic and the other study compared a PVB with an intrapleural administration of local anaesthetic (Table 2). Only one study provided adequate detail about randomisation of patients, which was achieved using sequential allocation to computer-generated random numbers in one study (Richardson 1998). Both studies failed to state whether or not allocation concealment was used. Data collection was performed by the ward nursing staff who were unaware of the randomisation group in one study (Richardson 1998), however in the other study, blinding status was not stated (Catala 1996). Neither study stated whether power calculations were used. One study provided adequate information regarding inclusion criteria, but failed to mention the exclusion criteria (Catala 1996), while in the other study, exclusion criteria were described in detail, however inclusion criteria were not stated (Richardson 1998). One study reported no losses to follow-up (Catala 1996), and in the remaining study the loss to follow-up was 8.3% (1/12 patients). In general, sample sizes were small (6 to 15 patients in each group) and the follow-up period was short (48 hours). 18 SECTION 3 INCLUDED STUDIES

33 Case Series and Reports Two Level IV studies, which utilised PVB for analgesia following thoracotomy, were included in this review (Conacher and Kokri 1987, Thomas 1999) (Table 2). The method of patient selection was not stated in one study (Thomas 1999), while the other study stated that patients to whom the procedure had been explained agreed to be studied (Conacher and Kokri 1987). One study did not report the length of follow-up or if there were any losses to follow-up (Conacher and Kokri 1987), while the other study reported that the patient was followed-up 2 weeks after the initial surgery attempt (Thomas 1999). Volunteers Comparative Studies One non-randomised comparative study (Level III-2 evidence), which utilised PVB for analgesia in a group of healthy volunteers, was included in this review (Saito 2001) (Table 2). This study compared a thoracic PVB using lidocaine, with a thoracic PVB using a control saline solution. The subjects appeared to be well matched for age and sex, however ASA classification was not stated. In general, sample sizes were small (16 patients in each group, with each volunteer acting as his own control two to seven days later). The length of follow-up was 2 hours, and there were no losses to follow up. SECTION 3 INCLUDED STUDIES 19

34 Table 2: Paravertebral blocks for analgesia: a summary of included studies Study Study type Level Interventio N n Breast Surgery Buggy and Kerin 2004 RCT II PVB M Kairaluoma 2004 RCT II PVB SC Cardiovascular Surgery Dhole 2001 RCT II PVB TE Kumar and Rajendran 2003 RCT II PVB TE Canto 2003 Case Series IV PVB 111* Richardson 1995 Case Series IV PVB 8 Ganapathy 1999 Case Report IV PVB 1 Shine 2004 Case Report IV PVB 1 Cholecystectomy Bigler 1989 RCT II PVB 10 TE Naja 2004 RCT II PVB+GA GA Hepatectomy Ho 2004 Case Report IV PVB 2 Herniorrhaphy Klein 2002 RCT II PVB PNB Wood 1981 Comparative Study III-2 PVB CA OA Labour Nair and Henry 2001 Case Series IV PVB 4 Okutomi 2002 Case Report IV PVB 1 Mixed Catala 1993 Comparative Study III-2 PVB M Naja 2004a Case Series IV PVB 28 Kirvela and Antila 1992 Case Series IV PVB 32 # Lonnqvist 1995 Case Series IV PVB 367 Richardson 1996 Case Series IV PVB 14 Post-traumatic pain Antila and Kirvela 1998 Case Series IV PVB 7 Cheema 2003 Case Series IV PVB 29 $ Cheema 1995 Case Series IV PVB 6 Karmakar 2003 Case Series IV PVB 15 Richardson 1998a Case Series IV PVB 10 Hall and Leach 1999 Case Report IV PVB 1 Karmakar 2000 Case Report IV PVB 1 Karmakar 2001 Case Report IV PVB 2 Karmakar 2001a Case Report IV PVB 1 Lall and Sharma 1971 Case Report IV PVB 1 Paniagua 2000 Case Report IV PVB 1 Williamson and Kumar 1997 Case Report IV PVB 1 Thoracic surgery Richardson and Sabanathan 1995 Case Series IV PVB 9 Thoracotomy Catala 1996 RCT II PVBCI PVBB Richardson 1998 RCT II PVB IPL Conacher and Kokri 1987 Case Series IV PVB 5 Thomas 1999 Case Report IV PVB 1 Volunteers Saito 2001 Comparative Study III-2 PVB SC SECTION 3 INCLUDED STUDIES