Consistent evidence from a number of sources. The primary studies reported in the high quality systematic reviews were of moderate quality.

Transcription

1 Injection: Epidural Steroid with/without Anaesthetic (Caudal / Sacral) 1. Volume of evidence Comment here on any issues concerning the quantity of evidence available on this topic and its methodological quality. One guideline document (Manchikanti 2003),one Health Technology Assessment (McQuay 1997) and five systematic reviews (Rozenberg 1999, Koes 1999, Koes 1995, van Tulder 1997, Watts 1995) assessed trials that included patients with sciatica or radicular pain or low back pain treated with caudal/sacral epidural steroid. One systematic review (Watts 1995) reported a pooled analysis of results for these patients separately to those treated with lumbar epidural steroid therapy. Another review (Koes 1995) was subsequently updated (Koes 1999); only the updated report was considered. There was considerable duplication of the reported trials a well as variation in review quality. One Health Technology Assessment (McQuay 1997), and two systematic reviews (van Tulder 1997, Watts 1995) were of a high quality; between them these studies reported on five (Beliveau 1971, Breivik 1976, Bush 1991, Mathews 1987, Swerdlow 1970) 1 of six randomized controlled trials assessed across all the systematic reviews. The remaining reviews were of low quality (Rosenberg 1999, Koes 1999). Two of these reviews (Koes 1999, Rosenberg 1999) reported on the remaining RCT (Yates 1978) 2, thus assessments of six RCTs reported in one HTA and four systematic reviews (McQuay 1997, Watts 1995, van Tulder 1997, Koes 1999, Rosenberg 1999) were used to inform the current review of effectiveness. The single guideline document (Manchikanti 2003) included evidence from nine RCTs, three prospective studies and four retrospective studies. There was no common outcome analysis and the report was of low quality. Three additional RCTs and one non-randomised comparative study (Manchikanti 2001, McGregor 2001, Meadeb 2001, Revel 1996) and eight case series (Delport 2004, Manchikanti 2004, Southern 2003, Manchikanti 2002, Waldman 1998, Bush 1992, Morelle1987, Kim and Sadove 1975) were also identified in searches carried out for the current review. One of the RCTs (Revel 1996) did not allow an assessment of the effectiveness of epidural steroid per se and was excluded from the review of effectiveness. Five of the case series were of poor quality and were also excluded from the assessment of effectiveness but not from the safety assessment (Delport 2004, Manchikanti 2002, Southern 2003, Morelle1987, Kim and Sadove 1975) The populations enrolled in the randomised and non-randomised controlled trial included patients with acute and chronic low back pain (McGregor et al, Manchikanti), lumbosciatic pain with epidural fibrosis at the corresponding level of pain (Revel et al 1996) and post-operative sciatica with or without low back pain (Meadeb 2001). The interventions were heterogeneous across the studies with the combination of steroid and anaesthetic being compared with conservative treatment and a caudal epidural of anaesthetic and sarapin in the study by Manchikanti (2001). Meadeb (2001) examined the comparison between a forceful injection of saline, saline with prednisolone and the use of prednisolone alone. The efficacy of lumbar epidurals was compared to caudal epidurals in the study by McGregor (2001). The populations included in the eligible case series comprised patients with chronic low back pain (Manchikanti 2004), MRI confirmed lumbar herniated disc (Waldman 1998) and sciatica due to suspected lumbosacral nerve root compromise (Bush 1992) The trials were found to be of moderate quality, with the most regularly occurring problem being losses to follow up, particularly where treatment was discontinued because of the positive effect of an initial injection. In addition, in some of the trial reports it is not clear who received more than one injection and if the outcomes were substantially different for the patients who received more injections. In most of the trials the outcome assessors appear not to be blind to the treatments the patients received which may introduce bias, particularly when examining subjective outcomes such as patient satisfaction and pain relief. The case series were all of reasonable quality i.e., prospective studies of consecutive patients, study size varied from The main quality issues related to length of follow-up, losses to follow up and in one study, patient population size. 1 Bush, K., & Hillier, S. (1991). A controlled study of caudal epidural injections of triamcinolone plus procaine for the management of intractable sciatica. Spine., 16(5), Mathews, J. A., Mills, S. B., Jenkins, V. M., Grimes, S. M., Morkel, M. J., Mathews, W., et al. (1987). Back pain and sciatica: controlled trials of manipulation, traction, sclerosant and epidural injections. British Journal of Rheumatology., 26(6), Believeau, P. (1971) A comparison between epidural anaesthesia with and without corticosteroid in the treatment of sciatica. Rheum.Phys.Med.; 11: Brevivik, H. Hesla, P. Molnar, I. Lind, B. (1976) The treatment of chronic low back pain and sciatica: comparison of caudal injections of bupivacaine and methyl-prednisolone with bupivacaine followed by saline. Advance Pain Res Ther;1: Swerdlow, M. Sayle-Creer, W. (1970). A study of extradural medication in the relief of the lumbo sciatic syndrome. Anaesthesia; 25 (3): Yates, D.W. (1978). A comparison of the types of epidural injection commonly used in the treatment of low back pain and sciatica. Rheum Rehab; 17:

2 2. Consistency Comment here on the degree of consistency demonstrated by the availability of evidence. Where there are conflicting results, indicate how the group formed a judgement as to the overall direction of the evidence. Consistent evidence from a number of sources. The primary studies reported in the high quality systematic reviews were of moderate quality. 3. Applicability Comment here on the extent to which the evidence is directly applicable in the New Zealand setting. Comment here on how reasonable it is to generalise from the results of the studies used as evidence to the target population for this guideline. 4. Clinical Impact Comment here on the potential clinical impact that the intervention in question might have e.g. size of patient population; magnitude of effect; relative benefit over other management options; resource implications; balance of risk and benefit. A single guideline document (Manchikanti 2003) included evidence from nine randomised controlled trials, three prospective studies and four retrospective studies. The report concluded that there was strong evidence for short term pain relief and moderate evidence of long term relief for caudal epidural steroid injections. There was no common outcome analysis and the report was of low quality; the strength of the evidence appeared to have been overstated in this guideline report. The results of six randomised controlled trials were reported in the HTA and systematic reviews. There was considerable overlap between the trials reported in the systematic reviews and the guideline report. A meta-analysis of five of the trials (Watts 1995) showed that caudal epidural steroid therapy had an analgesic effect on sciatica compared to placebo controls. Odds ratios of 1.48 (95% confidence intervals of 0.43 to 5.09) to 6.60 (95% confidence interval of 1.31 to 33.16) were reported in four trials for short term pain relief of up to 60 days. Odds ratios ranging between 0.92 (95% confidence interval of 0.32 to 2.68) and 2.66 (95% confidence interval of 0.43 to 16.43) were reported for long term relief. The pooled odds ratio for caudal epidural steroid was 3.80 (95% confidence interval of ). McQuay (1997) used the data reported by Watts (1995) to estimate the number of patients that would need to be treated with caudal epidural steroid for one patient to obtain pain relief of 50%-75%. The number needed to treat (NNT) to obtain short term pain relief of 75% varied between 2.1 (95% confidence interval of 1.2 to 7.5) and 12 patients (95% confidence interval of 3.0 to infinity). The comparators varied considerably across these studies and included epidural injections of saline, epidural injections of anaesthetic and ligament injections of anaesthetic. Three studies reported on long term (12-52 weeks) pain relief. Numbers need to treat to obtain 50% long term pain relief was reported as 5.0 (95% confidence intervals of 1.8 to infinity) and 5.6 patients (95% confidence intervals of 3.5 to 14.0) in two studies using epidural saline as the comparator. Numbers needed to treat were not reported for the remaining trial. The other high quality review (van Tulder 1999) reporting a non-statistical analysis of the same trials as Watts (1995) and McQuay (1997) concluded that there was moderate evidence that epidural steroid injections were more effective than placebo injections in the short term for chronic low back pain but that there was no evidence that they were more effective than injections of local anaesthetic. These conclusions did not distinguish between trials using caudal and lumbar routes for epidurals and were determined using evidence from both routes. The remaining trial (Yates 1978) was reported in two lower quality reviews (Koes 1999, Rosenberg 1999). Caudal epidural steroid injections with and without anaesthetic were reported to be better than epidural injections of saline or anaesthetic alone. Two additional randomised (McGregor 2001, Meadeb 2001, Revel 1996) and one non-randomised (Manchikanti 2001) controlled trial of modest quality were included in the assessment of effectiveness. McGregor (2001) found that epidural steroid injections were of little to no benefit in a 45-patient pilot study regardless of whether the injections were given by the caudal route or the lumbar route, although no statistical test results were given to confirm this. Although there were no significant changes reported from baseline to six weeks of follow up there were slight improvements in Oswestry disability scores and self-rated health status as measured by the EuroQoL instrument for both groups. However by six months scores for both outcomes had reverted to pre-treatment levels. In a second trial Meadeb (2001) demonstrated a non-significant effect for forceful epidural steroid injection compared to saline in the improvement of post-disectomy sciatica four months after surgery. In the non-randomised study by Manchikanti (2001) significant differences were found from baseline to follow up for the two groups who received epidural injections of either anaesthetic and Sarapin or anaesthetic and betamethasone, for the outcomes of narcotic intake, overall health status (pain, physical health, mental health, and functional status) and for employment status. However, this was not a randomised study and there was no presentation of statistical tests examining differences between the steroid and sarapin groups at follow up which limits the extent to which conclusions can be drawn about the effectiveness of the different combinations of medications given in this study. Moreover, it is not clear exactly when during the period of follow up these measurements were taken. In addition, only 65 of a potential 200 patients were included 2

3 in the study and it was unclear why 40 patients were excluded. Considered Judgement Form - Injection: Epidural Steroid with/without Anaesthetic (Caudal/Sacral) In the trial by Revel (1996) where forceful injection of methyprednisolone and saline were compared with methylprednisolone alone, at six months post-treatment there were significant differences favouring the forceful injection group in the outcomes of patient-rated evaluation of back and nerve root pain, Schober s index. There was also a significant result favouring the forceful injection group at 18 months for visual analogue pain scores for back pain. There were no differences between the groups for finger-to-floor tests or straight leg raising tests, use of analgesics and return to work and leisure activities. The effect was suggested by the authors to be indicative of better diffusion of the steroid and saline in the areas of impingement rather than to a mechanical effect. The authors acknowledged that the sample size of their study was small and they highlighted strict inclusion criteria as a potential cause for this. This was a relatively well conducted albeit small underpowered study. There were a large number of patients who discontinued treatment in the treatment group, nearly half of which was because of treatment efficacy, and it is difficult to know if these patients were included in the analysis at 18 months. This study didn t assess the effectiveness of epidural steroid per se. Three case series of reasonable quality were identified from the additional searches (Manchikanti 2004, Waldman 1998, Bush 1992). In an evaluation of caudal epidural injections with and without fluoroscopic guidance (Manchikanti 2004) 90% of patients reported 50% or greater pain relief at 72 hours. The authors recommended that caudal epidural steroid injections should be performed with fluoroscopic guidance. A smaller case series (Waldman 1998) of 53 patients with lumbar hernatiated disc reported that 93% of patients reported an average decrease in pain (visual and verbal pain scores) of 63% after six weeks with 83% of patients experiencing a 67% reduction at three months. Bush (1992) reported that patients presenting with sciatica receiving on average three injections of caudal epidural steroids and anaesthetic showed a high degree of resolution at one year and that these patients were significantly younger and had a shorter duration of symptoms (less than one year before presentation) than those who had no resolution of symptoms. Each of thee studies had methodological limitations that could result in an overestimation of effectiveness. Summary of evidence for effectiveness There is moderate evidence that caudal epidural steroid injections are more effective than placebo injections in the short term for chronic low back pain and some evidence to suggest that a small proportion of patients may receive long term modest pain relief. There was no evidence that epidural steroid injections were more effective than epidural injections of local anaesthetic. A number of studies compared caudal and lumbar approached, this was not looked a further for this current document. Adverse Events/Complications Nineteen studies, one guideline, one HTA, four systematic reviews, three additional controlled trials, eight case series and two safety reports (one case report and one retrospective case series) were evaluated for safety. The single HTA and four systematic reviews did not report on safety in this group specifically, where reported adverse events and complications were pooled for all of the assessed trials which included epidural steroid delivered by the lumbar and caudal rotes. The guideline report did not report on safety. Randomised controlled trials Of the three additional RCTs and one non-randomised comparative study three reported on adverse reactions / complications. Manchikanti 2001 indicated that no complications were reported during the study; Meadeb 2001, and Revel 1996 both ereported greater pain associated with theforceful injection than no forceful. Observational studies Three of seven cases series reported on safety. In the case series by Manchikanti et al (2004) there were no complications resulting from the epidural injection procedure in 76% of patients. The most common complaint was soreness at the injection site which varied significantly with the number of attempts to inject at the site. No complications were reported in the case series reported by Waldman (1998) and Bush (1992) although a few patients felt faint after the procedure in the latter study. Other safety reports Botwin (2001) reported a retrospective cohort design study in which chart reviews were performed on 139 patients who presented with radiculopathy and received 257 fluoroscopically-guided caudal epidural steroid injections. No major complications occurred; minor complications included the following: 12 episodes of insomnia the night before the injection 9 transient non-positional headaches that resolved within 24 hours 8 increased pack pain 6 facial flushing 2 vasovagal reactions 2 episodes of nausea 1 increased leg pain. No dural punctures occurred. All reactions resolved without morbidity and no patient required hospitalization. 3

4 A case report (Yue 2003) of a 73 year old female with mild diabetes reported distant skip level discitis and vertebral osteomyelitis one month after caudal epidural injection of steroid and anaesthetic Intravenous ciprofloxacin and gentamicin followed by oral ciprofloxacin was administered and at five months the resolution of infection was confirmed. Pooled safety reports from the HTA and systematic reviews The systematic reviews which included trials of caudal epidural steroid injection did not distinguish between routes of entry (lumbar or caudal) in the reporting of safety and the following comments cover epidural steroid injections in general. McQuay (1997) reported a NNT for dural tap of approx. 40 patients in the assessed studies; no other complications were reported. Koes (1999) reported only transient minor complaints such as headaches and nausea after epidural steroid injection. Watts (1995) reported pooled data from five trials (56%) that detailed adverse events. The incidence of dural taps was 2.5% (11/431), transient headaches 2.3% (10/431), transient increase in pain 1.9%, one patient (0.2%) complained of irregular periods. It is a safe procedure. Headache & low blood pressure are less likely than with the lumbar route. The exact mixture of drugs & volume of fluid injected varies. Higher doses of steroids are likely to raise blood sugar levels for a few days in those patients with diabetes, and raise blood pressure in those with high blood pressure. Rare and serious complications can occur including paraplegia as a result of bleeding or infection in the epidural space. These complications are so rare it is difficult to put an accurate figure to them but the incidence is probably in the order of one in every fifty thousand epidurals. Infection can present one to two weeks following caudal epidural steroid and is usually readily identifiable clinically with a picture of severe back pain & associated rigidity, fevers, sweats. Infection can present one to two weeks following epidural steroid & is usually readily identifiable clinically with a picture of severe back pain and/or leg pain associated with fevers, sweats, muscle spasm & general unwellness. Rapidly developing weakness or numbness in the legs or perineum must be treated as a medical emergency with prompt referral to a hospital with MRI scanning facility. This list of side effects is by no means exhaustive. Anaesthetists would argue that it is a less reliable way of getting into the epidural space than the lumbar approach & also that the concentration of drugs obtained in the lumbar area may not be high enough for a good therapeutic effect. 5. Other Factors Indicate here any other factors that you took into account when assessing the evidence base. The Meadeb study uses high dose prednisolone. 6. Evidence Statement Please summarise the development group s synthesis of the evidence relating to this key question, taking all the above factors into account, and indicate the evidence level which applies. Initial statement There is limited evidence that caudal/sacral epidural injections are effective for the treatment of radiating leg pain and or back pain. Updated May 05 There is medium quality evidence that epidural caudal/sacral injections of steroid are effective for the treatment of sciatic or radicular pain in the short term (up to 60 days). Weight and consistency of evidence* + + Evidence Level** ~ ~ 7. Recommendation What recommendation(s) does the guideline development group draw from this evidence? Please indicate the grade of recommendation(s) and any dissenting opinion within the group. Initial statements There is insufficient evidence of benefit to make a recommendation about the effectiveness of caudal/sacral epidural injection for radiating leg pain and or back pain. Good Practice Point Forceful injection at large volume into the epidural space may be associated with cranial nerve palsy and subdural heamatoma. Grade of Recommendation*** I Where there is no substantial clinical benefit it is difficult to justify more than three caudal/sacral injections. Updated May 05 We recommend that epidural caudal/sacral injections of steroid can be used for the 4

5 short-term treatment of sciatica or radicular pain. Considered Judgement Form - Injection: Epidural Steroid with/without Anaesthetic (Caudal/Sacral) B We do not recommend that epidural caudal/sacral injections of steroid can be used for the treatment of low back pain Good Practice Point Where there is no substantial clinical benefit it is difficult to justify more than three caudal/sacral injections. C *weight and consistency of evidence: + good ~ moderate - poor **Evidence level: + strong studies where all or most of the validity criteria are met - weak studies where very few of the validity criteria are met and there is a high risk of bias ~ studies where not all the of the criteria are met but the results of the study are not likely to be affected ***Grade of recommendation: A= The recommendation (course of action) is supported by good evidence The evidence consists of results from studies of strong design for answering the question addressed B = The recommendation (course of action) is supported by fair evidence The evidence consists of results from studies of strong design for answering the question addressed but there is some uncertainty attached to the conclusion either because of inconsistencies among the results from the studies or because of minor flaws; or the evidence consists of results from weaker study designs for the question addressed but the results have been confirmed in separate studies are reasonably consistent. There is fair evidence that the benefits of the course of action being proposed outweigh the harms. C = The recommendation (course of action) is supported by expert opinion only For some outcomes, trials or studies cannot be or have not been performed and practice is informed by only expert opinion I = No recommendation can be made because the evidence is insufficient Evidence for a course of action is lacking, of poor quality or conflicting and the balance of benefits and harms cannot be determined. 5