Research reviews. RCTs. Research Reviews

Transcription

1 Research reviews This section is designed to give a synopsis of some of the latest research published in Medline listed journals over the last year or so. It will concentrate on controlled trials and systematic reviews, but will also include other papers that may be of interest to the readership. Some papers will be reviewed in more detail than others. If summaries and comments are based on an abstract only, this will be indicated. The main reviewer in this section is Mike Cummings, London. Other reviewers are indicated after the relevant review. Key x/5 Jadad this refers to the Jadad score, the well-known measure of internal validity of a trial. The maximum score is five; points are awarded for randomisation (up to 2), blinding (up to 2), and description of withdrawals (1). Trials scoring three or above (annotated as 3/5, 4/5 or 5/5) are considered to be of higher quality. As investigator blinding in acupuncture trials is problematic, a modification of the original description of the scoring system is used, so that subject and observer blinding is counted as double-blind. RCTs Formula acupuncture in chronic low back pain (n=131) Leibing E, Leonhardt U, Koster G, Goerlitz A, Rosenfeldt JA, Hilgers R et al. Acupuncture treatment of chronic low back pain - a randomized, blinded, placebo-controlled trial with 9-month follow-up. Pain 2002;96(1-2): One hundred and thirty-one consecutive outpatients of the Department of Orthopaedics, University Goettingen, Germany, (age=48.1 years, 58.5% female, duration of pain: 9.6 years) with non-radiating low back pain (LBP) for at least 6 months and a normal neurological examination were randomised to one of three groups over 12 weeks. Each group received active physiotherapy over 12 weeks. The control group (n=46) received no further treatment, the acupuncture group (n=40) received 20 sessions of traditional acupuncture and the sham acupuncture group (n=45) received 20 sessions of minimal acupuncture. Changes from baseline to the end of treatment and to nine-month follow-up were assessed in pain intensity and in pain disability, and secondary in psychological distress and in spine flexion, compared by intervention groups. Acupuncture was superior to the control condition (physiotherapy) regarding pain intensity (p=0.000), pain disability (p=0.000), and psychological distress (p=0.020) at the end of treatment. At the nine-month follow-up, the superiority of acupuncture compared to the control condition became less and acupuncture was not different from sham acupuncture. The authors concluded that traditional acupuncture resulted in a significant improvement in chronic LBP compared with routine care (physiotherapy) but not compared with sham acupuncture. They contend that the trial demonstrated a placebo effect of traditional acupuncture in chronic LBP. This appears to be a fairly rigorous trial in terms of methodology. It was randomised, subjects and assessors were masked, and dropouts were accounted for. Traditional acupuncture was a formula approach using 20 body points and six ear points. The body points were needled to a depth of 10 to 30mm and stimulated to achieve de qi. The points used were GV3, 4, BL23, 25, 31, 32, 40, 60, GB34, SP6, and an extra point on the back of the hand (Yautungdien). The latter refers, in some texts, to a pair of points on the back of the hand, between the second and third and the third and fourth metacarpals. It is not clear which one of these points was used. Six ear points were needled in one ear with indwelling studs and left for a week. The ear used was alternated each week. The sham group received minimal needling to nonpoints, superficial insertion and no stimulation, 10 to 20mm from the real body points. There is no report of ear needling in this group, so there may 126

2 have been potential for unmasking. This is another study to add the ever-increasing list of acupuncture studies in which needling is compared with needling, and no difference is demonstrated between the interventions. The authors conclude that their trial demonstrated a placebo effect for traditional acupuncture in PENS returns as PNT for chronic low back pain still electrical stimulation of acupuncture needles (n=72) White PF, Ghoname EA, Ahmed HE, Hamza MA, Craig WF, Vakharia AS. The effect of montage on the analgesic response to percutaneous neuromodulation therapy. Anesth Analg 2001;92(2): (Author s abstract) The analgesic response to percutaneous neuromodulation therapy (PNT) is influenced by the location, frequency, and duration of electrical stimulation. We evaluated the effect of different patterns of stimulation (montages) on the acute analgesic response to PNT when applied at the same dermatomal levels in 72 consenting patients with low back pain. All of the patients received a standardized montage (I) and three alternative montage (II-IV) patterns according to a randomized, single-blinded, crossover study design. All of the PNT treatments were administered at identical alternating stimulation frequencies of 15 and 30 Hz for a period of 30 min, three times per week for two consecutive weeks, with a one-week washout between each modality. Pretreatment assessments included the health status survey short form (SF-36) questionnaire, as well as visual analog scale scores for pain, physical activity, and quality of sleep (with 0=the best to 10=the worst). The pain visual analog scale was repeated 5-10 min after each treatment session. The daily oral analgesic usage was recorded in a patient diary. All four montages produced significant improvements in pain (42%-64%), physical activity (35%- 51%), and quality of sleep (28%-46%), as well as 23% to 47% reductions in the daily oral analgesic usage. However, Montage II was significantly more effective than the standard (Montage I) and the other two montages studied. These data suggest chronic LBP. This reviewer would prefer to conclude that a significant effect of deep and superficial needling compared with no (additional) treatment was demonstrated, but that the trial was not designed to determine the difference between the specific and non-specific effects of the needling therapies used. that the pattern of stimulation (i.e., montage) can influence the acute analgesic response to PNT. This paper continues on from the series of PENS papers that were previously reviewed in this journal (Acupunct Med 2001;19(1):32-5). The trial was partially sponsored by the White Mountain Institute (PF White, director). It is not stated whether the authors have any commercial interest in the sale of PNT equipment. There is reference to dermatomal levels in determining placement of the needles, but the latter are placed 2 to 4cm deep, and the dermatomes in the low back rarely overlap their equivalent myotomes. In each montage there are positive and negative electrode labels, but the electrical stimulus used is stated to be biphasic, so the needles would alternate in polarity during stimulation. The electrical generator used was capable of producing an output of 37mA, though during treatment the intensity of stimulation was maintained below the level required to produce muscle contraction. Currents as low as 30mA can be fatal in humans. Robust muscle contraction is produced by currents of 2 to 3mA passing between acupuncture needles in the same muscle. The pulse width used was 0.7ms. This is wider than the standard pulse width produced by most commercial electroacupuncture devices, which is generally between 0.06 to 0.5ms. The authors describe an unusually clear-cut positive effect, when the subjects in each group were treated with deep needling and electrical stimulation in the area of pain. Perhaps it is timely for this method to be verified by an independent group, to determine whether the perceived efficacy of the treatment results from the atypical characteristics of stimulation or from other factors. 127

3 Real acupuncture superior to non-point sham in tennis elbow (n=45) Fink M, Wolkenstein E, Karst M, Gehrke A. Acupuncture in chronic epicondylitis: a randomized controlled trial. Rheumatology (Oxford) 2002;41(2): (Author s abstract) Objective: to evaluate the clinical efficacy of acupuncture in the treatment of chronic lateral epicondylitis. Methods: in a randomized, investigator- and patient-blinded, controlled clinical study, 23 patients were treated with real acupuncture and 22 patients received sham acupuncture. Patients each received 10 treatments, with two treatments per week. The primary outcome variables were maximal strength, pain intensity (verbal rating scale) and disability scale (Disabilities of the Arm, Shoulder and Hand questionnaire). Patients were examined at baseline (1 week before the start of treatment) and at follow-up two weeks and two months after the end of treatment. Results: there was no significant difference between the groups at baseline for any outcome parameter. Two weeks and two months after the end of treatment, there were significant reductions in pain intensity and improvements in the function of the arm and in maximal strength in both treatment groups. At the two-week follow-up these differences were significantly greater for all outcome parameters in the group treated with real acupuncture. At two months the function of the arm was still better in this group than in the sham acupuncture group; however, the differences in pain intensity and maximal strength between the groups were no longer significant. Conclusion: in the treatment of chronic epicondylopathia lateralis humeri, acupuncture in which real acupuncture points were selected and stimulated was superior to non-specific acupuncture with respect to reduction in pain and improvement in the functioning of the arm. These changes are particularly marked at early follow-up. The real acupuncture in this trial involved needling one Ash point (presumably equivalent to ah shi ), LI11, 10, LU5, LI4 and TE5. Needling sensation (de qi) was produced and the needles retained for 25 minutes. Sham acupuncture involved needling in the same manner, but non-tender (not ah shi or trigger points) points at least 5cm away from the classical acupuncture points and their interconnecting lines (meridians). This seems reasonable until you consider that there are six meridians fairly evenly distributed around the arm, and in order to find a point 5cm from any meridian the arm would need to be in excess of 60cm in girth. This adult male reviewer s forearm is 31cm. This was put to the authors of the paper and they confirmed that this was a mistake and provided additional information. Four of the sham points were in the forearm and two in the upper arm. They particularly avoided extensor carpi radialis (presumably longus and brevis), PC6 and TE5. So how did Fink et al manage to show a difference between two groups that were needled, when Leibing et al could not, particularly as the sham acupuncture in the positive trial was deep rather than superficial needling to non-points? There are, of course, several possible reasons. The key differences probably relate to the conditions, and the proximity of the real and sham needling. Lateral epicondylalgia almost always has a local myofascial component and trigger points were avoided by at least 5cm in the sham group. In patients with chronic non-radiating back pain there may be tender or trigger points at the sites of the classical acupuncture points used in Leibing et al, and the sham points, while superficial, were within 2cm of the real points. Such superficial treatment is used in practice, 1 and has been shown to be superior to mock TENS. 2 Reference list 1. Baldry P. Management of myofascial trigger point pain. Acupunct Med 2002;20(1): Macdonald AJ, Macrae KD, Master BR, Rubin AP. Superficial acupuncture in the relief of chronic low back pain. Ann R Coll Surg Engl 1983;65(1):

4 Large study finds acupuncture ineffective for cocaine dependence (n=620) Margolin A, Kleber HD, Avants SK, Konefal J, Gawin F, Stark E, Sorensen J, Midkiff E, Wells E, Jackson TR, Bullock M, Culliton PD, Boles S, Vaughan R. Acupuncture for the treatment of cocaine addiction: a randomized controlled trial. JAMA 2002;287: Auricular acupuncture is commonly offered as a treatment for cocaine addiction, though good evidence of its effectiveness is scarce. The NADA protocol, with four or five points in each ear, is most widely used in America. This research group has considerable experience in designing trials in this area and has done pilot studies to determine the best sham treatment. They conducted a randomised, controlled, single-blind clinical trial over three years. It was based in six community clinics in the United States, three of which were hospital-affiliated clinics and three were methadone maintenance programs. They recruited 620 cocaine-dependent adult patients with a mean age of 38.8 years. Over two thirds of them were men. Four hundred and twelve used cocaine only and 208 used both opiates and cocaine and were receiving methadone maintenance. Patients were randomly assigned to receive auricular acupuncture according to the NADA protocol (n=222), or sham needling to the helix of the ear (n=203), or instructions to relax for the same period of time (n=195). Treatments were offered five times weekly for eight weeks, and concurrent drug counselling was also offered to all patients. The main outcome measures were cocaine use (measured in urine screens) during treatment and at the three and six month followup, and whether the patients stayed in treatment. Intention-to-treat analysis of urine samples showed a significant overall reduction in cocaine use (odds ratio, 1.40; 95% confidence interval, ; p=0.002). However, there were no differences between the three groups (p=0.90 for acupuncture vs. both control conditions). There were also no differences between the groups in the numbers of patients remaining in treatment, about 45% for the full eight weeks. Counselling sessions in all three groups were poorly attended. The authors concluded that, within the clinical context of this study, acupuncture was not more effective than a sham-needle insertion or relaxation control in reducing cocaine use. This study does not support the use of acupuncture as a stand-alone treatment for cocaine addiction or in contexts in which patients receive only minimal concurrent psychosocial treatment. Research will be needed to examine acupuncture s contribution to addiction treatment when provided in an ancillary role About one year earlier, the same group published a similar study, conducted at one centre in 82 patients with cocaine dependency. Acupuncture was clearly more effective than sham acupuncture or relaxation in reducing urinary cocaine levels. Now this much larger, multi-centre study yields equally clear results pointing in the opposite direction. The conclusion of the first study was slightly hesitant because more patients dropped out from the acupuncture group than the other groups, and one of the claims for acupuncture is that it helps retain patients in contact with the addiction centre. The results of the earlier study relied on only 13 patients who received acupuncture. This new study was carefully designed over many years, with pilot studies and extensive experience, in order to have high internal validity. In the discussion section, the authors list seven features that would be worth including in any study on dependence, including for example, the sample size calculation. It was designed to retain a reasonable power of 80% even if as many as 60% of patients dropped out, which is not uncommon in addition research. There were no differences detected in the amount of trial therapy or adjunctive intervention that could explain the negative result, and patients in the three groups were equally satisfied with treatment, although patients in the relaxation group scored higher relaxation after treatment than the other two groups. Certain differences did occur between the studies, such as the style of counselling and group therapy, giving financial reward for attendance at clinic rather than for abstinence, (which might have encouraged the more severely addicted, less hopeful case to reattend) and differences in 129

5 treatment context. But the most likely conclusion is that the larger study gives a more accurate estimation of the treatment effect of acupuncture, and that the previous study gave a false positive result. It does seem likely that the NADA Auricular acupuncture for pre-operative anxiety I (n=55) Wang SM, Kain ZN. Auricular acupuncture: a potential treatment for anxiety. Anesth Analg 2001;92(2): (Author s abstract) Acupuncture can be an effective treatment for chronic anxiety disorders. The purpose of this study was to assess the effectiveness of acupuncture in reducing anxiety in a volunteer population. If found effective, this modality could be introduced as a treatment of anxiety before surgery. Adult volunteers (n=55), were randomized to three treatment groups: a) Shenmen group bilateral auricular acupuncture at the shenmen point; b) Relaxation group bilateral auricular acupuncture at a relaxation point; and c) Sham group bilateral auricular acupuncture at a sham point. Press-acupuncture needles were inserted at the respective auricular areas for 48h. State anxiety, blood pressure, heart rate, and electrodermal activity were assessed at 30min, 24h, and 48h after insertion. Analyzing anxiety levels using repeated-measures analysis of variance has demonstrated a significant difference [F(2,51)=8.8, p=0.001] between the three treatment groups. Post hoc analysis demonstrated that patients in the Relaxation group were significantly less anxious at 30min (p=0.007) and 24h (p=0.035) as compared with patients in both the Shenmen group and the Sham group, and less anxious at 48h (p=0.042) as compared with patients in Shenmen Auricular acupuncture for pre-operative anxiety II (n=91) Wang SM, Peloquin C, Kain ZN. The use of auricular acupuncture to reduce preoperative treatment protocol has no specific effects on the withdrawal process. Acupuncture s role seems to be more to attract addicted patients into treatment and keeping them in touch with the services. Adrian White, Exeter group. Repeated-measures analysis of variance performed for electrodermal activity, blood pressure, and heart rate demonstrated no group differences (p=ns). We conclude that auricular acupuncture at the relaxation point can decrease the anxiety level in a population of healthy volunteers. This trial is described as blinded, but the nature of the masking is not detailed. The reader must assume that it was subject blind. The subjects had no prior experience of acupuncture and were instructed not to discuss the study with one another. The sham point used corresponded to the stomach point. This is at the end if the helix crus at the tip of the concha. Shenmen is on the superior border of the triangular fossa, one third of the way along the superior antihelix crus. The relaxation point is on the superior lateral wall of the triangular fossa at the end of the superior antihelix crus. These are three quite distinct locations, so, with press needles in place, it would not have been too difficult to distinguish the group allocation at a glance, and the subjects were all operating room staff, from the same environment, who were present in their working environment for the 48 hours of the intervention. There is no description in the paper of how allocation was masked, nor any discussion of credibility testing. Significant differences between groups were recorded in the subjective measure only (Stait-Trait Anxiety Inventory), the outcome most likely to be influenced by expectation resulting from inadvertent unmasking. anxiety. Anesth Analg 2001;93(5): This was a follow-up study to the previous one performed by Wang and Kain on operating room staff. On this occasion the subjects were

6 elective surgery patients treated pre-operatively with one of three sets of three auricular acupuncture points. Subjects were randomly allocated to a traditional Chinese medicine (TCM) group, a relaxation group, or a control group. The TCM group had press needles applied to shenmen, kidney and heart points. The relaxation group had press needles applied to relaxation, tranquilizer and master cerebral points. In the control group points were used that have no documented effect on anxiety. Press needles were applied to the ear on the side of the non-dominant hand, and left in place for 30 minutes. The Stait-Trait Anxiety Inventory was used as the only outcome measure for this trial. It was assessed prior to, and 30 minutes after, the intervention; the second assessment being performed before the needles were removed. Compared with the control group, the relaxation group were significantly less Auricular acupuncture for alcohol dependence (n=501) Bullock ML, Kiresuk TJ, Sherman RE, Lenz SK, Culliton PD, Boucher TA et al. A large randomized placebo controlled study of auricular acupuncture for alcohol dependence. J Subst Abuse Treat 2002;22(2):71-7. (Author s abstract) We report clinical data on the efficacy of acupuncture for alcohol dependence. 503 patients whose primary substance of abuse was alcohol participated in this randomized, single blind, placebo controlled trial. Patients were assigned to either specific acupuncture, nonspecific acupuncture, symptom based acupuncture or conventional treatment alone. Alcohol use was assessed, along with depression, anxiety, functional status, and preference for therapy. This article will focus on results pertaining to alcohol use. Significant improvement was shown on nearly all measures. There were few differences associated with treatment assignment and there were no treatment differences on alcohol use measures, although 49% of subjects reported acupuncture reduced their desire for alcohol. The placebo and preference for treatment measures did not materially anxious (p=0.01), however, anxiety levels in the TCM group were not significantly different from either of the other two groups. The potential for unmasking in this trial was not as great as in the previous one performed by Wang and Kain, principally because the intervention was 30 minutes as compared to 48 hours. The anatomical distribution of the points was such that only in the control group were all three points located outside the area of innervation of the vagus nerve. If the effect measured is a specific effect of the point combination used rather than a type 1 statistical error, or some form of bias, we still need to know whether the effect size is of clinical relevance. Once this has been determined, consideration can be given to testing this point combination in a larger multicentre RCT. affect the results. Generally, acupuncture was not found to make a significant contribution over and above that achieved by conventional treatment alone in reduction of alcohol use Two-point discrimination on the surface of the pinna is 15 to 20 mm. Despite this simple and reproducible finding, researchers continue to test biologically implausible hypotheses relating to specific points on the ear. In this trial three of the four groups received acupuncture to points on the ear. In two of the groups the points were deliberately placed within 5mm of each other so that the external appearance would be similar. The authors refer to the observation made by Wen and Cheung that auricular electroacupuncture (EA) reduced withdrawal symptoms in opiate addicts, 1 but they do not explain why EA was abandoned as a form of treatment in the US. I believe the authors are mistaken to state that there have been no controlled studies of EA in the treatment of substance abuse in humans. 2 Wen and Cheung worked in a neurosurgical unit, and the stimulation sites they used are likely to have been chosen for their close proximity to the operative field. The effect, if any, on withdrawal symptoms may be more related to the nature of the stimulus 131

7 (strong electroacupuncture) than to the site it was applied. This reviewer suggests that it is now time to test mode specific effects in this clinical situation, and that continuing to focus of the site of needling is probably missing the point! Bloodletting acupuncture at the thumb nail bed exacerbates post-extubation stridor (n=60) Saghaei M, Razavi S. Bloodletting acupuncture for the prevention of stridor in children after tracheal extubation: a randomised, controlled study. Anaesthesia 2001;56(10): (Author s abstract) Bloodletting acupuncture has been used for the treatment of a variety of upper respiratory tract problems, especially those of laryngeal origin. This study assesses its efficacy in reducing the incidence of stridor after tracheal extubation in children undergoing general anaesthesia with halothane. Sixty children were randomly allocated to an acupuncture group and a control group. In the acupuncture group, bloodletting acupuncture was performed before extubation. A blinded observer determined the presence and severity of postextubation stridor. The incidence of stridor in the acupuncture group was significantly higher than in the control group. In addition, the severity of stridor was significantly greater in the acupuncture group. It is concluded that in children undergoing halothane anaesthesia, Acupuncture is as effective as ondansetron for postoperative nausea and vomiting (n=90) Somri M, Vaida SJ, Sabo E, Yassain G, Gankin I, Gaitini LA. Acupuncture versus ondansetron in the prevention of postoperative vomiting. A study of children undergoing dental surgery. Anaesthesia 2001;56(10): (Author s abstract) This study compares the anti-emetic effect of acupuncture with that of ondansetron and a placebo. Ninety children Reference list 1. Wen HL, Cheung SYK. Treatment of drug addiction by acupuncture and electrical stimulation. Asian J Med 1973; Rampes H, Pereira S, Mortimer A, Manoharan S, Knowles M. Does electroacupuncture reduce craving for alcohol? A randomized controlled study. Complement Ther Med 1997;5(1): the incidence of postextubation stridor cannot be reduced by bloodletting acupuncture The first thing that occurs to this reviewer is how such a protocol got past an ethics committee. The point tested in this study was at the medial corner of the thumb nail bed. It was needled with a hypodermic needle and allowed to stop bleeding spontaneously. This procedure was performed bilaterally just prior to extubation. The study was subject and observer blind. The observer was waiting outside the operating room door. No description was given of how blinding of the observer was ensured. Presumably drops of blood were collected rather than allowing them to fall on the patient s gown, and the puncture sites, or potential puncture sites (control group), could have been covered with an adhesive plaster, for example. What was surprising about this study was how much worse the intervention group faired than the control group. The patients receiving acupuncture had an incidence of stridor of 33% as compared to 3% in the control group (p=0.006). This reviewer would hope that such a trial is not considered appropriate to repeat. undergoing dental treatment under general anaesthesia were randomly allocated to one of the three equal groups, to receive acupuncture needle insertion, intravenous ondansetron 0.15 mg per kg or a placebo. Parental satisfaction scores and the incidence of emetic episodes were recorded. A significant difference was found in the number of patients who vomited and the total number of the emetic episodes when comparing the two treatment groups with the placebo group (p<0.0001). A significant difference was also found between the treatment groups and the placebo group with 132

8 respect to parental satisfaction score (p<0.03). We conclude that traditional Chinese acupuncture is a valid non-pharmacological alternative anti-emetic treatment that can be recommended as a prophylactic technique in children undergoing dental surgery under general anaesthesia. This is a good result for acupuncture. Bilateral PC6 plus CV13 proved to be as effective as a high dose of ondansetron in the prevention of postoperative vomiting. The interventions were applied in the anaesthetic room following induction, where the subjects stayed for 20 minutes. This makes the trial somewhat unrealistic as dental EA for PONV after tonsillectomy in children (n=120) Rusy LM, Hoffman GM, Weisman SJ. Electroacupuncture prophylaxis of postoperative nausea and vomiting following pediatric tonsillectomy with or without adenoidectomy. Anesthesiology 2002;96(2): (Author s abstract) Background: electrical stimulation of acupuncture point PC6 reduces the incidence of postoperative nausea or vomiting (PONV) in adult patients. However, acupressure, laser stimulation of PC6, and acupuncture during anesthesia have not been effective for reducing PONV in the pediatric population. The authors studied the effect of electrical PC6 acupuncture in awake pediatric patients who had undergone surgery associated with a high incidence of PONV. Methods: patients aged 4-18 yr undergoing tonsillectomy with or without adenoidectomy were randomly assigned to acupuncture, sham acupuncture, or control groups. Acupuncture needles at PC6 and a neutral point were placed while patients were anesthetized, and lowfrequency electrical stimulation (4Hz) was applied to these points for 20 min in the recovery room while the patients were awake (PC6 Acu group). This treatment was compared with sham needles along the arm at acupuncture points not associated with antiemesis (sham group) and a no-needle control group. The arms were wrapped to prevent identification of treatment group, and anesthetic, surgery operating lists often have a number of short procedures on them, and it would probably not be acceptable in practice to keep a patient in the anaesthetic room for so long under these circumstances. The paper states that the random group allocation was performed before induction. This seems unnecessary, and it allows for the potential of unmasking the subjects. Allocation could have been performed immediately after induction. Despite these minor flaws, the trial scores well for internal validity (5/5 Jadad), and adds to the growing weight of evidence in favour of the routine use of acupuncture for postoperative nausea and vomiting. analgesic, and surgical technique were standardized. Assessed outcomes were occurrence of nausea, occurrence and number of episodes of vomiting, time to vomiting, and use of antiemetic rescue medication. Results: one hundred twenty patients were enrolled in the study, 40 per group. There were no differences in age, weight, sex, or opioid administration between groups. The PONV incidence was significantly lower with P6 acupuncture (25 of 40 or 63%; odds ratio, 0.135; number needed to treat, 3.3; p<0.001) compared with controls (37 of 40 or 93%). Sham puncture had no effect on PONV (35 of 40 or 88%; p=ns). Occurrence of nausea was significantly less in PC6 Acu (24 of 40 or 60%; odds ratio, 0.121; p<0.01), but not in the sham group (34 of 40 or 85%) compared with the control group (37 of 40 or 93%). Vomiting occurred in 25 of 40 or 63% in P6 Acu; 35 of 40 or 88% in the sham group, and 31 in 40 or 78% in the control group (p=ns). Patients receiving sham puncture vomited significantly earlier (p<0.02) and needed more rescue treatment (33 of 40 or 83%; odds ratio, 3.48; p<0.02) compared with PC6 Acu (23 of 40 or 58%) and the control group (24 of 40 or 60%). Conclusions: perioperative PC6 electroacupuncture in awake patients significantly reduced the occurrence of nausea compared with the sham and control groups, but it did not significantly reduce the incidence or number of episodes of emesis or the use of rescue antiemetics. Sham acupuncture may exacerbate the severity but not the incidence of emesis. The 133

9 efficacy of PC6 acupuncture for PONV prevention is similar to commonly used pharmacotherapies. Its appropriate role in prevention and treatment of PONV requires further study. At first glance this appears to be another positive study for PC6 acupuncture (or electroacupuncture in this case) in the prevention of PONV. The trial scores well for internal validity (5/5 Jadad), although it is not clear how masking was maintained when the subjects became fully conscious. Needles were placed before the patients emerged from anaesthesia and their arms were bandaged, presumably to prevent observers recognising group allocation. However, as one group was stimulated in the forearm, one in the upper arm and one not at all (wires simply attached to the insides of the arm covers), it seems possible that some of the subjects could have complained of the electroacupuncture sensation being uncomfortable and pointed to the site of Intradermal acupuncture reduces postoperative pain and PONV (n=191) Kotani N, Hashimoto H, Sato Y, Sessler DI, Yoshioka H, Kitayama M et al. Preoperative intradermal acupuncture reduces postoperative pain, nausea and vomiting, analgesic requirement, and sympathoadrenal responses. Anesthesiology 2001;95(2): (Author s abstract) Background: in a controlled and double-blind study, the authors tested the hypothesis that preoperative insertion of intradermal needles at acupoints 2.5 cm from the spinal vertebrae (bladder meridian) provide satisfactory postoperative analgesia. Methods: the authors enrolled patients scheduled for elective upper and lower abdominal surgery. Before anesthesia, patients undergoing each type of surgery were randomly assigned to one of two groups: acupuncture (n=50 and n=39 for upper and lower abdominal surgery, respectively) or control (n=48 and n=38 for upper and lower abdominal surgery, respectively). In the acupuncture group, pain, thus unmasking observers. The reader is left wondering about the intensity of stimulation that was applied. If the stimulus was adjusted with the subjects emerging from anaesthetic the perceived intensity would have been difficult to determine. Muscle twitch is unlikely to have been used as this would have been uncomfortable, and added to the potential for unmasking of observers. In the methods section of the paper we are given details of the sample size calculation. The figures were based on previous literature on rates of PONV in paediatric tonsillectomy patients, and the efficacy of PC6 in adults. The authors used an estimate of PONV of 65% and an expected reduction to 30% in the acupuncture group. The results of the trial indicate that the PONV rate was 93% in the control group. There is no explanation for the greater than expected incidence of PONV in this cohort of patients undergoing tonsillectomy. Unusually the sham acupuncture appeared to significantly exacerbate the incidence of postoperative vomiting. intradermal needles were inserted to the left and right of bladder meridian and in upper and lower abdominal surgery before induction of anesthesia, respectively. Postoperative analgesia was maintained with epidural morphine and bolus doses of intravenous morphine. Consumption of intravenous morphine was recorded. Incisional pain at rest and during coughing and deep visceral pain were recorded during recovery and for 4 days thereafter on a four-point verbal rating scale. We also evaluated time-dependent changes in plasma concentrations of cortisol and catecholamines. Results: starting from the recovery room, intradermal acupuncture increased the fraction of patients with good pain relief as compared with the control (p<0.05). Consumption of supplemental intravenous morphine was reduced 50%, and the incidence of postoperative nausea was reduced by 20-30% in the acupuncture patients who had undergone either upper or lower abdominal surgery (p<0.01). Plasma cortisol and epinephrine concentrations were reduced 30-50% in the acupuncture group during recovery and on the first postoperative day (p<0.01). Conclusion: preoperative 134

10 insertion of intradermal needles reduces postoperative pain, the analgesic requirement, and opioid-related side effects after both upper and lower abdominal surgery. Acupuncture analgesia also reduces the activation of the sympathoadrenal system that normally accompanies surgery It is hard to find fault with this study. It was well Study of acupressure for nausea is not conclusive (n=60) Werntoft E, Dykes A-K. Effect of acupressure on nausea and vomiting during pregnancy. J Reprod Med 2001;46: Although there are several studies of the effect of acupressure on nausea and vomiting of pregnancy, the evidence is not yet convincingly positive. These authors conducted a randomised, partly patient-blind clinical trial with three parallel arms. The trial took place in a Hospital ante-natal clinic. About 80 women attending ante-natal clinic who experienced nausea and vomiting of pregnancy were included if they were not using any other treatment for nausea. They were allocated to three groups by choosing an envelope from a pile with three possible contents. They opened the envelope after returning home. One group (n=20) found a wrist-band with a button (sea band) and instructions to position the button over PC6. The second group (n=20) also found a wrist-band with button, but instructions to place it on the wrong side of the wrist as a placebo. The third group (n=20) found an empty box with an explanation. All envelopes also contained an evaluation questionnaire containing a Visual Analogue Scale for nausea. The mean nausea scores for PC6 acupressure and placebo groups were no different from each other after one and three days, but were significantly lower than for the no treatment group. After 6 and 14 days, scores in the PC6 group were significantly lower than both the placebo and the no treatment groups. The authors conclude that it is possible to reduce nausea and vomiting of pregnancy significantly with designed, executed and reported (5/5 Jadad). The intradermal needles are new to this reviewer. They were 5mm long and 0.16mm diameter, with pencil point tips, and a loop to prevent the needle becoming embedded in the skin. The procedure was quick and simple, and there were no reported adverse effects. This technique certainly seems worthy of further investigation. acupressure at P6 compared to acupressure at a placebo point or no treatment at all in health women with normal pregnancy. There are some neat, inventive ideas in this trial; such as letting the patients choose an envelope and including the treatment, the instructions and the outcome measures. However, this meant losing control over the study, which suffered as a result. For example, there were dropouts, though the authors seem uncertain how many, and we are not told. The dropouts are not included in the results as the analysis is not on an intention-to-treat basis. What this study really teaches us is that, if research is going to provide clear answers, it must be rigorously conducted and reported. For example, most clinicians who use acupuncture for nausea find its effect is immediate; and some previous studies have confirmed this. Therefore, it is unsatisfactory to call the study positive when it showed no effect after one or three days of treatment, only after 6 or 14 days. This seems a kind of spin that will get acupuncture a bad name, just like it does to politics! The results showed that the placebo group had less nausea than the controls for three days, but was no different from them after the sixth day. The authors interpret this as showing a short-lived effect of placebo compared with a genuine effect of acupuncture. However, knowledge of acupressure and of the correct site to administer it is now widespread, and the findings could equally well be explained (i.e. confounded) by purely placebo effect in the acupressure group, and disappointment in the other groups at not receiving genuine treatment. This disappointment would be immediate in the no treatment group, but would be delayed for a few days in the placebo 135

11 group, until they had talked to their friends. This is not to say, of course, that acupressure is ineffective at reducing nausea of pregnancy, just that this study seems inconclusive. It may be surprising to readers that I believe that a strict review of the Acupuncture reduces nausea in early pregnancy compared to no treatment (n=593) Smith C, Crowther C, Beilby J. Acupuncture to treat nausea and vomiting in early pregnancy: a randomized controlled trial. Birth 2002;29(1):1-9. (Author s abstract) Background: nausea and vomiting in early pregnancy are troublesome symptoms for some women. We undertook a single blind randomized controlled trial to determine whether acupuncture reduced nausea, dry retching, and vomiting, and improved the health status of women in pregnancy. Methods: the trial was undertaken at a maternity teaching hospital in Adelaide, Australia, where 593 women less than 14 weeks pregnant with symptoms of nausea or vomiting were randomized into four groups: traditional acupuncture, pericardium 6 (PC6) acupuncture, sham acupuncture, or no acupuncture (control). Treatment was administered weekly for four weeks. The primary outcomes were nausea, dry retching, vomiting, and health status. Comparisons were made between groups over four consecutive weeks. Results: women receiving traditional acupuncture reported less nausea (p<0.01) throughout the trial and less dry retching (p<0.01) from the second week compared with women in the no acupuncture control group. Women who received p6 acupuncture (p<0.05) reported less nausea from the second week of the trial, and less dry retching (p<0.001) from the third week compared with women in the no acupuncture control group. Women in the sham acupuncture group (p<0.01) reported less nausea and dry retching (p<0.001) from the third week compared with women in the current evidence on acupuncture or acupressure against sham for nausea of pregnancy remains balanced, with the most rigorous study so far being negative. Adrian White, Exeter no acupuncture group. No differences in vomiting were found among the groups at any time. Conclusion: acupuncture is an effective treatment for women who experience nausea and dry retching in early pregnancy. A time-related placebo effect was found for some women. This is the largest study to date investigating the effects of acupuncture on nausea in early pregnancy. The traditional group received needling to arms (including PC6), legs and abdomen. The PC6 group received needling to this point alone, and the sham group received needling to non-points in the arms and legs, but not to the abdomen. So the traditional group were the only group to receive needling that could be considered segmental for the stomach. This is one possible explanation for why there appeared to be a more rapid response in this group compared to the no treatment group; however, none of the three groups receiving acupuncture differed significantly in terms of the primary outcomes throughout the trial. It is unfortunate that the design of this trial did not include a non-invasive sham procedure. The trial demonstrates a highly significant effect for needling, but does not determine whether the effect is specific to needle insertion or to other non-specific factors of treatment, such as expectation. The trial does provide evidence in favour of the idea that the site of needling is unimportant. The authors found a number of differences in the SF36 when analysed by domains, but these appear to be inconsistent, and as highly subjective measures are arguably more susceptible to bias introduced by an unblinded therapist. 136

12 Experimental Studies (humans) Facial TENS decreases anaesthetic requirement in simulated surgical pain (n=20) Greif R, Laciny S, Mokhtarani M, Doufas AG, Bakhshandeh M, Dorfer L et al. Transcutaneous electrical stimulation of an auricular acupuncture point decreases anesthetic requirement. Anesthesiology 2002;96(2): (Author s abstract) Background: German anesthesiologists have long used transcutaneous electrical stimulation of an acupuncture point near the tragus to reduce anesthetic requirement in unblinded and uncontrolled trials. This is known as auricular electrically stimulated analgesia. The authors therefore tested the hypothesis that auricular electrically stimulated analgesia reduces anesthetic requirement. Methods: in a randomized, double-blind, crossover trial, volunteers were anesthetized twice with desflurane. Electrical stimulation of an auricular acupuncture point in the vicinity of the tragus was used on 1 randomly assigned day, and no electrical stimulation of the same point was used on the other study day. Treatment consisted of bilateral electrical stimulation of the lateralization control point, 3 cm anterior to the tragus. The 10-mA current was set to 299 Hz on the dominant side of the face and to 149 Hz on the contralateral side. Anesthetic requirement was determined by the Dixon up-anddown method and was defined by the average desflurane concentration required to prevent purposeful movement of the extremities in response to noxious electrical stimulation. Results: ten men and 10 women completed the protocol. Electrical stimulation of the lateralization control point reduced anesthetic requirement by 11 +/- 7% (p<0.001), with the reduction being similar in women and men. Women required more desflurane to prevent movement on the control day than the men (5.5 +/- 1.0 vs /- 0.6 vol%; p=0.028). Conclusion: this double-blinded trial with an objective outcome demonstrates that electrical stimulation of the lateralization control point significantly reduces anesthetic requirement. This was a rigorous trial (5/5 Jadad) of a relatively simple technique, which has demonstrated a reduced anaesthetic requirement in the active intervention group. The points used are rather curiously described as auricular points, despite being 3cm anterior to the ear. The points were vertically aligned either side of the zygoma, the lower point being over the upper portion of the masseter muscle. The frequencies used are also odd, but probably do not differ substantially from any high frequency transcutaneous stimulus. A current of 10mA applied transcutaneous would probably produce a robust sensation in the conscious patient. Further high quality studies are needed to confirm this result in surgical patients, and consideration can then be given to introducing the technique into routine practice, if the effect size is considered to be worthwhile. 137

13 Reports of Adverse Events & Related Papers Another case of silicone granuloma (n=1) Alani RM, Busam K. Acupuncture granulomas. J Am Acad Dermatol 2001;45(6 Suppl):S225-S226. (Author s abstract) Silicone compounds have recently been a source of controversy with regard to their potential role in the genesis of collagen vascular diseases. Foreign body reactions to injectable silicone were noted early in its cosmetic use and led to subsequent abandonment of this procedure. Here we report the first documented case of silicone granulomas to occur after acupuncture. Readers of Acupuncture in Medicine may find it strange that this is reported as the first Septic arthritis of a lumbar facet joint following pyonex (n=1) Ishibe M, Inoue M, Saitou K. Septic arthritis of a lumbar facet joint due to pyonex. Arch Orthop Trauma Surg 2001;121(1-2):90-2. (Author s abstract) We present a case of septic arthritis of a lumbar facet joint with an associated epidural abscess. A 13-year-old boy was hospitalized with acute severe back pain and fever after pyonex was done. The infection was precisely localized with magnetic resonance imaging, bone and gallium scintigraphy. He responded to antibiotic therapy. We suppose that documented case of silicone granulomas to occur after acupuncture, since a case was reported in 2000, 1 and reviewed in this journal in the same year (Acupunct Med 2000;18(2):143). This is the second case ever reported in a human (there has been one other similar report in horses), 2 so it would be premature to alter practice at this stage. It may not be appropriate to market your needles based on such rare adverse events. Reference list 1. Yanagihara M, Fujii T, Wakamatu N, Ishizaki H, Takehara T, Nawate K. Silicone granuloma on the entry points of acupuncture, venepuncture and surgical needles. J Cutan Pathol 2000;27(6): Slovis NM, Watson JL, Affolter VK, Stannard AA. Injection site eosinophilic granulomas and collagenolysis in 3 horses. J Vet.Intern Med 1999;13(6): the infection was caused by pyonex because the blood cultures were negative, and the patient had an abrupt onset of severe pain and fever 24 h after the acupuncture. This is a rather unsatisfactory report from the acupuncturist s perspective because of the lack of information regarding the acupuncture used. The boy had had dull pain in the lumbar spine one month previously. Two days prior to admission a pyonex was done to his right paravertebral muscles The stitch technique was used. This is all the information that is provided about the acupuncture. The authors supposed that the infection was caused by a contaminated pyonex needle. 138

14 Basal cell carcinoma of the earlobe after auricular acupuncture (n=1) Brouard M, Kaya G, Vecchietti G, Chavaz P, Harms M. Basal cell carcinoma of the earlobe after auricular acupuncture. Dermatology 2002;204(2): (Author s abstract) The genesis of familial and sporadic basal cell carcinomas involves activation of the Sonic hedgehog signal transduction pathway. Other known factors for the development of basal cell carcinoma are ultraviolet exposure, X-rays, race, age, gender and decreased DNA repair capacity. We here report the case of a right earlobe sporadic basal cell carcinoma in a 65- year-old woman. This case is unusual because of its earlobe localization and its association with multiple auricular acupuncture treatments. This observation suggests a connection between local Bilateral pneumothorax after acupuncture (n=1) Kao CL, Chang JP. Bilateral pneumothorax after acupuncture. J Emerg Med 2002;22(1): (Author s abstract) A 28-year-old woman who did not smoke and had no history of pulmonary disease was admitted to our hospital because of bilateral pneumothorax. She had been receiving acupuncture therapy twice a week to the upper thoracic bilateral paraspinal regions because of chronic back pain. Because of no chest pain or respiratory distress, she was managed with oxygen traumas, which occur in the course of acupuncture treatment and ear piercing, and the genesis of basal cell carcinoma. The incidence of minor adverse events associated with acupuncture is high, but serious events are uncommon. Acupuncture is not known to date for promoting the development of tumors. This connection remains to be elucidated. This is an interesting and well-reported case. The patient had had auricular acupuncture performed twice a year, using at least five needles to each ear, for the previous 12 years. There was no X-ray or arsenic exposure. She had noticed some lesions that healed spontaneously on the sun exposed surface of the right earlobe for many years, but a lesion became persistent, and showed no spontaneous healing for several months. It seems most likely that the acupuncture was coincidental in this case, however, we should keep an open mind. and bed rest. A repeat chest X-ray study the following day showed improvement of the bilateral pneumothorax. She was discharged two days later with a complete resolution of the pneumothorax. The image accompanying this short account was of a plain chest X-ray showing a bilateral 15 to 20% pneumothorax. There is insufficient detail to gain more than a reminder of the potential risk of needling over the chest. It is not clear from the account why the patient attended the hospital in the first place. The reader must assume it was with mild shortness of breath, although the authors comment that there was no respiratory distress. 139