- Correspondence to: S O'Meara, NHS Centre for Reviews & Dissemination, University of York, York YO10 5DD, UK;
The research evidence on the effectiveness of homoeopathy presented in a recent issue of Effective Health Care is reviewed.
Homoeopathy is a system of treating patients using very low dose preparations according to the principle: “like should be cured with like”. This paper summarises the research evidence presented in a recent issue of Effective Health Care on the effectiveness of homoeopathy.1
Increasing numbers of patients are seeking information on complementary medicines from NHS health professionals.2 Results of a 1998 survey of use and expenditure on complementary medicine in England suggested that 28% of respondents had either visited a complementary therapist or had purchased an over the counter herbal or homoeopathic remedy in the past year.3 From this survey it was estimated that there could be over 470 000 recent users of homoeopathic remedies in England.3
Homoeopathy has been part of the NHS since its inception.4 There are currently five homoeopathic hospitals, of which the two largest in Glasgow and London have inpatient units. These hospitals provide a range of conventional and complementary treatments in addition to homoeopathy.
Most of the conditions treated by homoeopathic practitioners are chronic or recurrent. They also treat a large number of patients with ill defined illnesses that have not been given a conventional diagnosis.5 Initially, a very detailed history is taken from the patient, a clinical examination is performed, and all signs and symptoms are recorded. Attention is paid to alternating or unusual symptoms and information is sought on the impact of modalities (conditions providing relief or aggravation of symptoms such as weather or activity). The symptoms are then matched to remedies using either a homoeopathic repertory or “pattern recognition”.
Homoeopathic remedies are often known as potencies and are prepared by a process of serial dilution with succussion (vigorous shaking).5,6 Such dilutions are known as ultramolecular in that they are diluted to such a degree that not even a single molecule of the starting substance is likely to be present. The claim that these dilutions have an active mechanism is the source of most of the scientific controversy surrounding homoeopathy.
Methods of prescribing vary among homoeopathic practitioners (see box 1).5,7 Following administration of a remedy, the homoeopathic practitioner follows the patient's progress and pays attention to the development of symptoms, and will repeat or adjust the prescription depending on what is observed.6
Classical: single remedy prescribed based on patient's presentation and history
Complex: more than one remedy used concurrently
Fixed: same single agent used for a group of patients
Isopathy: preparation based on causal agent
Phytotherapy: administration of herbs or low potencies of herbs
NATURE OF THE EVIDENCE
Around 200 randomised controlled trials (RCTs) evaluating homoeopathy have been conducted, and there are also several systematic reviews of these trials. This paper is based mainly on an overview of existing systematic reviews of RCTs. Some reviews are general overviews, some focus on individualised (classical) homoeopathy, while the remainder have a more specific focus. Individual RCTs published subsequent to the included reviews of individualised homoeopathy and those with a specific scope are also included (more detail on the included RCTs is available at www.york.ac.uk/inst/crd/ehcb.htm). Details of the review methods are available elsewhere.1
There are a number of problems and controversies surrounding the existing evidence base for homoeopathy. Firstly, there is much debate over whether homoeopathy shows any effect over and above placebo (a dummy medication or treatment given to participants in trials). Sceptics have argued that homoeopathy cannot work because of the use of remedies that are diluted to such a degree that not even a single molecule of the starting substance is likely to remain. Given the absence of a plausible mechanism of action, it has been argued that the existing evidence base represents little more than a series of placebo versus placebo RCTs.8,9
Others have argued that much of the research conducted on the effectiveness of homoeopathy is not representative of routine homoeopathic practice as homoeopathic treatment is highly individualised—that is, two patients with similar symptoms may receive different treatments.10 While it is possible to carry out RCTs evaluating the efficacy of homoeopathy, researchers have tended to focus on conducting placebo controlled RCTs either to test the effects of a single remedy on a particular condition and/or to explore the placebo issue. As such, conditions such as delayed onset muscle soreness (DOMS) have been subject to study whereas skin conditions such as eczema, which are commonly treated by homoeopathy, have been overlooked.10
Most RCTs of homoeopathy have involved small numbers of patients and have suffered from low statistical power. Given the controversy surrounding the plausible mechanism of action for homoeopathy, there have been calls for stronger levels of evidence for its effectiveness than would normally be required for more conventional interventions.7,11
REVIEWS WITH A GENERAL SCOPE
Four systematic reviews were identified (table 1).7,12–14 The purpose of these reviews was to determine whether there is any evidence for the effectiveness of homoeopathic treatment generally. Patients with any disease were included rather than investigating effects within a specific group such as those with asthma. Because of the general nature of all four reviews, characteristics of the participants and outcomes were not specified in the selection criteria for primary studies and both participants and interventions varied greatly. All four reviews included RCTs and one also included non-randomised studies.6 Each review covered several different types of homoeopathy including classical, fixed, complex, and isopathy. All reviews identified methodological problems within the primary studies and, as such, were unable to draw firm conclusions about the general effectiveness of homoeopathy. It should be noted that the analyses undertaken in two of the reviews involved the statistical pooling of clinically heterogeneous data and therefore the estimates shown should be viewed with caution (table 1).13,14
REVIEWS OF INDIVIDUALISED (CLASSICAL) HOMOEOPATHY
Two reviews were identified (table 1).15,16 Again, the scope of these reviews was general and selection criteria relating to participant characteristics and outcome measurements were unspecified. Methodological problems with the primary studies were reported in both reviews.15,16
One review assessed the effectiveness of individualised homoeopathy compared with placebo, no treatment, or another therapy, and included randomised, quasi-randomised, or double blind trials (n=32).15 The results from a pooled analysis of 19 trials indicated a statistically significant result in favour of homoeopathy. However, when the analysis was limited to six trials of higher methodological quality, the difference between homoeopathy and control treatments was no longer statistically significant (table 1). It should be noted that clinically heterogeneous data were combined in the analyses, and assessments of statistical heterogeneity were not reported. The results should therefore be interpreted with caution.
The second review assessed the effectiveness of individualised homoeopathy compared with allopathic (conventional) medications and included RCTs and non-randomised controlled trials.16 Six studies were included, each involving a different disease. The results suggested that homoeopathic remedies may be superior to conventional drug therapy for rheumatoid arthritis and otitis media in children. However, conventional drug therapy may be better than homoeopathy for proctocolitis (inflammation of the rectum and colon) and tonsillitis in children. No between group differences were found for trials of irritable bowel syndrome and malaria. This review did not present details of individual studies, including aspects of methodological quality, and therefore it was difficult to judge the validity of the findings of the review.
Four further RCTs of classical homoeopathy, all of reasonable methodological quality, were identified,17–20 two of which were included in one of the above reviews15 but had been reported only in abstract form.17,18 In addition, a follow up study relating to a trial of classical homoeopathy included in a review on homoeopathic prophylaxis of headaches and migraine was identified21 and will be described later.22
In the earliest trial patients with mild traumatic brain injury were recruited.17 After 4 months, statistically significant effects in favour of homoeopathy were observed for changes in some scores of physical, cognitive, and affective symptoms and functional disability.
A small trial (n=23) compared homoeopathy with placebo in relieving symptoms associated with the premenstrual syndrome (PMS).18 The results were in favour of homoeopathy for improvement in menstrual symptoms at 3 months (p=0.057), mean symptom improvement rate (p=0.048), and the proportion of women experiencing more than 30% improvement (38% versus 90%, p=0.037).
Another trial assessed the effects of classical homoeopathy in treating children with a recent history of diarrhoea.19 The results suggested that homoeopathy was significantly more effective than placebo in reducing the frequency of diarrhoea and the duration of illness. The same research group conducted another trial (n=75) on children with acute otitis media.20 No statistically significant between group differences were seen for treatment failure or middle ear effusion.
REVIEWS WITH A MORE SPECIFIC FOCUS
Since all of the reviews described so far have aimed to assess whether homoeopathy as a general system shows any effect over and above placebo, no specific implications can be derived for clinical practice. The following sections provide details of nine reviews with a more specific focus in terms of the homoeopathic agent being evaluated or the type of participants recruited (table 2).21,23–30
One review focused on the effectiveness of homoeopathic arnica.24 The findings did not indicate that homoeopathic arnica is any more effective than placebo. Some study details were lacking, particularly with regard to results and methodological quality, and therefore it is difficult to assess the reliability of the evidence.
Eight placebo controlled trials (including four RCTs) were included. The conditions represented included: DOMS, postoperative care, trauma, stroke, and experimental bruising (bruising deliberately induced in healthy volunteers under laboratory conditions). Two trials showed a statistically significant result in favour of arnica when used to treat DOMS and to prevent postoperative complications. However, the remaining six trials did not show statistically significant between group differences.24
A further five RCTs of the use of homoeopathic arnica were identified.31–35 Three were concerned with DOMS31–33 and two with surgical patients.34,35 In the trials of DOMS, two of the three studies did not show statistically significant between group differences.31,32 The surgical trials focused on recovery after total abdominal hysterectomy34 and saphenous stripping (stripping of varicose veins).35 Neither trial found statistically significant differences between groups.
Postoperative ileus (bowel muscle paralysis)
Postoperative ileus refers to cessation of peristalsis due to paralysis of the bowel muscle following surgery or trauma to the bowel. One review assessed the effectiveness of homoeopathic treatment versus placebo in resolving postoperative ileus and included six trials (four RCTs) of patients undergoing abdominal or gynaecological surgery.23 All trials used fixed homoeopathic preparations (as opposed to individualised prescription). The findings indicated that homoeopathic treatment administered immediately after abdominal surgery may reduce the time to first flatus compared with placebo. However, the possibility of bias and inappropriate pooling of data means that these findings should be treated with caution. In addition, the largest and most well conducted study, as rated by the authors of the review, showed no difference between homoeopathy and placebo. No further RCTs were identified.
Delayed onset muscle soreness (DOMS)
The effectiveness of homoeopathy in reducing DOMS was assessed in a review of eight trials, including three RCTs.27 The results suggested that homoeopathic remedies were no more effective than placebo in alleviating DOMS.
Participants were healthy volunteers who had undergone some form of exercise in order to induce DOMS. There was a high level of heterogeneity between included studies in terms of the homoeopathic remedies and the type of exercise used to induce DOMS. The three RCTs all reported non-significant differences between treatment groups, while results from the non-randomised studies were inconsistent.27
Arthritis and other musculoskeletal disorders
Two reviews were identified.28,29 One examined the effectiveness of homoeopathy in people with rheumatoid arthritis, osteoarthritis, and other types of musculoskeletal disorders.28 Most of the trials were rated by the authors of the review as being of high methodological quality. Although the overall pooled estimate indicated that homoeopathy was superior to placebo, the data were clinically heterogeneous. In addition, the outcome measurements used in the pooling were not defined but, when referring to a related publication, it seems likely that these were highly heterogenous.13 The findings of this review should therefore be treated with a great deal of caution.
The second review focused more specifically on osteoarthritis and included four RCTs.29 Fixed rather than individualised treatments were used in all trials. Results between trials were inconsistent and the authors noted methodological problems in all cases. This meant that firm conclusions could not be drawn. The methodological quality of the review was fair to good.
One additional RCT was identified.36 Patients with gonarthrosis (joint disease) received either Zeel compound tablets (a preparation containing several homoeopathic remedies) or diclofenac (a non-steroidal anti-inflammatory drug). No statistically significant between group differences were observed in pain, stiffness, functional ability, and global symptoms.
A systematic review of fair methodological quality focused on the effectiveness of homoeopathy as a prophylactic agent for headaches and migraine.21 The results suggested that homoeopathy was not effective. Four trials of classical homoeopathy versus placebo were included. One trial of poor methodological quality found a statistically significant improvement in all outcomes in favour of homoeopathy, whereas the trials of better quality all reported no statistically significant differences between groups.21
No new RCTs were identified. However, follow up data were identified for one trial rated in the review as having good methodological quality.37 At 1 year, between group differences for headache frequency, duration, and intensity were still not statistically significant.22
A well conducted review assessed the effectiveness of homoeopathy in treating stable chronic asthma or asthma-like symptoms.26 The three included RCTs were of variable methodological quality. Two showed results in favour of homoeopathy (symptom improvement, lung function improvement, and reduced use of corticosteroids) and one found no statistically significant differences between groups.
Two additional RCTs recruited patients with chronic asthma treated with corticosteroids for at least 5 years before study entry and assessed changes in respiratory function and corticosteroid use.38,39 Neither study detected statistically significant between group differences for change in respiratory function. However, one study showed results in favour of homoeopathy for a reduction in the daily dose of corticosteroids and number of infections.39 The results from both studies should be interpreted with caution due to lack of details on patient and intervention characteristics, and methodological problems such as failure to analyse by intention to treat.
A good quality systematic review assessed the use of homoeopathic oscillococcinum in preventing and treating influenza.25 Three prevention and four treatment RCTs were included. Findings indicated that oscillococcinum may reduce the duration of influenza by 0.26 days (95% CI 0.47 to 0.05) but there was insufficient evidence to suggest a preventive effect. One trial reported a higher rate of adverse events in the homoeopathy group (most frequent symptoms were aching muscles and fever). Problems with methodological quality and reporting were noted in all the trials. No further RCTs concerning the use of homoeopathic oscillococcinum or any other homeopathic preparation in the prevention or treatment of influenza were identified.
Induction of labour
One systematic review assessing the role of homoeopathy in the induction of labour was identified.30 Only one RCT (n=40) was identified which found no statistically significant differences between homoeopathic caulophyllum and placebo. However, this trial may have been too small to detect the true treatment effect. This trial has not been shown in table 2 as only one trial was involved. No further RCTs were identified.
The evidence base for homoeopathy needs to be interpreted with caution. Many of the areas researched are not representative of the conditions that homoeopathic practitioners usually treat. In addition, all conclusions about effectiveness should be considered together with the methodological inadequacies of the primary studies and some of the systematic reviews.
Common problems with the methodological quality of the primary studies included underpowered studies, failure to analyse by intention to treat, and failure to use allocation concealment (process used to prevent investigators having prior knowledge of group assignment in an RCT). The main problem with some of the systematic reviews was the pooling of clinically heterogeneous data.
There are currently insufficient data to either recommend homoeopathy as a treatment for any specific condition or to warrant significant changes in the provision of homoeopathy. The authors of many of the systematic reviews recommended further primary research to clarify or confirm conclusions relating to the effectiveness of homoeopathy. Any future research evaluating homoeopathy should address the methodological inadequacies of the existing evidence base.