Does homoeopathy work? Part I: A review of studies on patient and practitioner reports^*

Introduction

Priya Chanda, Adrian Furnham

The question of homoeopathy’s effectiveness as a cure continues to attract both public and scientific scrutiny and debate.¹–4 Founded roughly 250 years ago by German physician Samuel Hahnemann, homoeopathy is a holistic therapy which maintains that it takes into account patients’ physical symptoms as well as medical history, personality and behaviour.⁵

It is based on two main theoretical tenets.⁶ The first is the ‘like cures like’ principle, which states that patients with certain symptoms can be cured if given a substance that produces the same symptoms in a healthy individual. The second tenet is that of ‘potentiation’ – a process whereby serial dilution is combined with vigorous shaking. It is thought that potency increases as the remedy is further diluted – a principle which opposes that found in modern science. Indeed, scientists are quick to point out that homoeopathic remedies are usually so highly diluted that no active substance remains in the remedy.⁶ They argue therefore that the effectiveness of homoeopathy cannot be anything more than a placebo response, citing rigorous RCTs supporting this. Equally, however, homoeopaths point to trials demonstrating the opposite. They argue that an underlying mechanism occurs whereby water is able to retain biological information through the repeated agitation in the dilution process – the so-called ‘memory of water’.⁷

Indeed both sides induce considerable passion, as exemplified by the ‘NHS alternative therapies row’.⁸ A letter to NHS Trusts signed by 13 leading scientists and doctors openly criticised CM. In particular, homoeopathy was described as an ‘implausible treatment’ not deserving of government funding.⁹ This letter was met by a speech made by Prince Charles to the World Health Assembly in Geneva, in which he urged every country to integrate holistic treatment with conventional medicine. It is certainly apparent, then, that despite its 250-year history, homoeopathy still remains in the limelight and very much shrouded by controversy.

This paper offers the lay or patient perspective on homoeopathy. Two points should be noted from the outset: first, only studies evaluating human subjects have been included in this review. Second, only RCTs comparing homoeopathy to placebo have been included because the scientific community’s primary concern is whether the response to homoeopathy is more than a placebo response.¹⁰

^*Part II: A review of recent scientific papers will be published in the next issue of FACT (September 2008).

The lay perspective: Clinical outcome studies

The papers on this topic are of varied quality and were selected using the following search criteria. First, the study must examine the efficacy of homoeopathy alone, rather than homoeopathy as part of CM in general. Second, the study must focus on patient or clinician assessment or patient and clinician assessment of outcome. Third, outcome assessment must include an evaluation of both physical and mental components of well-being. As Vincent and Furnham point out, both are important in assessing efficacy in healthcare.¹¹ Additionally, homoeopathy’s holistic approach encompasses both body and mind – thus studies focusing purely on one or the other would not fully reflect the treatment’s efficacy. Finally, outcome measures must be clearly defined and explained.

The majority of studies found did in fact fit these criteria and are summarised in Table 1. On a basic level one can see from Table 1 that outcome studies have been carried out for a number of different illnesses. Some focus on patient/clinician assessment generally for a variety of conditions. Others focus on a set of illnesses or a subset of illnesses and yet others focus on a specific condition. Each of these types of papers has been grouped together and will be analysed in turn.

Despite differences in focus it should be noted that all of the outcome studies identified point to the same result –that the majority of homoeopathic patients/clinicians report a positive outcome in response to homoeopathy. What constitutes a ‘positive’ result is, however, not necessarily uniform across all studies. For example, success may be defined after quite different periods of time. Indeed it will become apparent on closer analysis that there are a number of limitations to these studies.

The studies focusing on a variety of illnesses assess outcome either by recording descriptive answers from patients/clinicians or by asking them to rate response according to a scale. One of the most recent studies using the descriptive answer method recorded the self-assessed response of 1223 patients attending homoeopathic centres in northern and central Italy.¹² An impressive 95.5% of patients reported homoeopathy to be ‘effective or very effective’. Similarly, Van Wassenhoven’s study of 782 Belgian patients reported 95% to be ‘fairly or very satisfied’, and this was re-inforced by a similar assessment from physicians (all members of the Unio Homoeopathic Belgica).¹³ Cross-sectional studies like these only assess outcome at a single point in time and so cannot shed light on long-term outcomes.¹⁴ Indeed Pomposelli et al. note that they gave ‘a standard questionnaire to all adults’ but do not give an indication of what stage in homoeopathic treatment patients were asked to fill this out.¹² Thus it is unclear at what stage homoeopathic patients believe their treatment to be effective.

Attena asked 648 Italian patients to fill out cards detailing their clinical diagnosis after their first consultation with a pluralist homoeopath.¹⁵ A year later patients rated their health and well-being according to Question 2 of the SF-36. This question gave five graded options ranging from ‘much better’ to ‘much worse’. In addition, patients were asked more generally if they were satisfied with their treatment, requiring a simple ‘yes’ or ‘no’ answer. Of 609 patients, 73.5% reported they were ‘much better’ or ‘somewhat better’ and 83.9% answered ‘yes’ regarding satisfaction. These results emphasise the importance of the phrasing of questions used in outcome studies and the effect this can have on response. They also point to individual differences between patients because 83.9% report general satisfaction whereas only 73.5% report significant improvement. Of course, the effectiveness of homoeopathy is not uniform across all patients for all diseases; a concept that will be built on in the analysis of further studies.

Certainly those studies employing a point scale to assess outcome also reflect individual differences. A study carried out by Richardson¹⁶ at a health centre in Liverpool, UK, monitored patients over a year, asking them to self-assess improvement according to the Glasgow Homoeopathic Hospital Outcome Scale (GHHOS),¹⁷ as shown below:

This scale takes into account improvement and deterioration not only in presenting complaints but also in energy level, mood and value of daily living.¹⁸ Success of a treatment is usually defined as +1 and according to this Richardson reports 76.6% of patients with successful outcomes to homoeopathy.¹⁶ However, defining success from +1, a ‘slight improvement’, seems an inappropriate measure of efficacy – for homoeopathy to be truly effective it should at least affect daily living. However, Richardson highlighted the point that 60.3% self-assess themselves as +2. This shows us that of those patients reporting successful outcomes only a small proportion report a slight improvement; the majority report a significant improvement impacting on their daily lives. This is further supported by Clover’s study in which 74% of 1372 patients report success but 55% record a rating of +2.¹⁹

Richardson also noted that 52% of patients were able to reduce their use of conventional treatment.¹⁶ Not all outcome studies, however, monitor use of conventional treatment while evaluating homoeopathy. Yet, out of those that do, all report reduced or discontinued use in conventional medicine (as shown in Table 1).

See Table 1: Clinical outcome studies of homoeopathy.

Slade carried out an outcome study at a GP-led practice also using a 7-point scale.²⁰ However, the study was conducted slightly differently in that patients were asked to rate primary symptoms, secondary symptoms and well-being separately. They made these ratings once after their first consultation and once after a mean time of 134 days. The difference in their scores was then calculated for these three distinct categories, a larger difference indicating greater improvement. Both primary and secondary symptoms improved by a mean of 2.49 points (95% CI, 2.08 to 2.90 and 2.00 to 2.98, respectively). However, well-being (encompassing emotional status) improved by a lower 1.41 points (95% CI, 1.02 to 1.80).

These results suggest that patients regard homoeopathy to be more effective in relieving physical symptoms. This is supported by two further studies. First, Robinson only assessed physical symptoms and found 78% of patients to report improvement.²¹ Second, Steinsbekk and Lüdtke²² separated physical complaints from emotional ones similarly to Slade,²⁰ finding that 71% of patients reported at least a 10 mm improvement (VAS) in their main physical complaint (95% CI, 67 to 74) and 51% reported at least a 10 mm improvement in their well-being (95% CI, 48 to 55). It should be noted that both Slade²⁰ and Steinsbekk and Lüdtke²² provide CIs that substantiate the observed difference between physical and emotional symptoms. In both studies, even if the upper limit of the physical complaint is compared with the lower limit of the well-being result there is still a significant statistical difference.

It seems then that the patient-rating evidence for homoeopathy’s slightly higher effectiveness in relieving physical symptoms is convincing. However, the approach these studies take in separating physical and emotional components can be called into question. Itamura points out that skin disease is inextricably linked to psychological state; therefore, it is very difficult to isolate variables.²³

Indeed the desire to split the physical from the emotional bears on biomedicine’s much criticised mind–body dichotomy.²⁴ It would be naive to presume that reported improvement in physical symptoms occurs without the influence of a psycho-biological response. Associations have been made, for example, between the outcome of cancer patients and their level of optimism.²⁵,²⁶

The above studies include both acute and chronic conditions. However, homoeopathy is especially well known for its popular use in treating chronic illness, no doubt because it represents a new hope for desperate individuals who find conventional treatment ineffective.¹⁸ Certainly a study carried out by Spence et al. showed positive results in homoeopathic treatment for chronic illnesses.²⁷ In this study 6544 patients rated improvement according to a 7-point scale: 70.7% reported improvement as +1; 50.7% reported improvement as +2. This supports the finding of Richardson¹⁶ and Clover¹⁹ that the majority of patients report significant improvement.

Sevar had 455 patients rate improvement together with their practitioners, according to the GHHOS; 66.8% reported a positive response to homoeopathy as +1.¹⁸ Most patients in this study rate improvement as +2 or +3. Only a minority report a slight improvement (+1) or a complete cure (+4). It should be noted that percentages such as those given by Spence et al.,²⁷ Richardson¹⁶ and Clover¹⁹ may mask the possibility that only a few patients experience a ‘full cure’. This is because the percentage broadly reflects significant improvement and does not separate those reporting a complete cure.

Witt et al. also conducted a study into chronic illnesses in which patients and doctors assessed outcome according to a VAS (0–10) after 12 months of treatment.²⁸ Patient assessments of homoeopathy improved in adults by 2.5 points for homoeopathy but by 1.5 points for conventional medicine. Physician assessments also supported those of patients. This study suggests, then, that in the case of chronic illness, both patients and physicians believe homoeopathic treatment to be more effective than conventional treatment. This study has an additional advantage in that it compares homoeopathy with a control group (in this case standard conventional treatment).

Respiratory conditions are a subset of illnesses most commonly treated by homoeopaths.²⁹ However, Table 1 shows that the number of studies meeting the required search criteria was substantially less for respiratory allergies than those focusing on a variety of illnesses or chronic diseases. Colin focused on the response of 147 patients to complex homoeopathy, asking homoeopaths to rate outcome according to a scale of +1 to +4, where +1 indicated deterioration and +4 a complete disappearance of symptoms. He found that after a minimum of 2 years of treatment, homoeopaths rated 87.6% as +1. It should be noted that this percentage took into account 31 patients who gave up homoeopathic treatment. Percentages that do not take into account patients who discontinue treatment therefore give a misleading impression of efficacy in terms of both patient compliance and benefit.

A study previously carried out by Riley et al. also demonstrated a positive response to homoeopathy.³⁰ It compared the response of 281 patients receiving homoeopathy to 175 receiving conventional treatment for a number of respiratory/ear complaints (patients were matched according to whether they had upper respiratory tract conditions, lower respiratory tract conditions or ear complaints); 82.6% of those receiving homoeopathy reported a cure or major improvement after 14 days as opposed to 68.0% receiving conventional treatment. Certainly, an advantage of this study in methodological terms is that Riley, like Witt et al.,²⁸ uses a control group. However, unlike Witt et al., Riley’s data lacks the support of appropriate statistical analysis.

The studies in Table 1 show that there are even fewer studies assessing the efficacy of homoeopathy on specific conditions. A study carried out by Itamura²³ asked 60 Japanese patients suffering from chronic skin disease to take individualised homoeopathy in addition to conventional treatment for at least 3 months. Afterwards patients self-assessed outcome according to a 9-point scale similar to the GHHOS; 88.3% reported improvement as +2. This positive result may appear then to support the positive results of studies assessing chronic diseases in general. However, one should be cautious of comparing this study with those discussed earlier as they differ fundamentally in methodology – those focusing on sets of chronic illnesses assessed the outcome of homoeopathy alone whereas Itamura’s study assesses the efficacy of homoeopathy in conjunction with conventional medicine. In addition, Itamura found that homoeopathy was particularly effective for atopic dermatitis and eczema. It is evident, then, that even ‘specific’ conditions may encompass individual diseases and these are deserving of further investigation.

Relton and Weatherley-Jones asked patients to complete the Measure Yourself Medical Outcome Profile (MYMOP) before, during and after treatment, which made them rate their two most troublesome complaints and well-being on a 7-point scale.³¹ 81.3% of patients reported improvement and on average these patients reported decrease in the severity of their physical symptoms by 2 points and an improvement in well-being of 1.5 points. This reiterates the pattern emerging from studies by Slade,²⁰ Robinson²¹ and Steinsbekk and Lüdtke,²² suggesting homoeopathy’s slightly higher effectiveness in treating physical symptoms. However, the approach in separating the physical and mental elements can be debated.

Relton and Weatherley-Jones also highlight the issue that homoeopathy seems particularly effective for headaches, tiredness, vasomotor symptoms, locomotor symptoms and sleeping difficulties.³¹ They point to two other outcome studies that appear to offer further support of homoeopathy’s efficacy in treating menopause. However, both of these studies –Clover¹⁴ focusing on hot flushes and Thompson and Reilly³² focusing on oestrogen withdrawal in breast cancer patients – address different symptoms to those specifically highlighted by Relton and Weatherley-Jones. It is curious that the authors chose to highlight these further studies – at first glance they appear to offer more support for homoeopathy’s effectiveness in treating menopause, but on closer inspection they analyse different symptoms of menopause to those specifically highlighted by Relton and Weatherley-Jones. It is again evident that ‘specific’ conditions may be further subdivided and that individual symptoms must be carefully considered when comparing studies.

Discussion

Overall, then, an analysis of the studies itemised in Table 1 suggests that:

• The majority of patients undertaking homoeopathy and practitioners monitoring their treatment report improvement in terms of both physical and emotional symptoms – essentially they believe homoeopathy to be an effective cure.
• It is possible that physical symptoms may be treated more effectively by homoeopathy than emotional symptoms – although it is debatable whether physical and mental components can be clearly separated.
• Homoeopathy can be integrated with conventional treatment.
• Homoeopathy can result in reduced or even discontinued use of conventional treatment.

It has certainly become apparent that although the studies in Table 1 all show positive results, one should be wary of taking these at face value. There are a number of limitations, some of which have already been highlighted:

• The lack of a control group in most of these studies.
• The shortcomings of a cross-sectional as opposed to longitudinal approach.
• Percentages that may mask the possibility that only a few patients experience a full cure.
• The impact of how exactly questions about effectiveness are phrased.
• The difficulty in comparing studies using different outcome measures.
• The lack of robust and relevant statistical analysis in most of these studies.
• The problems with grouping illnesses or symptoms together for analysis.

Indeed it is curious that so many outcome studies group together a variety of illnesses or sets of illnesses. As Vincent and Furnham point out, the effectiveness of a standard drug treatment would never be analysed using such a grouped approach.¹¹ In order to gain a better understanding of homoeopathy’s efficacy, ideally the form of treatment (complex/individualised), the name and potency of remedy and individual diseases/ symptoms should all be specified. None of the outcome studies analysed combine all of these – only Itamura²³ and Attena¹⁵ distinguish between complex or individualised homoeopathy; only Sevar¹⁸ specifies the effect of a particular remedy; only Itamura²³ and Relton and Weatherley-Jones³¹ point to specific diseases/symptoms (although these require further study) and none detail the potency of the remedy. The lack of specificity in these respects results in great variation between studies, making it difficult to distinguish whether a particular homoeopathic remedy is effective for a particular disease.

Over and above the limitations thus far discussed there are several others which apply to the outcome studies in Table 1 as a whole:

• Sample size – larger samples are naturally thought to be more representative of the population. However, as Pomposelli et al.¹² point out, even large samples may not be completely representative of all those who seek homoeopathy because they do not take into account those who self-administer.
• ‘Ideological’ bias – even if the sample was completely representative of all who seek homoeopathy, a fundamental criticism is that the study only encompasses those open to, and accepting of, homoeopathy’s principles – thus their response to treatment may be strongly influenced by the placebo response. Furthermore, homoeopaths themselves are likely to be biased as they will be eager to obtain positive results in support of their profession and source of income. Linked to this is the problem of patients filling out questionnaires positively in order to please their clinician. To combat this problem, only Attena uses an independent researcher to question patients.¹⁵
• Multiple variables – outcome may be influenced by the expertise and prescribing strategy of the clinician,²¹ the patient–doctor relationship,³³ patient compliance and individual differences such as gender and age, as discussed earlier. Thus establishing whether reported improvement/deterioration is due to the specific effect of homoeopathy remains difficult to distinguish. Indeed the variables relating to both doctors and patients are by no means unique to homoeopathy and must also be considered when evaluating the efficacy of other medical treatments.
• Subjectivity – on a basic level both patient and clinician assessment is subjective, symptoms often being difficult to quantify.²³ Furthermore, McGee distinguishes between patient and clinician subjectivity, highlighting that clinicians usually rate the patient as having a worse condition.³⁴
• Dissonance reduction – according to Festinger, cognition dissonance occurs when different ideas or beliefs contradict each other.³⁵ Reduction in response to this occurs when people try to ignore or deny conflicting information. When patients pay for homoeopathic treatment and it does not work this creates dissonance. In order to reduce this dissonance patients are likely to convince themselves that the treatment has worked rather than admit their money has been wasted. Thus a patient’s success ratings in outcome studies may be affected by dissonance reduction.
• Reasons for discontinuation – although some studies take into account the number of patients discontinuing treatment, the reasons for discontinuation are often not indicated. This is important – one might assume a patient gives up because the treatment is not effective but, equally, a patient may give up because their treatment was successful and is no longer necessary.³⁶ These reasons do, of course, bear quite differently on the question of efficacy.

However there is also the external problem of publication bias. As stated from the outset, every single outcome study in Table 1 demonstrates a positive result in response to homoeopathy. It is also clear that the majority of these studies are published in the journal Homoeopathy or CM journals. In light of this, Lehmacher points out: ‘… the fact that the majority of homoeopathic studies are positive is no proof of homoeopathy’s efficacy’.³⁷

It should also be noted that during the search process no negative results were found in either journals of CM or journals of conventional science and medicine, which are perhaps more inclined to publish negative or critical evidence about CM.

Indeed, despite their limitations, the results of clinical outcome studies should not be completely negated. At the very least they form hypotheses that may be tested by experimental studies.³⁸

Conclusion

There is a paradox at the heart of the literature on the effectiveness of homoeopathy. Most studies on patients point to the fact that they claim the treatment works for them, yet most good empirical studies in the randomised, double-blind control condition suggest that any effects must be due to a placebo effect. Beyerstein has asked the simple but important question of why any therapy is thought to work if it is (a) implausible on a priori grounds; (b) it lacks a scientific acceptable rationale of its own; (c) it has insufficient supporting scientific evidence; (d) it has failed well-controlled clinical trials conducted by impartial evaluators; and (e) it is improbable even to the layperson.³⁹ He then lists ten errors and biases which may account for why ‘bogus’ therapies seem to work:

1. The disease may have run its natural course, because many diseases are self-limiting.
2. Many diseases are cyclical and have ‘natural ups and downs’ and it may look as if the treatment is causing the recovery periods.
3. Spontaneous recovery such that psychoneuroimmunological factors affect both susceptibility to, and recovery from many diseases.
4. The ubiquitous but little understood placebo effect, which causes misattributions as to the precise cause of the ‘cure’.
5. Some allegedly cured symptoms are psychosomatic to begin with, thus the cause (stress) is cured by reassurance and psychosocial not physiological processes.
6. There is symptomatic relief rather than cure, reducing pain and discomfort rather than really dealing with the serious underlying problem.
7. Consumers or patients ‘hedge their bets’ by using complementary rather than solely alternative treatment and hence bogus treatments, used to supplement orthodox treatments, get a disproportionate share of the credit.
8. There is misdiagnosis by either the patient and/or the physician in the sense that they ‘recover’ from an incorrect diagnosis.
9. There are clear, real, derived benefits from charismatic or caring practitioners that cause the mood and outlook of patients really to increase.
10. Patients have a psychological distortion of reality because they are reducing their dissonance, having paid for and endured their therapies.

Beyerstein concludes that, ‘individual testimonies count for very little in evaluating therapies’. He also notes: ‘To people who are unwell, any promise of a cure is especially beguiling … Erstwhile savvy consumers, felled by disease, often insist upon less evidence to support the claims of alternative healers than they would previously have demanded from someone hawking a used car. Caveat emptor.’ (p. 7)

This may therefore lead to the unusual conclusion that no further work of this kind (i.e. studies on patient reports of treatment) needs to be done because of the systematic errors and biases in the existing reports.

References

1. Ernst E. Homoeopathy: A Critical Appraisal. Oxford: Butterworth-Heinemann , 1998. 107–17
2. Ernst E. Classical homeopathy versus conventional treatments: a systematic review. Perfusion 1999; 12: 13–15.
3. Ernst E, Hahn GH. Introduction. In: Ernst E, Hahn GH (Eds). Homoeopathy: A Critical Appraisal. Oxford: Butterworth-Heinemann , 1998. ix–x.
4. Ernst E, Pittler MH. Efficacy of homeopathic arnica. A systematic review of placebo-controlled clinical trials. Arch Surg 1998; 133: 1187–90.
5. Zollman C, Vickers A. ABC of Complementary Medicine. London: BMJ Books , 2000.
6. Ernst E. Is homeopathy a clinically valuable approach? Trends Pharmacol Sci 2005; 26: 547–8.
7. Schiff M. The Memory of Water. Northampton: Thorsons , 1995.
8. Boseley S. Doctors’ letter sparks NHS alternative therapies row. Guardian 2006; p. 3. 24 May
9. BBC News. Doctors’ letter: In full. The open letter from some of the UK’s leading doctors urging NHS trusts to stop using complementary therapies. [online document], http://news.bbc.co.uk/go/pr/fr/-/1/hi/health/5008034.stm accessed May 23, 2006.
10. Kron M, English JM, Gaus W. Guidelines on methodology of clinical research in homoeopathy. In: Ernst E, Hahn GH (Eds). Homoeopathy: A Critical Appraisal. Oxford: Butterworth-Heinemann , 1998. 9–47.
11. Vincent F, Furnham A. Complementary Medicine: A Research Perspective. Chichester: John Wiley and Sons Ltd , 1998.
12. Pomposelli R, Andreoni C, Costini G et al. Opinions and self-reported health status of Italians seeking homeopathic treatment. Homeopathy 2006; 95: 81–7.
13. Van Wassenhoven, Ives G. An observational study of patients receiving homeopathic treatment. Homeopathy 2004; 93: 3–11.
14. Shibata M, Flather M, Perez de Arenaza D. Design issues in observational studies and randomized trials. In: Flather M, Aston H, Stables R (Eds). Handbook of Clinical Trials. London: Remedica Publishing Ltd , 2001. 1–15.
15. Attena F, Del Giudice, Verrengia G, Granito C. Homoeopathy in primary care: self-reported change in health status. Complement Ther Med 2000; 8: 21–5.
16. Richardson WR. Patient benefit survey: Liverpool Regional Department of Homeopathic Medicine. Br Homeopath J 2001; 90: 158.
17. Reilly D. The Evidence for Homeopathy. [online document], http://www.adhom.com/adh_download/EVIDENCE_9.0_Sept_06.pdf accessed February 12, 2007
18. Sevar R. Audit of outcome in 455 consecutive patients treated with homeopathic medicines. Homeopathy 2005; 94: 215–21.
19. Clover A. Patient benefit survey: Tunbridge Wells Homoeopathic Hospital. Br Homeopath J 2000; 89: 68–72.
20. Slade K, Chohan BP, Barker PJ. Evaluation of a GP practice based homeopathy service. Homeopathy 2004; 93: 67–70.
21. Robinson T. Responses to homeopathic treatment in National Health Service general practice. Homeopathy 2006; 95: 9–14.
22. Steinsbekk A, Lüdtke R. Patients’ assessments of the effectiveness of homeopathic care in Norway: A prospective observational multicentre outcome study. Homeopathy 2005; 94: 10–16.
23. Itamura R. Effect of homeopathic treatment of 60 Japanese patients with chronic skin disease. Complement Ther Med 2007; 15: 115–20.
24. Scheper-Hughes N, Lock M. The mindful body: a prolegomenon to future work in medical anthropology. Medical Anthropology Quarterly 1987; 1: 6–41.
25. Gustavsson-Lilius M, Julkunen J, Hietanen P. Quality of life in cancer patients: The role of optimism, hopelessness, and partner support. Qual Life Res 2007; 1: 75–87.
26. Von Ah, Kang DH, Carpenter JS. Stress, optimism, and social support: impact on immune responses in breast cancer. Res Nurs Health 2007; 1: 72–83.
27. Spence DS, Thompson EA, Barron SJ. Homeopathic treatment for chronic disease: a 6-year, university-hospital outpatient observational study. J Altern Complement Med 2005; 11: 793–8.
28. Witt C, Keil T, Selim D et al. Outcome and costs of homoeopathic and conventional treatment strategies: A comparative cohort study in patients with chronic disorders. Complement Ther Med 2005; 13: 79–86.
29. Colin P. Homeopathy and respiratory allergies: a series of 147 cases. Homeopathy 2006; 95: 68–72.
30. Riley D, Fischer M, Singh B et al. Homeopathy and conventional medicine: an outcomes study comparing effectiveness in a primary care setting. J Altern Complement Med 2001; 7: 149–59.
31. Relton C, Weatherley-Jones E. Homeopathy service in a National Health Service community menopause clinic: audit of clinical outcomes. J Br Menopause Soc 2005; 11: 72–3.
32. Thompson EA, Reilly D. The homeopathic approach to the treatment of symptoms of oestrogen withdrawal in breast cancer patients. A prospective observational study. Homeopathy 2003; 92: 131–4.
33. Bikker AP, Mercer SW, Reilly D. A pilot prospective study on the consultation and relational empathy, patient enablement, and health changes over 12 months in patients going to the Glasgow Homoeopathic Hospital. J Altern Complement Med 2005; 11: 591–600.
34. McGee HM. Outcomes: do health professionals and patients have the same objectives? In: Meade T (Ed). Science-based Complementary Medicine. London: Royal College of Physicians , 1998. 23–31.
35. Festinger L. A Theory of Cognitive Dissonance. Stanford, CA: Stanford University Press , 1957.
36. Endrizzi C, Rossi E. Patient compliance with homeopathic therapy. Homeopathy 2006; 95: 206–14.
37. Lehmacher W. Methodological principles in clinical trials. In: Ernst E, Hahn GH (Eds). Homoeopathy: A Critical Appraisal. Oxford: Butterworth-Heinemann , 1998. 3–8.
38. Mathie RT. Clinical outcomes research: contributions to the evidence base for homeopathy. Homeopathy 2003; 92: 56–7.
39. Beyerstein B. Why bogus therapies seem to work. Skeptical Inquirer 1997: Sept–Oct [online publication], http://csicop.org/si/9709/beyer.html accessed February 12, 2007.