Abstract
In this double-blind, three-armed, randomized controlled study (RCT) of 90 participants, individualized homeopathic interview and remedy treatment will be compared against a group receiving inert placebo with homeopathic interview, and another group receiving no additional treatment. Placebo and remedy groups will receive individualized homeopathic consultation guided by classical case-taking methods. Progress will be measured by SADS-C (Schedule for Affective Disorders and Schizophrenia C Version) for baseline, mid-study, and end-of-study in all three groups. The goal of this trial is to shed more light on RCT model design for use within the context of classical homeopathy by exploring what effects, if any, homeopathic interviewing and treatment have beyond that of placebo and conventional maintenance on persons suffering from auditory hallucinations. A secondary purpose is to explore the homeopathic principle of likenesses through case analysis by investigating what “likeness” exists between reported symptoms and remedies selected for each participant.
Research Design: Homeopathy and the Treatment of Auditory Hallucination
Controversy is on-going over the efficacy and scientific authenticity of homeopathy, often giving rise to professional angst over the modality itself, the validity of research models, and their subsequent results (Bell, 2005). Denizens of the placebo camp argue their meta-analysis clearly demonstrates that any result from treatment with a Homeopathic remedy amounts to placebo effect (Shang, et al., 2005), yet others find homeopathy trials to be of superior quality to allopathic trials (Rutten, 2009). Ullman and Frass (2010) cite inadequate external validity and strong bias in selecting trials for the Shang, et al. (2005) meta-analysis, providing a review of several overlooked trials demonstrating efficacy of homeopathic remedies. In fact, more than ten years prior to the placebo meta-analysis, one RCT (randomized controlled trial) model demonstrated consistent evidence of reproducibility (Reily & Taylor, 1994) in the homeopathic treatment of allergic rhinitis versus placebo using a daily visual analogue scale to report overall symptom intensity and outcome. Three independent studies using this model found that homeopathy acted differently and better than placebo.
Homeopaths are still working to arrive at testing models supportive of the philosophical approach to treatment (Oberbaum, Vithoulkas, & Van Haselen, 2003) and satisfying to their allopathic counterparts when it comes to reliability and reproducibility (Ullman & Frass, 2010), yet not much in this debate has settled. Oberbaum, et al. (2003) suggest there are two possible models for homeopathic research: one which is not reproducible because its focus is so closely linked to a single homeopath deeply analyzing an individual case to arrive at the most similar remedy (which they state cannot be repeatable with any other person or homeopath, since disease is considered unique for each person), and a second model using a panel of homeopaths and a process known as ‘keynote’ prescribing, which is less accurate and less likely to produce consistent remedy results. These models overlook other options and assume that classical homeopathy cannot be researched using conventional models.
The Reily and Taylor (1994) trial found evidence of efficacy and reproducibility in the treatment of allergic rhinitis, but there is a bit of a problem. When this trial is deconstructed, one finds that “placebo vials were prepared with globules impregnated with the same batch of diluent [a 70/30 water/alcohol solution] which, without the addition of any antigen, had been identically diluted and vibrated” (p. 1604). This was apparently done to eliminate confounds between preparation of remedy and preparation of placebo for the trial. The trouble is that according to classical principles of homeopathy, any substance diluted and vibrated is now ‘energized’ in such a way as to act in some capacity as a medicinal substance (Hahnemann, 1843). It is therefore not inert. While this does not necessarily invalidate the RCT model, it does invalidate the rhinitis trial results in terms of homeopathy acting differently than placebo. In this light, the trials show only that a similar remedy has actions better than a dissimilar one.
According to Bell (2005), “An incorrectly chosen remedy for a patient (a not uncommon occurrence in clinical reality) is also not homeopathic…” and constitutes “…an inactive “active” intervention that degrades the average outcome of the active/verum group” (p. 763). But any substance prepared according to the principles of classical Hahnemannian pharmacopeia is a homeopathic remedy. Giving it incorrectly means only that—it has been given incorrectly. This is not unlike giving an allopathic medicine incorrectly, so it is unfair to suggest this potential complicates homeopathic research. One thing is clear: any research model for homeopathic treatment must take into account that homeopathy is fundamentally different from conventional medicine in its approach to disease (Bell, 2005; Oberbaum, et al., 2003, Rutten, 2009;) in that homeopathy seeks to treat the individual suffering specific symptoms with a closely matched remedy that, if given to another, would cause those symptoms to arise (Hahnemann, 1843).
RCTs, in general, are not considered a great model for homeopathy (Rutten, 2009), yet reproducibility (Reilly & Taylor, 1994) and interesting results have been achieved using randomized models to trial homeopathic intervention. One study suggests it was the consultation process and not the remedy that produced clinical benefits for rheumatoid arthritis pain levels (Brien, Lachance, Prescott, McDermott & Lewith, 2011). Using a five-armed design to determine between-groups clinical benefit, patients were randomized into groupings in order to look at differences between homeopathic treatment, conventional treatment, remedy, placebo, and rheumatoid homeopathic complex (a combination remedy). This study explored physical, mental, and emotional symptoms of the participants to approach the more “whole person” ideology of homeopathic treatment, but it does not report the remedies used or any details of the consultation process, so it is impossible to analyze in terms of classical homeopathic principles (other than to say that complex remedies are against the precepts of Hahnemann (1843)).
In a pilot RCT for treatment of ADHD with homeopathy (Jacobs, Williams, Girard, Njike, & Katz, 2005), researchers had hoped to include an arm in which no intervention took place in order to compare placebo and homeopathically treated participants to non-intervention, but funding fell short. The study concluded no evidence to support therapeutic effects of individually prescribed remedies after homeopathic consultation, but also suggested a potential for the encounter to be therapeutic in and of itself. This study used an individualized approach to consultation, but researchers cited multiple limitations, including the potential that 18 weeks may not be long enough to observe changes with remedy (a common concern for supporters of homeopathic research (Bell, 2005; Ullman & Frass, 2010)). The two homeopaths in the study had used a new method of prescribing. Further, average baseline participant scores on the CPT (Continuous Performance Test) worsened over time (an average score implies a degree of acceptable or non-ADHD attributes) which confounded results. These are subjective assessments conducted by teachers and parents.
It is important to remember in all the fuss that science has, on many occasions, proven what it must later rescind after making new discoveries. Generally speaking, good research should promote this potential. Homeopathy stands only to gain from a united, community response to the call for valid research and it seems imperative the purpose of such research be clearly motivated. If homeopathy seeks acceptance from the conventional community through research, it is a poor motive. If, on the other hand, the purpose of homeopathic research is to increase scientific understanding of this approach to health and disease and to ultimately discover the workings of highly diluted/vibrated (succussed) remedies, then this is great motive for research. Conventional medical theory uses conventional research models, so it seems reasonable that those adhering to homeopathic theory must either develop appropriate models to fulfill fundamental research requirements or be willing to exit the scientific debate altogether (Rutten, 2009).
To date, no research in homeopathy addresses the issue of “likeness” in remedy choice by evaluating the prescribed remedy against a repertorized list of patient symptoms, yet this is the very foundation of homeopathy—that like things cure like things (Hahnemann, 1843). This trial does not aim to settle once and for all the argument between homeopathy and conventional medicine, nor does it hope to finally establish that homeopathy is a valid and/or legitimate modality. The goal of this trial is to shed more light on RCT model design for use within the context of classical homeopathy by exploring what effects, if any, homeopathic interviewing and treatment have beyond that of placebo and conventional maintenance on persons suffering from auditory hallucinations as measured by the SADS-C (Schedule for Affective Disorders and Schizophrenia C Version). A secondary purpose is to explore the homeopathic principle of likenesses through case analysis by investigating what “likeness” and/or “difference” exists between reported symptoms and remedies selected for each participant.
By using the Reilly and Taylor (1994) RCT model with an inert placebo, a classical approach to individualized homeopathic interviewing (Schmidt, 2004), a panel analysis of the interview with consensus prescribing (Oberbaum, et al., 2003), focusing on a single, dominant symptom (auditory hallucination) as a focal point to begin individualized treatment, and by contrasting modalities and treatments, this trial seeks to enrich understanding and work toward developing an acceptable model for homeopathic research.
Research Design
This is a three-armed randomized, controlled trial (RCT). Ninety participants suffering auditory hallucinations (specifically defined as “hearing voices believed to be outside myself”) are required. Hallucinations must be considered by the participant to hinder quality of life. Subjects must have experienced auditory hallucinations for a period of greater than six months. If currently under conventional treatment, such treatment will be maintained throughout the course of the trial (that is, participants agree to continue compliance with their respective conventional regimen). If not under conventional treatment, participants agree to forego such treatment until after trial completion. All participants must agree to and sign release for video-taped interviews and consent to baseline, mid-study, and end-of-study SADS-C.
Participant selection
All 90 Participants must be 18 years of age and legally able to sign release, disclosure, and consent forms (no requirement for legal guardian signatures). Participants will be screened and selected from respondents to national (United States) mass-media advertising and mass notification to major metropolitan area psychiatric hospitals, mental health clinics, and homeless shelters. Selection methods will attempt to maintain World Health Organization reported demographics for persons suffering schizophrenia (see Appendix) to enhance generalization. Once screening has occurred, participants will be sequentially numbered 1 through 90 by order of acceptance into the trial. When selection activities are completed, participants will be allocated to one of three groups by way of a computerized random sequence generator[1] by a third party statistician who will remain blinded to the study (see Table 1 for group composition). Participants may be currently or previously under conventional treatment with or without medications typically prescribed for auditory hallucinations. Participants may or may not have a conventional diagnosis of schizophrenia according to current diagnostic criteria. Interviews will be video-taped in their entirety at the location of the participant, coded for identification, and retained for use in this and future research. Participants will remain masked to treatment[2]. Exclusion criteria: Under age 18, under legal guardianship, prior homeopathic treatment for any reason, consistent auditory hallucinations of less than six months duration, auditory hallucinations attributable to specific cause (substance use/abuse, auditory disease), refusal to sign release, disclosure, and consent forms.
Table 1
Randomized Groups
| Control 1 (N=30) | Control 2 (N=30) | Experimental Group 3 (N=30) |
| Maintain Conventional Treatment, if any | Maintain Conventional Treatment, if any | Maintain Conventional Treatment, if any |
| SADS-C | SADS-C | SADS-C |
| Visual Analogue Scale | Visual Analogue Scale | Visual Analogue Scale |
| Interview/Placebo | Interview/Remedy |
Note 1: The SADS-C will be given for baseline, mid-study, and end-of study assessments.
Note 2: Placebo and remedy will be administered according to the precepts of classical treatment (minimum dose, then wait and watch).
Note 3: Groups 2 and 3 (N=60) will be randomly assigned to one of six homeopaths for the interviewing process.
Random Allocation
Viera & Bangdiwala (2007) recommend multiple ways to eliminate bias in RCT research, emphasizing the critical importance of maintaining secrecy in allocation and masking participants and researchers inasmuch as possible. Accordingly, there will be two copies of the randomly generated group allocation assignments, one to be retained by the research supervisor and the other to be retained by the pharmacy coordinator. These will be wrapped in foil and sealed in an opaque envelope with a wax seal applied. The pharmacy coordinator will break the seal and use the list at his/her location to dispense placebo or remedy, keeping it locked in a safe at all times other than when dispensing. The other copy will remain sealed unless there is an emergency requiring knowledge of group allocation for a specific participant.
SADS-C Assessments
Three psychologists (P1, P2, and P3) formally trained in the use of the SADS-C assessment tool are required, with one additional stand-by in the event a primary member must withdraw. These individuals will remain blind to the study and will only know they are conducting SADS-C to document patient symptoms. They will be rotated so that each interviews a participant only one time (each participant will be assessed by each psychologist for a total of three assessments.). Assessors will be remunerated for their assistance in a manner consistent with the average wage for their profession and travel expenses reimbursed. Exclusion: Psychologist cannot show a record of formal training for administering the SADS-C.
Statistician
A statistician will be retained and compensated to analyze all results. The statistician will remain blind to the entire study to prevent bias, will assist with selection and randomizing allocation, random assignment of homeopathic interviews, and will be provided coded study results for analysis.
Coding and Administrative Team
A coding and administrative team of five will be employed to catalogue and code data. The coding team will remain blinded to remedy/placebo prescriptions. This team will also be used for the secondary study to deconstruct video interviews and synthesize into homeopathic rubrics using classical methods (Kent, 2002). Members will be experienced homeopaths or students of homeopathy. Rubric assignment will be by consensus. Once completed, rubrics will be compared to those known for each remedy prescribed during the study, and results analyzed to determine “likeness” of remedy selection to participant symptoms.
Interviewers and Placebo/Remedy Assignment
Six professional homeopaths are required to conduct interviews, with two additional on stand-by in the event a primary member must withdraw. A professional homeopath is defined as one who has completed formal training from an acknowledged school of classical homeopathy (founded on the principles of Hahnemann), is currently in practice, and is certified/licensed according to the requirements of their geographic location). All will remain blinded to whether participants are from the placebo or experimental group in an effort to limit bias. Homeopaths will be randomly assigned placebo and treatment participants for individual consultation using classical methods in video-recorded sessions and will follow the participant for the duration of the study (each homeopath will have 10 participants for interviewing). Recorded sessions will be analyzed by panel, then remedy and potency selection determined by consensus exclusive of the consulting homeopath (see example in Table 2).
Table 2
Homeopath Interviewing and Panel Assignments
| Interviewing Homeopath | Analysis and Prescription Panel |
| H1 | H2, H3, H4, H5, H6 |
| H2 | H1, H3, H4, H5, H6 |
| H3 | H1, H2, H4, H5, H6 |
| H4 | H1, H2, H3, H5, H6 |
| H5 | H1, H2, H3, H4, H6 |
| H6 | H1, H2, H3, H4, H5 |
Using this method should alleviate pressure to prescribe, commonly confounding to homeopathic research (Bell, 2005; Jacobs et al., 2005) and permit the interviewer to focus only on obtaining symptom information through the classical process of case-taking.
Prescriptions from the panel will be given to an on-site research administrator who will submit the order to a pharmacy coordinator (this person will not be blinded to group allocation) at an unidentified location via generic email address, who will then dispense ordered remedies (coded so as not to reveal identity) to experimental group participants and substituting placebo to control group 2 participants. The pharmacy coordinator will maintain a coded catalogue of all remedies/placebo prescribed and by whom, date of prescription receipt, date of dispatch, and to whom. Homeopaths and pharmacy coordinator will be remunerated in a manner consistent with the average wage for their profession.
Trial Schedule
The schedule is outlined in Table 3. An initial thirty day period will be allocated for baseline SADS-C and homeopathic interviews to accommodate time requirements for initial intake of sixty participants.
Table 3
Trial Schedule
| Control 1 | Control 2 | Experimental Group | |
| Day 1 | SADS-C Baseline (P1) | SADS-C Baseline (P2) | SADS-C Baseline (P3) |
| Interview Period | Interview Period | ||
| Day 30 | Dispense Placebo | Dispense Remedy | |
| Day 60 | SADS-C Mid (P2) | SADS-C Mid (P3) | SADS-C Mid (P1) |
| Interview Period | Interview Period | ||
| Day 80 | Dispense Placebo | Dispense Remedy | |
| Day 110 | SADS-C End (P3) | SADS-C End (P1) | SADS-C End (P2) |
| Interview Period | Interview Period | ||
| Day 130 | Exit Interview | Exit Interview | Exit Interview |
Note 1: During the second interview period, a remedy will only be dispensed if it is determined necessary by panel consensus within the precepts of homeopathic treatment.
Note 2: During the third interview period, remedy decisions will be sent to the pharmacy coordinator who will not act on them.
The study will conclude with a twenty day period for end-of-study SADS-C, follow-up homeopathic interviews, and an exit interview for each participant during which they will be asked whether they wish to continue with homeopathic treatment. If so, those already in the interviewing process will receive the remedy recommended by consensus during the third interview period and will be scheduled for follow-up treatment, while those allocated to control group 1 will be scheduled for intake.
Results
Results will be compiled and reported in a variety of ways following analysis by the statistician. SADS-C results will be compared both within and between groups to determine whether statistically significant differences in progress and outcome exist between the conventional, placebo, and homeopathic treatment groups and/or between assessment periods (baseline, mid-study, end-of-study) for each participant within each group.
Remedy and potency decisions from the placebo and experimental groups will be analyzed and tabulated to show any changes and for what reason (for instance, change in remedy choice, change in remedy potency, determination to ‘wait and watch”, etc.). Videos will be deconstructed as previously described and results analyzed against remedy selections to determine “likeness” to patient symptoms and then to determine whether a “more alike” remedy exists in the homeopathic material medica (this will be conducted through Radar and MacRepertory homeopathic software analyses). Through statistical and repertorial analysis, this study aims to answer the following questions:
- Is there a statistically significant difference in treatment progress and outcomes between homeopathy vs. placebo, homeopathy vs. conventional treatment, and/or placebo vs. conventional treatment as measured by the SADS-C and visual analogue scale assessment information gathered at baseline, mid-study, and end-of-study?
- What difference is there between the assessment results when grouped by the psychologist conducting the assessment? This question seeks to explore the consistency of SADS-C assessment methodology between psychologists.
- What difference is there between homeopaths’ prescriptions and results when grouped by case-taker and analysis panel? This question seeks to explore the consistency of case-taking and analysis methodology between homeopaths and analysis panels.
- What is the degree of “likeness” between prescribed remedies and participant symptoms when interviews are deconstructed and repertorized?
- What changes were made to remedy prescriptions and for whom? This question seeks primarily to explore the degree to which interviewers accessed clinical progress in the group receiving placebo and for what reason changes in remedy and/or potency were prescribed.
- Are there other remedies that can be considered more “alike” than those prescribed when deconstructed interviews are analyzed against rubrics of the homeopathic material medica?
References
Bell, I. R. (2005). All evidence is equal, but some evidence is more equal than others: Can logic prevail over emotion in the Homeopathy debate? The Journal of Alternative and Complimentary Medicine, 11(5), 763 – 769.
Brein, S., Lachance, L., Prescott, P., McDermott, C. & Lewith, G. (2011). Homepathy has clinical benefits in rheumatoid arthritis patients that are attributable to the consultation process but not the homeopathic remedy: A randomized controlled clinical trial. Rheumatology, 50(11), 1070 – 1082.
Hahnemann, S. (1843). Organon of the medical art, 6th Ed. O’Reilly, W. B., Decker, S. eds. Redmond, WA: Birdcage Books.
Jacobs, J., Williams, A., Girard, C., Njike, V. Y., & Katz, D. (2005). Homeopathy for attention-deficit/hyperactivity disorder: A pilot randomized-controlled trial. The Journal of Alternative and Complimentary Medicine, 11(5), 799 – 806.
Kent, J. T. (2002). Repertory of the homeopathic materia medica and a word index. Reprint ed. New Delhi, India: B. Jain Publishers.
Oberbaum, M., Vithoulkas, G., & Van Haselen, R. (2003). Clinical trials of classical Homeopathy on appropriate research designs. The Journal of Alternative and Complimentary Medicine, 9(1), 105 – 111.
Reilly, D. & Taylor, M. A. (1994). Is evidence for homeopathy reproducible? Lancet, 344(8937), 1601 – 1607.
Rhutten, A. L. B. (2009). The 2005 Lancet review proved superior quality of Homeopathy trails: what’s next? International Journal of High Dilution Research, 8(28), 110 – 118.
Schmidt, P. (2004). The homeopathic consultation: The art of interrogation. New Delhi, India: B. Jain Publishers.
Shang A., Huweiler-Muntener K., Narty L, Juni, P., Dorig, S., Sterne, J. A. C., & Egger, M. (2005). Are the clinical affects of homeopathy placebo affects? Lancet, 366, 726 – 732.
Ullman, D, & Frass, M. (2010). A review of Homeopathic research in the treatment of respiratory allergies. Alternative Medicine Review, 15(1), 48 – 58.
Viera, A. J., Bangdiwala, S. I. (2007). Eliminating bias in randomized controlled trials: Importance of allocation concealment and masking. Family Medicine, 39(2), 132 – 137.
Appendix
Demographics of Schizophrenia
Retrieved November 22, 2012, from http://www.minddisorders.com/Py-Z/Schizophrenia.html
In the United States, Canada, and Western Europe, the sex ratio in schizophrenia is 1.2:1, with males being affected slightly more often than females. There is a significant gender difference in average age at onset, however; the average for males is between ages 18 and 25, whereas for women there are two peaks, one between ages 25 and 35, and a second rise in incidence after age 45. About 15% of all women who develop schizophrenia are diagnosed after age 35. In some women, the first symptoms of the disorder appear postpartum (after giving birth). Many women with schizophrenia are initially misdiagnosed as having depression or bipolar disorder, because women with schizophrenia are likely to have more difficulties with emotional regulation than men with the disorder. In general, however, females have higher levels of functioning prior to symptom onset than males.
The incidence of schizophrenia in the United States appears to be uniform across racial and ethnic groups, with the exception of minority groups in urban neighborhoods in which they are a small proportion of the total population. A recent study done in the United Kingdom replicated American findings: there are significantly higher rates of schizophrenia among racial minorities living in large cities. The rates of schizophrenia are highest in areas in which these minority groups form the smallest proportion of the local population. The British study included Africans, Caribbeans of African descent, and Asians.
The incidence of schizophrenia in most developed countries appears to be higher among people born in cities than among those born in rural areas. In addition, there appears to be a small historical/generational factor, with the incidence of schizophrenia gradually declining in later-born groups. Schizophrenia is a leading cause of disability, not only in the United States, but in other developed countries around the world. In 1997, the World Health Organization (WHO) listed schizophrenia as the world’s ninth leading cause of disability. According to the National Institute of Mental Health (NIMH), 2.2 million American adults, or 1.1% of the population over age 18, suffer from schizophrenia. Other estimates run as high as 1.5% of the population.
Schizophrenia is disproportionately costly to society for reasons that go beyond the sheer number of people affected by the disorder. Although patients with schizophrenia are little more than 1% of the population, they account for 2.5% of all health care costs—$40 billion per year in the United States, $2.35 billion in Canada (in Canadian dollars), and 2.6 billion pounds sterling (about $7.28 billion in US dollars) in Great Britain. In the United States, patients with schizophrenia fill 25% of all hospital beds and account for about 20% of all Social Security disability days.
In addition, the onset of the disorder typically occurs during a young person’s last years of high school or their first years in college or the workforce; thus it often destroys their long-term plans for their future. According to the federal Agency for Healthcare Research and Quality, 70%–80% of people diagnosed with schizophrenia are either unemployed or underemployed (working in jobs well below their actual capabilities). Ten percent of Americans with permanent disabilities have schizophrenia, as well as 20%–30% of the homeless population.
[1] At www.Random.org, http://www.random.org/sequences/. A minimum value of 1 and a maximum value of 90 formatted into three columns will provide random assignment of participants into three groups.
[2] Those in the non-treatment group who will maintain their current status will obviously recognize they are not receiving additional treatment. At the end of the study, these individuals will be invited to enter into homeopathic treatment if they wish.
