HRI Malta 2017-Cutting Edge Research in Homeopathy: HRI's Third International Research Conference in Malta.

The third international conference on "Cutting Edge Research in Homeopathy" organised by the Homeopathy Research Institute (HRI) was held on the inspiring and historic island of Malta from 9th to 11th of June, 2017. One hundred and two abstracts underwent peer review by the HRI Scientific Advisory Committee and external experts to produce the programme of 36 oral presentations and 37 posters, presented by researchers from 19 countries. The 2.5-day programme covered a diverse range of topics, including quantitative and qualitative clinical research, basic research, veterinary research, and provings. These intensive plenary and parallel sessions were interspersed with multiple opportunities for delegates to discuss and exchange ideas, in particular through interactive panel discussions and a pre-conference workshop. The continuing commitment of the homeopathy research community to generate high-quality studies in this rapidly evolving field was clear. In this conference report, we present highlights from this memorable event.

Model validity of randomised placebo-controlled trials of non-individualised homeopathic treatment.

BACKGROUND: The comprehensive systematic review of randomised placebo-controlled trials (RCTs) in homeopathy requires examination of a study's model validity of homeopathic treatment (MVHT) as well as its risk of bias (extent of reliable evidence). OBJECTIVE: To appraise MVHT in those RCTs of non-individualised homeopathy that an associated investigation had judged as 'not at high risk of bias'. DESIGN: Systematic review. METHODS: An assessment of MVHT was ascribed to each of 26 eligible RCTs. Another 49 RCTs were ineligible due to their high risk of bias. MAIN OUTCOME MEASURES: MVHT and the prior risk of bias rating per trial were merged to obtain a single overall quality designation ('high', 'moderate', 'low'), based on the GRADE principle of downgrading. RESULTS: The trials were rated as 'acceptable MVHT' (N = 9), 'uncertain MVHT' (N = 10) and 'inadequate MVHT' (N = 7); and, previously, as 'reliable evidence' (N = 3) and 'non-reliable evidence' (N = 23). The 26 trials were designated overall as: 'high quality' (N = 1); 'moderate quality' (N = 18); 'low quality' (N = 7). CONCLUSION: Of the 26 RCTs of non-individualised homeopathy that were judged 'not at high risk of bias', nine have been rated 'acceptable MVHT'. One of those nine studies was designated 'high quality' overall ('acceptable MVHT' and 'reliable evidence'), and is thus currently the only reported RCT that represents best therapeutic practice as well as unbiased evidence in non-individualised homeopathy. As well as minimising risk of bias, new RCTs in this area must aim to maximise MVHT and clarity of reporting.

Comparison of effectiveness of frequently and infrequently used homeopathic medicines.

RATIONALE: Patients treated with homeopathy may respond to infrequently used or even 'new' medicines. But does the introduction of an unlimited number of (new) medicines improve homeopathy? Do new medicines solve old problems? METHODS: 1. Consensus meetings to evaluate best cases. 2. Patient outcome study in 10 Dutch practices. RESULTS: Good cases are scarce for many medicines, random variance is an important source of uncertainty. 50 Medicines are responsible for 72% of all successful prescriptions. There is no difference in effectiveness of frequently and less frequently used medicines. Confirmation bias is found for a few well-known symptom-medicine combinations. CONCLUSION: 'New' and infrequently medicines are as effective as 'old' frequently used medicines. Improving the use of frequently used medicines is more effective in improving results than seldom used medicines. Large numbers are required and old mistakes should be avoided developing new medicines. A research agenda for improving the use of homeopathic medicines is imperative.

Semi-standardised homeopathic treatment of premenstrual syndrome with a limited number of medicines: feasibility study.

BACKGROUND: Individualised homeopathy involves a large number of possible medicines. For clinical research purposes it is desirable to limit this number, create more consistency between prescribers and optimising the accuracy of prescription. Using a semi-standardised treatment protocol, we aimed to improve homeopathic management of targeted subgroups of women with premenstrual syndrome/symptoms (PMS/S). OBJECTIVES: To design a semi-standardised protocol for individualised prescribing in PMS/S with a limited number of homeopathic medicines, and to explore the feasibility of working with it in daily homeopathic practice. METHODS: With help of an expert panel, homeopathic medicines were selected, as well as predictive symptoms and characteristics (keynotes) for each medicine. With those, we designed a patient questionnaire and a diagnostic algorithm. The patient questionnaire contained 123 questions, representing potential predictive symptoms for 11 homeopathic medicines for PMS/S. The medicines selected (in rank order) were Sep, Nat-m, Lach, Cimic, Lac-c, Puls, Calc, Lil-t, Mag-p, Mag-c, Phos. In a feasibility study 20 homeopathic doctors used the protocol in daily practice. The diagnosis was confirmed by daily rating of pre-defined symptoms during two consecutive menstrual cycles. The acceptability and feasibility of the protocol were evaluated after 3 months follow-up, at which time we also measured changes in premenstrual symptom scores and patient-reported changes in symptoms and general health. RESULTS: The doctors mostly complied with the protocol and valued the computerised diagnostic algorithm as a useful tool for homeopathic medicine selection. 33 patients completed 3 months follow-up. By then, 19 patients still taking the first medicine on the basis of the algorithm. We received valid symptom records of 30 patients. Premenstrual symptom scores dropped by 50% or more in 12 patients and by 30-50% in 6 patients; scores dropped by less than 30% or increased in 12 patients. Recruitment of patients (n=38 in 9 months) proved difficult. Adherence to the diaries and the questionnaire was satisfactory. CONCLUSIONS: It is feasible to use a semi-standardised protocol for individualised homeopathic prescribing in PMS, in daily practice. Its predictive value and the percentage of women with PMS/S helped by the selected medicines remain to be evaluated in further research. In future research, active promotion will be needed to recruit patients.

Opposite repertory-rubrics in Bayesian perspective.

INTRODUCTION: Hitherto entries have been added to a rubric in the repertory when patients responding well to a specific medicine showed the corresponding symptom. Continuing like this, theoretically every medicine will eventually appear in every rubric. METHOD: This becomes clear if we compare opposite symptom-rubrics. Polarity Analysis (PA) subtracts opposite rubrics and has been shown to improve clinical results. CONCLUSION: The source of this problem and the reason for the success of PA are clear from Bayesian perspective. A reliable repertory should be based on Bayesian principles.

Statistical analysis of six repertory rubrics after prospective assessment applying Bayes' theorem.

BACKGROUND: After prospective assessment of six homeopathic symptoms we validated some rubrics of the homeopathic repertory using Bayesian theory. In this paper we introduce statistical arguments for introducing or discarding entries from the repertory. METHODS: 4094 patients entered the prospective study and 4072 prescriptions were evaluated. After translating typeface into Likelihood Ratios (LRs), Confidence Intervals and the probability of existing repertory entries compared to our findings were calculated. OUTCOME: Our assessment yielded 121 relevant results to validate existing repertory entries. Five symptoms could be compared with Kent's original repertory; they have about the same prevalence (range 3.9-6.5%) in the whole population, but the size of the corresponding repertory rubrics varies from 3 to 103 entries. LR assessment reduced the larger rubrics and supplemented the smaller ones. Our results do not correspond with 56% of the existing repertory entries regarding five symptom-rubrics. This result cannot be generalised for the whole repertory.

Diagnostic test evaluation by patient-outcome study in homeopathy: balancing of feasibility and validity.

BACKGROUND: Signs and symptoms can be considered as diagnostic tests, updating prior odds by applying Bayes' theorem. In homeopathy, signs and symptoms guide the doctor when prescribing appropriate medicines but the powers of these indicators are largely based on common experience. OBJECTIVE: We want to know whether it is possible to calculate the power of arguments of signs and symptoms indicating homeopathic medicines, expressed in likelihood ratios. METHODS: An observational patient-outcome study as advocated for diagnostic test evaluation. There was no independent observation of symptoms and results. Instead, observers were extensively trained in assessing each other's cases and the process was regularly monitored. RESULTS: A total of 4072 prescriptions for 4094 patients were recorded. The relation between six clinical symptoms and outcome was calculated. Variance between observers in assessing symptoms and results were considerable. Some indications of confirmation bias were detected by follow-up, and 48 statistically significant likelihood ratios regarding six symptoms were calculated. CONCLUSION: A diagnostic patient-outcome study within homeopathy collecting a large amount of data is demonstrated. Results partly confirm clinical practice at a 95% confidence level. This kind of research could validate knowledge from practical experience.

The conclusions on the effectiveness of homeopathy highly depend on the set of analyzed trials.

OBJECTIVE: Shang's recently published meta-analysis on homeopathic remedies (Lancet) based its main conclusion on a subset of eight larger trials out of 21 high quality trials (out of 110 included trials). We performed a sensitivity analysis on various other meaningful trial subsets of all high quality trials. STUDY DESIGN: Subsets were defined according to sample size, type of homeopathy, type of publication, and treated disease/condition. For each subset, we estimated the overall odds ratios (ORs) from random effect meta-analyses. RESULTS: All trials were highly heterogeneous (I2=62.2%). Homeopathy had a significant effect beyond placebo (OR=0.76; 95% CI: 0.59-0.99; p=0.039). When the set of analyzed trials was successively restricted to larger patient numbers, the ORs varied moderately (median: 0.82, range: 0.71-1.02) and the P-values increased steadily (median: 0.16, range: 0.03-0.93), including Shang's results for the eight largest trials (OR=0.88, CI: 0.66-1.18; P=0.41). Shang's negative results were mainly influenced by one single trial on preventing muscle soreness in 400 long-distance runners. CONCLUSIONS: The meta-analysis results change sensitively to the chosen threshold defining large sample sizes. Because of the high heterogeneity between the trials, Shang's results and conclusions are less definite than had been presented.

New repertory, new considerations.

The criteria for entering medicines in repertory rubrics are unclear and partly incorrect. A new repertory should be based on clear and objective criteria. Retrospective and prospective assessment of medicines and symptoms by the Dutch Committee for Methods and Validation gives an indication of the validity of existing repertory entries. Reliance on the experience of one expert is the cause of part of the shortcomings of the repertory. This experience is highly influenced by chance. Another part is due to the use of absolute rather than relative occurrence of symptoms. Yet another part is caused by not comparing prevalence in the population 'cured' by a medicine with the remainder of the population. In many cases we need better definitions of symptoms. A clear protocol and prospective research could overcome many shortcomings of the repertory. Statistics help to get more objective criteria, but we still need to reach consensus about how to handle probabilities of outcomes of our assessments.

The 2005 meta-analysis of homeopathy: the importance of post-publication data.

BACKGROUND: There is a discrepancy between the outcome of a meta-analysis published in 1997 of 89 trials of homeopathy by Linde et al and an analysis of 110 trials by Shang et al published in 2005, these reached opposite conclusions. Important data were not mentioned in Shang et al's paper, but only provided subsequently. QUESTIONS: What was the outcome of Shang et al's predefined hypotheses? Were the homeopathic and conventional trials comparable? Was subgroup selection justified? The possible role of ineffective treatments. Was the conclusion about effect justified? Were essential data missing in the original article? METHODS: Analysis of post-publication data. Re-extraction and analysis of 21 higher quality trials selected by Shang et al with sensitivity analysis for the influence of single indications. Analysis of comparability. Sensitivity analysis of influence of subjective choices, like quality of single indications and of cut-off values for 'larger samples'. RESULTS: The quality of trials of homeopathy was better than of conventional trials. Regarding smaller trials, homeopathy accounted for 14 out of 83 and conventional medicine 2 out of 78 good quality trials with n<100. There was selective inclusion of unpublished trials only for homeopathy. Quality was assessed differently from previous analyses. Selecting subgroups on sample size and quality caused incomplete matching of homeopathy and conventional trials. Cut-off values for larger trials differed between homeopathy and conventional medicine without plausible reason. Sensitivity analyses for the influence of heterogeneity and the cut-off value for 'larger higher quality studies' were missing. Homeopathy is not effective for muscle soreness after long distance running, OR=1.30 (95% CI 0.96-1.76). The subset of homeopathy trials on which the conclusion was based was heterogeneous, comprising 8 trials on 8 different indications, and was not matched on indication with those of conventional medicine. Essential data were missing in the original paper. CONCLUSION: Re-analysis of Shang's post-publication data did not support the conclusion that homeopathy is a placebo effect. The conclusion that homeopathy is and that conventional is not a placebo effect was not based on comparative analysis and not justified because of heterogeneity and lack of sensitivity analysis. If we confine ourselves to the predefined hypotheses and the part of the analysis that is indeed comparative, the conclusion should be that quality of homeopathic trials is better than of conventional trials, for all trials (p=0.03) as well as for smaller trials (p=0.003).

How can we change beliefs? A Bayesian perspective.

How can Randomised Controlled Trials (RCTs) change our beliefs? The fact that they do update prior beliefs to different posterior beliefs is explained by Bayesian philosophy. Crucial points in Bayesian analysis include setting the first prior expectation right and sequential updating of the prior in the light of new evidence. Bayesian analysis depends highly on the evidence included. RCT evidence can only falsify the placebo hypothesis, it cannot indicate which mechanism of action could be responsible for an intrinsic effect and therefore cannot overturn existing beliefs. Bayesian reasoning could structure further discussion, but subjectivity is an inherent element of this process. In the case of homeopathy the first prior is not a common prior shared by all parties to the debate, but a paradigm, this prevents common updating of beliefs. Only by keeping an open mind towards other paradigms and all possible hypotheses can a low Bayesian prior be elevated to the point of accepting a new paradigm, this is more relevant than Bayesian calculations.

Bayesian homeopathy: talking normal again.

Homeopathy has a communication problem: important homeopathic concepts are not understood by conventional colleagues. Homeopathic terminology seems to be comprehensible only after practical experience of homeopathy. The main problem lies in different handling of diagnosis. In conventional medicine diagnosis is the starting point for randomised controlled trials to determine the effect of treatment. In homeopathy diagnosis is combined with other symptoms and personal traits of the patient to guide treatment and predict response. Broadening our scope to include diagnostic as well as treatment research opens the possibility of multi factorial reasoning. Adopting Bayesian methodology opens the possibility of investigating homeopathy in everyday practice and of describing some aspects of homeopathy in conventional terms.

A Bayesian perspective on the reliability of homeopathic repertories.

An interim assessment of likelihood ratio of homeopathic symptoms shows that there are serious flaws in Kent's repertory. The system of expressing relationship between symptoms and expected results from medicines by type-face is unclear and unreliable. Bayesian methods can improve this and expand the possibilities of the repertory including the possibility of interpreting the absence of expected symptoms.

Repertory and the symptom loquacity: some results from a pilot study on likelihood ratio.

Treatment outcome in a pilot study indicates that it is possible to assess likelihood ratios of homeopathic symptoms. Entries in repertory rubrics can be validated, but must still be handled carefully. Prospective research is the only acceptable way. Software to support this research must be carefully designed to export correct data.

Repertory and likelihood ratio: time for structural changes.

If the likelihood ratio (LR) method is introduced, the repertory will gradually change as more symptoms are assessed. It will also change the use of the repertory: the most important medicines of each symptom rubric can be identified and relied on, even in large rubrics. This is also a good opportunity to correct structural shortcomings of the repertory, for instance, entries should be based on systematic analysis of materia medica instead of casual observations.

Assessing likelihood ratio of clinical symptoms: handling vagueness.

Clinical symptoms including homeopathic symptoms are often vague. There is reluctance to assess clinical symptoms as diagnostic instruments because they are hard to define. Still, clinical symptoms appear effective in daily practice. Expert systems and neural networks handle vague data successfully. Theoretical considerations predict the kind of problems we may expect. There is a difference between quantitative and qualitative vagueness. Vague data cause problems if we try to prove a hypothesis because of expectation bias. We assess likelihood ratio of homeopathic symptoms only to improve the method.

Is assessment of likelihood ratio of homeopathic symptoms possible? A pilot study.

A pilot study was performed to investigate the possibilities and restrictions of likelihood ratio (LR) investigation using three symptoms. Qualitative vagueness and expectation bias is inherent in our method, but is, in part avoidable. It appears that experienced observers assess common homeopathic symptoms quite similarly. Clinical judgement is an essential part of our work and should be preserved during assessment of LR. The assessment does not influence clinical practice and can be maintained for a long period, provided the appropriate software is used. A limited range of symptoms seems most suitable for LR investigation.

Improving homeopathic prescribing by applying epidemiological techniques: the role of likelihood ratio.

A committee of the Dutch Association of Homeopathic Physicians is trying to validate materia medica by evaluating successful cases. These cases are presented and assessed by a group of experienced homeopathic physicians to provide indications about the prevalence of symptoms related to particular homeopathic medicines. The next logical question is whether epidemiological techniques can be applied to them. We have some ideas concerning the information these data can provide, based on epidemiological theory and limited experimental data. Theoretical investigation suggests that the epidemiological concept of likelihood ratio is well adapted to homeopathy. Researching and applying likelihood ratio will lead to more accurate materia medica and repertory. These considerations already indicate some shortcomings in the representation of rare remedies and the use of grading in the current repertories.