Do homeopathic pathogenetic trials generate recognisable and reproducible symptom pictures?Results from a pilot pathogenetic trial of Ozone 30c - homeopathy360
Articles Research

Do homeopathic pathogenetic trials generate recognisable and reproducible symptom pictures?Results from a pilot pathogenetic trial of Ozone 30c

Background:  Homeopathic Pathogenetic Trials (HPTs) are a pillar of homeopathy, a key source of the symptoms characteristic of a particular homeopathic medicine. Homeopaths choose homeopathic medicines by comparing these remedy pictures with the symptoms the patient is presenting. Thus, recognition of these symptom sets underpins he clinical practice of homeopathy.
Objective: To test whether HPTs generate consistent and recognisable sets of symptoms in consecutive trials.
Design:  Practising homeopaths, blinded to the homeopathic medicine under investigation, were given the set of symptoms generated during an unpublished HPT and asked to identify the homeopathic medicine used.
Homeopathic trial substance: Ozone, prepared by homeopathic method to the ultra-molecular dilution of 30c (1060dilution), was chosen at random from twenty potential medicines.
Results: Seven practising homeopaths were asked to make three guesses as to the identity of the remedy. Initially from the full list of possible remedies (N= 2372). Two of the seven homeopaths guessed the identity of the remedy correctly (p< 0.0001). Subsequently, when their choice of possible medicines was restricted to a list of 20, the same two homeopaths selected the correct medicine, however none of the other practising homeopaths did so (p= 0.2).
Discussion: The selection of the correct homeopathic medicine from the unrestricted list (N= 2372 medicines) by two homeopaths is noteworthy given that the homeopathic medicine used during the HPT was diluted well beyond Avogadro’s number and would not be expected to produce any detectable or recognisable symptomatology. Possible reasons why the remaining five homeopaths did not guess correctly are discussed.
Conclusion: The results show that practising homeopaths may be able to correctly identify a homeopathic medicine from the set of symptoms generated during an HPT. This suggests that such symptom pictures generated by taking an ultramolecular homeopathic medicine are recognisable and specific to the substance taken. Since identification of the remedy was based on past HPT information held in the materia medica, this demonstrates that HPT-generated symptom pictures are reproducible, thus validating the HPT methodology. These promising preliminary findings warrant replication; possible improvements to the trial design to be incorporated in future studies were identified. Homeopathy (2014) 103, 108e112.


A Homeopathic Pathogenetic Trial (HPT) (also known as a “proving”) is a clinical study designed to explore the possible clinical applications of a homeopathic medicine. In an HPT, healthy participants take a substance prepared according to the homeopathic pharmacopeia and record any symptoms which develop within a specified time frame after taking the dose; participants are blinded as to the identity of the substance. The set of symptoms generated during an HPT is referred to as the ‘remedy picture’ or ‘symptom picture’ of the particular homeopathic medicine being tested. This information is subsequently collated into the homeopathic materia medica. In the clinical setting, homeopaths choose homeopathic medicines by comparing these remedy pictures with the symptoms presented by the patient. Since Hahnemann’s first proving, HPT information has been entered, along with toxicology and clinical data into homeopathic materia medica documents, books and computer programs. In modern times, protocols for administrating HPTs have been established, influencing most trials conducted since 1994.1
The literature contains few published studies that have xplored the topic of HPTs. In 2000, Vickers et al. conduct-ed an inter-rater reliability study.2Using a database of 557symptoms, including those recorded by placebo-takers, two practitioners were instructed to indicate the symptoms produced by Mercurius solubilis. There was a moderate degree of agreement (70%) between the two practitioners. In2001, Vickers et al. studied the ability of homeopathic practitioners to detect a homeopathic medicine experien-tially.3
Participants, randomized and blinded, took either a well-known homeopathic medicine (Bryonia alba) or a placebo. Based on the symptoms they experienced, the participants were to determine if they had taken verum or placebo. In this pilot study 60% of the homeopaths correctly identified when they had taken the verum.
In 2004 Walach et al. performed an HPT of Cantharisvesicatoria.4The final symptoms collected during the trial were rated by a homeopathic expert as typical or not of Cantharis. The results showed that the symptoms produced in the verum group were more typical of Cantharis(p= 0.02) than placebo. In 2008 Walach et al. expanded their 2004 study to test for specific and non-specific symptoms.5 Two HPTs were conducted in parallel: the first had two arms (Ozone and placebo), the second study had three arms (Ozone,Iridium and placebo). Using computerized materia medica software, experts were asked to determine which symptoms were typical (specific) or non-specific for the homeopathic medicine tested. When combined, the results of the two studies indicated that the group taking Ozone showed significantly more symptoms typical of the remedy than the control group (p= 0.011).
In 2009 Mollinger et al. performed an HPT with three arms: two well-characterised homeopathic medicines (Natrum muriaticum 30C and Arsenicum 30C) and placebo.6 An expert was then asked to label each symptom generated during the trial as typical of Natrum muriaticum, typical of Arsenicum or non-specific. The results of the study showed that healthy volunteers exhibit significantly more symptoms specific to the homeopathic medicine they had ingested compared to placebo (p< 0.001).
The same year, Piltan et al. published the results of their HPT of Aconitum Napellus 30C using a cross-over doubleblind randomised placebo-controlled trial design.7Anexpert was asked to determine the correct group allocation of the volunteers based on the symptoms they reported. The expert achieved a statistically significant success rate of70.4% (p= 0.004).
A meaningful and relatively unexplored question is whether multiple HPTs of the same homeopathic medicine produce consistent sets of symptoms. Earlier studies identified issues with comparing symptom sets generated indifferent trials. For instance, how does one equate a symptom such as “head pain” with a modified symptom “head pain, in the morning” or “head pain in the morning after eating?” It is well-known amongst experienced HPT investigators that the retrials of homeopathic medicines do not al-ways produce identical symptoms, even in retrials with the same participants.
Yet, in a clinical setting, homeopaths routinely and effectively assess the similarity of non-identical symptom sets (comparing those of the patients and the remedies).The best explanation for how homeopaths accomplish this is that the set of symptoms produced in HPTs createa meaningful totality or pattern of symptoms that is sufficiently characteristic as to be recognisable by a homeopath.
Therefore, to answer the question of whether multiple HPTs of the same homeopathic medicine produce consistent symptom pictures, a new approach was conceived for the present study. Homeopathic practitioners would be asked to identify a homeopathic medicine from the set of symptoms it generated during an unpublished HPT. If the practising homeopaths ea priori aware of the results of previously published HPTs are able to correctly identify the medicine, this would mean that symptom pictures generated by different HPTs carry qualitatively similar in-formation, characteristic of the medicine.


Replication of previous HPT
The first HPT of Ozone was performed in 1993 by Schadde and co-workers. The results were published in19978and the corresponding set of symptoms was then added to the homeopathic literature and computerised materia medica. The present pilot study used the symptoms drawn from a more recent HPT of Ozone conducted by Walach et al., the results of which were published in 2008.5Unblinding of the Walach et al. HPT was delayed until after the present study was conducted, so as to preserve the blinding required for the present study. The full list o fsymptoms from the Walach et al. 2008 replication has not yet been published, nor included in the materia medica.
A total of 53 participants took part in the replicated trial,17 were novice volunteers from UK/Canada and 36 were experienced volunteers from Freiburg, Germany. The Ozone HPT replication was part of a blinded, randomised, placebo-controlled parallel study of two previously tested homeopathic medicines, Ozone and Iridium. 21 participants had Ozone (10 in the UK/Canada group, 11 in the Freiburg group) while 20 participants had placebo (7 in the UK/Canada group, 13 from the Freiburg group); the remaining 12 participants in the Freiburg group had Iridium(not relevant to the present study). The UK/Canada groupcontained student homeopaths naive to HPTs while the Freiburg group had experienced participants. The set of symptoms obtained for Ozone in the trial was made avail-able before publication for use in the present pilot study.
Homeopaths taking part in the study
The PI (JS) invited seven practising homeopaths, one man and six women who had homeopathic experience ranging from two to twenty years, to participate in this study. All but one were known to the PI as well-established teachers of homeopathy. They were chosen based on their experience and reputation for expertise in homeopathic materia medica and practice. The practicing homeopaths and the PI were blind to the name of the correct medicine. The name of the correct remedy for both the Wallach et al. 2008 study and the present study wererevealed only after both studies had been completed.
Medicine identification
The practising homeopaths were informed that the study was based on a recent HPT of a previously trialled homeopathic medicine. They were emailed the full set of 1439symptoms from the HPT of Ozone conducted by Walachet al.5Symptoms from placebo-takers were included within the data given to the practising homeopaths (20 participants had placebo, 21 had Ozone) and the practising homeopaths were informed of this fact. At this stage the trial codes had not yet been revealed. Had the codes been un-locked to remove the placebo symptoms, blinding of the Walach et al. trial. would have been jeopardized.
Identification of the medicine by the practising homeo-paths was undertaken in two steps. In the first phase they were given an Excel sheet containing all 1439 symptoms (including placebo symptoms), pre-sorted according totheir homeopathic classification (Mind, Dreams, Head, Stomach etc). It is worth noting that the symptoms had not been converted into materia medica symptoms or rep-ertory rubrics and hence required a deeper level of interpretation by the practising homeopaths. The practicing homeopaths were asked to study the symptoms and identify the most likely homeopathic medicine that could have produced the set of symptoms. They were allowed to use any tools they normally use in practice, i.e. books and computerised materia medica and repertories. They were instructed to make three choices from the unrestrictedtotal of homeopathic medicines known to date and listed in the materia medica (N= 2372).9In a second phase, having handed in their answers, and having received no feedback as to accuracy of their choices, the practising homeopaths were then instructed to make another three attempts at naming the medicine from a restricted list of twenty homeopathic medicines.
Selection of medicine
The twenty medicines on the restricted list were newly-developed medicines that were fully represented in the homeopathic literature. Table 1 depicts the restricted list of the twenty medicines. The PI chose the initial list of twenty homeopathic medicines but was blind to the one selected for the trial. Ozone was randomly selected from these twenty medicines by the pharmacist, who was the only per-son in the study to know the medicine. The name of the medicine was revealed only at the end of this study.


Unrestricted choice
When asked to select the homeopathic medicine from the unrestricted list of all available homeopathic medicines (N= 2372), two practising homeopaths were successful (numbers 2 and 7, see Table 2) in identifying Ozone. The probability of one individual choosing the correct homeopathic medicine from the full list within three attempts isp= 0.0013. The probability of two individuals selecting the correct homeopathic medicine is calculated from the binomial distribution to be p= 0.000033, which is statistically highly significant.

Restricted choice
When asked to select the medicine from a restricted list of twenty medicines, the same two practising homeopaths (number 2 and 7, see Table 3), selected the correct medicine. The probability of one homeopath choosing the correct answer from the restricted list within three attemptsis p= 0.16, which is not statistically significant. The probability of two homeopaths selecting the correct medicine from the restricted list is p = 0.3, which is also not statistically significant. On the other hand the medicine Haliaeetus leucocephalus sanguinaria was selected four times from the restricted list – an interesting and statistically significant result (p= 0.01) discussed below.


In this study, the correct homeopathic medicine was identified far beyond the possibility of mere chance. Two homeopaths succeeded in determining the correct medicine out of the 2372 possible medicines, which is a statistically highly significant result (p< 0.0001). This demonstrates that HPTs generate specific and recognizable sets of symptoms. If the symptoms generated by HPTs were merely random placebo effects, identifying the correct medicine would not be possible as there would be no pattern to recognise.
It is interesting to note that the homeopathic medicine, Ozone 30c, used in Wallach’s HPT was diluted beyond Avogadro’s number – the point where theoretically no molecule of the original substance remains. This means that the symptom picture generated could not have been due to any material effect of the Ozone, thus demonstrating a physiological effect of ultramolecular homeopathic dilutions.
Given the complications inherent in the present study, it is not entirely unexpected that five out of the seven practising homeopaths were unable to recognize Ozone. Firstly, the set of symptoms given to homeopaths included not only symptoms from the Ozone group but also those coming from the placebo group. Adding these random symptoms made it harder to identify the correct medicine. Secondly, Ozone is a relatively new addition to the homeopathic pharmacopeia and so the homeopaths would be less familiar with its symptom picture. In the case of the restricted choice, we see that the most often selected medicine was not the correct one, Ozone, but Haliaeetus leucocephalus sanguinaria (American bald eagle’s blood). This medicine was selected three times in the unrestricted choices (highly significant, p< 0.0001) and four times from the restricted list (p= 0.01). There is a well-known overlap of themes in homeopathic medicines developed from birds and from gases, which make them seem superficially alike. The HPT databases of Ozone and Haliaeetus leucocephalus sanguinaria share the themeof “upward” or floating sensations, such as feeling light andfree, feelings of elation and euphoria and dreams of being high up in the mountains.8,10 Several gas provings have references to seeing or feeling like birds.9
Furthermore, the PI is well-known in the field for con-ducting numerous HPTs of new substances, in particular for his HPT of Haliaeetus leucocephalus sanguinaria in1997.10
Awareness of the PI’s association with newly-developed homeopathic medicines may have caused the practising homeopaths to be somewhat predisposed to choose a modern proving or one of the PI’s provings, such as Haliaeetus leucocephalus sanguinaria. It is interesting to note that only four out of the ten medicines chosen by the practising homeopaths in the unrestricted choice were classic polychrest homeopathic medicines already known in the 19th century, possibly indicating that such abias occurred.
Let us consider further the hypothesis that the practicing homeopaths were biased towards a modern homeopathic medicine due to the known involvement of the PI in modern HPTs. According to (a database of provings in which the PI is involved), there were 600 modern HPTs registered at the time of the study. If were strict the number of possible medicines to this list of 600 modern HPTs, the p-value corresponding to this configuration remains highly significant  (p< 0.001). Even when considering the case where the possible list of medicines was only 150 of the modern medicines, the results would remain significant  (p< 0.01).
Modern and older HPTs can be distinguished: modern HPTs tend to generate more symptoms and more fully developed symptoms, and the description of the symptoms tends to use more modern language. This in principle would allow a homeopath to differentiate between an old and a new proving based on the number of symptoms and language used in the descriptions. However, in this case the practising homeopaths were aware that the data originated from a modern repeat HPT of a known medicine and would expect a large number of symptoms with modern descriptions. They would therefore not be able to infer anything about the medicine from this information.
The mix of experienced and inexperienced HPT volunteers in the replicated HPT may have increased the challenge for the practising homeopaths trying to identify the medicine. Experienced HPT participants generate shorter lists of symptoms and are more specific in their descrip-tions.5Experienced HPT volunteers are more aware of their reactions and better able to discern newly-appeared, and sometimes subtle symptoms, from pre-existing ones. With the inclusion of inexperienced provers, some less-specific symptoms were added to the list rendering the task of the practising homeopaths more difficult.
The results of the present study support the conclusions of the Walach et al. 2004, 2008 and the Mollinger et al.2009 studies, which showed that symptoms typical of the homeopathic medicine occur more frequently when volunteers take verum than placebo.4e6It would be interesting toreplicate the present study using more commonly known and classically proved medicines, as well as experienced HPT participants. Using a larger number of experienced homeopaths to judge the symptoms would improve the significance of the results. We also recommend removing placebo symptoms from the final HPT data.


In this study, some practising homeopaths were able to correctly identify an ultramolecular homeopathic medicine from the symptoms it generated when taken by healthy volunteers during an HPT. Two homeopaths out of seven taking part in the study correctly guessed the remedy, Ozone, out of the N= 2372 remedies in the materia medica. This highly significant result (p< 0.0001) supports the hypothesis that ultramolecular homeopathic dilutions are not ‘simply water’.
Further, this study demonstrates that symptoms generated by taking an ultramolecular homeopathic medicine are recognisable and specific to the substance taken, unlike the random symptoms generated by a placebo. Also, since identification of the remedy was based on past HPT information held in the materia medica, this demonstrates that HPT-generated symptom pictures are reproducible.
Each homeopathic medicine has an associated symptom picture generated primarily through HPTs. In the clinical setting, homeopaths prescribe by selecting the homeopathic medicine with the symptom picture most similar to that of the symptoms presented by the patient.
The homeopathic prescribing process therefore hinges on two factors: the generation of consistent symptom pictures by HPTs and the ability of homeopaths to correctly recognise those symptom pictures. This study tends to validate both these aspects of homeopathic practice.
 Conflict of interest statement
The authors declare no known conflict of interest.
We thank the Samueli Institute, Alexandria, Virginia,USA for sponsoring the Walach et al. 2008 study and repli-cated HPT from which the database for this study wasderived, as well as Harald Walach and Heribert M€ollingerfor sharing the trial database with us. We are indebted toHeribert M€ollinger, Nadia Bakir and Mirsada Vinc for or-ganising the replicated trials; John Morgan and HeliosPharmacy for preparing the homeopathic medicines;Shelly Been for sorting the replicated trial symptoms andMary Jo Aloi, Joanna Daly, Silvie Gowan, Racheli Shabi,Camilla Sherr and Steve Waldstein for participating aspractising homeopaths. We are grateful to Rachel Robertsfor helping with the editing of the paper.

  1. Sherr J. Dynamics and methodology of homoeopathic provings.Malvern, UK: The Proving Fund, 1994.
  2. Vickers AJ, van Haselen RA, Pang L, Berkovitz S. Inter-rater reli-ability of symptom repertorisation: a pragmatic empirical study.Br. Homeopath. J. 2000; 89: 188e190.
  3. Vickers A, McCarney R, Fisher P, van Haselen R. Can homeopathsdetect homeopathic medicines? A pilot study for a randomised,double-blind, placebo controlled investigation of the proving hy-pothesis. Br. Homeopath. J. 2001; 90: 126e130.
  4. Walach H, et al. Homeopathic proving symptoms: result of a local,non-local, or placebo process? A blinded, placebo-controlled pilotstudy. Homeopat. J. Fac. Homeopat. 2004; 93: 179e185.
  5. Walach H, M€ollinger H, Sherr J, Schneider R. Homeopathic patho-genetic trials produce more specific than non-specific symptoms: results from two double-blind placebo controlled trials. J.Psychopharmacol. Oxf. Engl. 2008; 22: 543e552.
  6. Mollinger H, Schneider R, Walach H. Homeopathic pathogenetictrials produce specific symptoms different from placebo. Forsch.Komplement€armed. 2009; 16: 105e110.
  7. Piltan D, Rist L, Sim~oes-W€ust P, Saller R. Test of a homeopathicdilution of Aconitum napellus. A clinical, randomized, double-blind, controlled crossover study in healthy volunteers. Forsch.Komplement€armed. 2009; 16: 168e173.
  8. Schadde A. Ozone: a homeopathic proving. Seattle, Washington: Alethea Book Company, 1997.
  9. Schroyens DF. Synthesis repertorium homeopathicum syntheticum9.1. Belgium: Archibel, 2004.
  10. Sherr J. Dynamic provings. Malvern, UK: Dynamis, 1997.
Jeremy Sherr 1*, Tina Quirk 2 and Alexander L Tournier 3
1 Dynamis School for Advanced Homoeopathic Studies, Worcester, UK
2 Dynamis School for Advanced Homoeopathic Studies, New York, NY, USA
3 Homeopathy Research Institute, London, UK

Do Homeopathic Pathogenetic Trials generate recognisable and reproducible symptom pictures?: Results from a pilot pathogenetic trial of Ozone 30c. Available from: [accessed Apr 21, 2017].

Leave a Comment