In one of his many comments, our friend Iqbal just linked to an article that unquestionably is interesting. Here is its abstract (the link also provides the full paper):
Objective: The objective was to assess the usefulness of homoeopathic genus epidemicus (Bryonia alba 30C) for the prevention of chikungunya during its epidemic outbreak in the state of Kerala, India.
Materials and Methods: A cluster- randomised, double- blind, placebo -controlled trial was conducted in Kerala for prevention of chikungunya during the epidemic outbreak in August-September 2007 in three panchayats of two districts. Bryonia alba 30C/placebo was randomly administered to 167 clusters (Bryonia alba 30C = 84 clusters; placebo = 83 clusters) out of which data of 158 clusters was analyzed (Bryonia alba 30C = 82 clusters; placebo = 76 clusters) . Healthy participants (absence of fever and arthralgia) were eligible for the study (Bryonia alba 30 C n = 19750; placebo n = 18479). Weekly follow-up was done for 35 days. Infection rate in the study groups was analysed and compared by use of cluster analysis.
Results: The findings showed that 2525 out of 19750 persons of Bryonia alba 30 C group suffered from chikungunya, compared to 2919 out of 18479 in placebo group. Cluster analysis showed significant difference between the two groups [rate ratio = 0.76 (95% CI 0.14 – 5.57), P value = 0.03]. The result reflects a 19.76% relative risk reduction by Bryonia alba 30C as compared to placebo.
Conclusion: Bryonia alba 30C as genus epidemicus was better than placebo in decreasing the incidence of chikungunya in Kerala. The efficacy of genus epidemicus needs to be replicated in different epidemic settings.
I have often said the notion that homeopathy might prevent epidemics is purely based on observational data. Here I stand corrected. This is an RCT! What is more, it suggests that homeopathy might be effective. As this is an important claim, let me quickly post just 10 comments on this study. I will try to make this short (I only looked at it briefly), hoping that others complete my criticism where I missed important issues:
- The paper was published in THE INDIAN JOURNAL OF RESEARCH IN HOMEOPATHY. This is not a publication that could be called a top journal. If this study really shows something as revolutionarily new as its conclusions imply, one must wonder why it was published in an obscure and inaccessible journal.
- Several of its authors are homeopaths who unquestionably have an axe to grind, yet they do not declare any conflicts of interest.
- The abstract states that the trial was aimed at assessing the usefulness of Bryonia C30, while the paper itself states that it assessed its efficacy. The two are not the same, I think.
- The trial was conducted in 2007 and published only 7 years later; why the delay?
- The criteria for the main outcome measure were less than clear and had plenty of room for interpretation (“Any participant who suffered from fever and arthralgia (characteristic symptoms of chikungunya) during the follow-up period was considered as a case of chikungunya”).
- I fail to follow the logic of the sample size calculation provided by the authors and therefore believe that the trial was woefully underpowered.
- As a cluster RCT, its unit of assessment is the cluster. Yet the significant results seem to have been obtained by using single patients as the unit of assessment (“At the end of follow-ups it was observed that 12.78% (2525 out of 19750) healthy individuals, administered with Bryonia alba 30 C, were presented diagnosed as probable case of chikungunya, whereas it was 15.79% (2919 out of 18749) in the placebo group”).
- The p-value was set at 0.05. As we have often explained, this is far too low considering that the verum was a C30 dilution with zero prior probability.
- Nine clusters were not included in the analysis because of ‘non-compliance’. I doubt whether this was the correct way of dealing with this issue and think that an intention to treat analysis would have been better.
- This RCT was published 4 years ago. If true, its findings are nothing short of a sensation. Therefore, one would have expected that, by now, we would see several independent replications. The fact that this is not the case might mean that such RCTs were done but failed to confirm the findings above.
As I said, I would welcome others to have a look and tell us what they think about this potentially important study.