Dengue: No Evidence Homoeopathy Works, No Reason Why It Should
by Adrian Philip Gaylard (2009-07-04)
A Flawed Trial
It is clear that no conclusions can be drawn from the reported treatment of 126 symptomatic patients. The author claims that, for these patients, “remission was earlier (average = 5 days), without complications, by comparison to patients who did not use homeopathy (average = 8.6 days)”.
It is notable that means are cited without any measure of variance (such as standard deviations or confidence limits) or statistical significance. In the absence of such data claims of actual differences in remission times cannot be sustained.
In addition, it seems that the trial did not blind either patients or clinicians to the treatment, or use an equivalent placebo in the control group. The allocation of patients to treatment or control groups was not randomised. These methodological flaws mean that the difference in remission times is most likely caused by a combination of bias and expectation effects.
Regression to the Mean
From the data provided in the paper (Figure 2) it is clear that dengue infection rates are cyclical, with an approximately five to six year period: peaks in one year are usually followed by dramatic falls the next year. This means that 2008 was likely to show a substantial fall in dengue infections, even in the absence of the homoeopathy campaign.
Fair Comparisons
As the intervention in Macaé did not include a control group, it is important to ensure that comparisons with different areas or the same location for previous years are fair.
For instance, the author does not disclose whether the comparator regions operated a conventional public health campaign against dengue. If they did not, it is likely that any relative improvement Macaé is attributable to the conventional aspects of this campaign, rather than the novel homeopathic element, such as:
“[...] control of the vector, epidemiological surveillance and training of both health workers and the general population [...] supportive and pharmacological management, as well as clinical follow up [...]“
Neither does the author establish that comparator regions are of similar socio-economic make-up to Macaé.
It also seems peculiar to aggregate reported infections for January to March, when infections during 2007 peaked in April and May. This would appear to run the risk of missing the months with the greatest potential for dengue infection.
An Unremarkable Result
Commenting on the start of the homoeopathy campaign the author states, “From that moment on, a dramatic fall was recorded, not observed in the remaining localities”. And yet the data show that similar “dramatic falls” are not uncommon.
The main claim for control of dengue in the population of Macaé relates to the first three months of 2008. Totalling the data provided in Figure 1 for this period in other years reveals that three of the previous six years had fewer reported dengue cases during these months than the thirty-five claimed for 2008. (2004: 12; 2005: 5 and 2006: 10).
Looking at the data for complete years, Figure 1 reveals that there were only 31 and 17 reported cases of dengue in Macaé during 2004 and 2005, respectively. Also, between 2003 and 2004 dengue incidence fell by 93%. Finally, Figure 2 shows a similar decrease Litoral Sul Fluminense between 2006 and 2007.
From these data the reported dengue infection rates for early 2008 are certainly not without recent precedent.
Unsupported Conclusions
I wholeheartedly agree with the author that, “confounding factors were not controlled for.” However, the methodological flaws identified preclude the conclusion that homoeopathy made any appreciable contribution to public health in this case. This view is strengthened by the paper not addressing itself to more likely alternative explanations, namely: bias, expectation effects, regression to the mean, inappropriate comparisons and natural variation in the incidence of dengue infections.
This view, rather than the author’s conclusions, is compatible with the outcome of the only RCT of homoeopathy for dengue currently listed in PubMed. A remedy containing Aconita, Bryonia, Eupatorium perforatum, Gelsemium, and Rhus toxicodendron was trialed in Honduras by Jacobs et al. [1]. They concluded, “the results […] do not suggest that this combination homeopathic remedy is effective for the symptoms that are characteristic of dengue fever.”
In the absence of any credible supporting evidence for the efficacy of the homeopathic intervention reported, it is surprising that the author appears to see this work as some sort justification for the promotion of, “acupuncture, phytotherapy, […] and phytotherapic pharmacy.” None of these other alternative approaches were tested and phytotherapy directly contradicts many of the central tenets of homeopathy: similia similibus curentur and the ‘law of infinitesimals’, for example.
This paper provides no credible evidence that homoeopathy made any difference to dengue infections in Macaé during the first three months of 2008. The vanishingly small probability that any molecules of the ingredients were actually administered to anyone during this trial remains the best available explanation for this.
References
Jacobs, J., Fernandez, E. A., Merizalde, B., Avila-Montes, G. A., and Crothers, D. The use of homeopathic combination remedy for dengue fever symptoms: a pilot RCT in Honduras. Homeopathy. 2007. 96: 22–26.