There are dozens of observational studies of homeopathy which seem to suggest – at least to homeopaths – that homeopathic treatments generate health benefits. As these investigations lack a control group, their results can be all to easily invalidated by pointing out that factors like ‘regression towards the mean‘ (RTM, a statistical artefact caused by the phenomenon that a variable that is extreme on its first measurement tends to be closer to the average on its second measurement) might be the cause of the observed change. Thus the debate whether such observational data are reliable or not has been raging for decades. Now, German (pro-homeopathy) investigators have published a paper which potentially could resolve this dispute.
With this re-analysis of an observational study, the investigators wanted to evaluate whether the observed changes in previous cohort studies are due to RTM and to estimate RTM adjusted effects. SF-36 quality-of-life (QoL) data from a cohort of 2827 chronically diseased adults treated with homeopathy were reanalysed using a method described in 1991 by Mee and Chua’s. RTM adjusted effects, standardized by the respective standard deviation at baseline, were 0.12 (95% CI: 0.06-0.19, P < 0.001) in the mental and 0.25 (0.22-0.28, P < 0.001) in the physical summary score of the SF-36. Small-to-moderate effects were confirmed for most individual diagnoses in physical, but not in mental component scores. Under the assumption that the true population mean equals the mean of all actually diseased patients, RTM adjusted effects were confirmed for both scores in most diagnoses.
The authors reached the following conclusion: “In our paper we showed that the effects on quality of life observed in patients receiving homeopathic care in a usual care setting are small or moderate at maximum, but cannot be explained by RTM alone. Due to the uncontrolled study design they may, however, completely be due to nonspecific effects. All our analyses made a restrictive and conservative assumption, so the true treatment effects might be larger than shown.”
Of course, the analysis heavily relies on the validity of Mee and Chua’s modified t-test. It requires the true mean in the target population to be known, a requirement that seldom can be fulfilled. The authors therefore took the SF-36 mean summary scores from the 1998 German health survey as proxies. I am not a statistician and therefore unable to tell how reliable this method might be (- if there is someone out there who can give us some guidance here, please post your comment).
In order to make sense of these data, we need to consider that, during the study period, about half of the patients admitted to have had additional visits to non-homeopathic doctors, and 27% also received conventional drugs. In addition, they would have benefitted from:
- the benign history of the conditions they were suffering from,
- a placebo-effect,
- the care and attention they received
- and all sorts of other non-specific effects.
So, considering these factors, what does this interesting re-analysis really tell us? My interpretation is as follows: the type of observational study that homeopaths are so fond of yields false-positive results. If we correct them – as the authors have done here for just one single factor, the RTM – the effect size gets significantly smaller. If we were able to correct them for some of the other factors mentioned above, the effect size would shrink more and more. And if we were able to correct them for all confounders, their results would almost certainly concur with those of rigorously controlled trials which demonstrate that homeopathic remedies are pure placebos.
I am quite sure that this interpretation is unpopular with homeopaths, but I am equally certain that it is correct.