For a start, the fact that homeopathic remedies are generally so diluted that they contain no active ingredient is glossed over with the phrase "Critics argue that the active substance is so diluted that homeopathic remedies have no more effect than placebo or dummy treatment".
The article reports on a study by Dr. David Reilly at, God help us, the Glasgow Homoeopathic Hospital. What do we know about the study? Virtually nothing. According to the article, the study looked at 50 patients suffering from nasal allergies. Some were given homeopathy and some a placebo. I tried to find the study, so I could understand what they did. A search on the pubmed database doesn't find it. There is no mention of the status of the study in the Mail article. Has it been submitted to a journal? Presented at a conference? Written up on the back of a fag packet? We don't know. So there is no way the claims made in the article can be evaluated in any sensible way. Was the study blinded? Did it have adequate randomisation? What was the placebo used? We don't know. The only thing I can say about it is that the study size is small, and so more likely to produce a false positive result.
So, the reporting of the particular study that the article discusses is lamentable. But perhaps even worse is the total failure to place the study in the context of the rest of the scientific evidence. Dr Reilly is allowed to get away with saying that the study replicates some of his previous studies, and that "there were positive findings in 70 per cent of a further 180 clinical trials" of homeopathy.
Perhaps 70% of trials of homeopathy did apparently show a positive result. But this could be a result of poor study designs or insufficient sample sizes. Luckily, this can be tested using a technique called a meta-analysis. Essentially, you get the data from all the published trials, pool it together, and apply statistical techniques to the pooled data to evaluate bias. Such an analysis was conducted by Aijin Shang and co-workers, and published in the Lancet in 2005. This is a study much misunderstood by homeopaths, but what it shows is that large and well-designed trials show no effect for homeopathy, and that positive results of some trials can be explained by poor experimental design, bias and small sample sizes. A previous meta-analysis by Linde et al., published in the Journal of Clinical Epidemiology, came to similar conclusions. Dr. Reillys' previous work on homeopathic treatment for allergies was included in the Shang study, and was not deemed to be of high quality.
So, the current state of the scientific evidence is that there is no plausible mechanism for homeopathy to work, and that the results of large, well-conducted studies are consistent with homeopathic remedies being no more than placeboes. Going by the Mail article, you would get the impression that homeopathy was dismissed by blinkered scientists, but that evidence was gradually stacking up in favour of it. That is simply not the case.
But at least the Mail didn't claim that homeopathy can cure cancer...
EDIT: It has come to my attention that the Mail story is actually reporting the results of a trial that was published in the BMJ...in 2000. So why it has appeared 8 years later is anyone's guess. This trial was also included in the Shang meta-analysis, and not deemed to be of high quality. This is perhaps because the trial was not double-blind, even though it was described as such in the BMJ [Edit: my mistake. The wash-in period of the trial was single blind. The intervention was double-blind]. The authors also calculated that they needed to recruit 60 subjects in each group (a total of 120) for the study to have the required statistical power. There was a total of 50 subjects. So the only sensible conclusion would be that there was insufficient evidence either way.
Edit 2: Here's the author of the piece, Jenny Hope, giving evidence to the House of Commons Select Committee on Health regarding the influence of drug companies on research and research reporting. Note this bit:
When a story is in the public domain—or being placed with the specific intention of generating publicity—someone will write about it. If it's me than I know where the information is coming from, the background, and carried out the interviews—there is some quality control.
The quality control on this particular story seems to be negligible. Perhaps Ms. Hope was having a bad day?
