Question:
Many of the claims suggest that the effect is immediate and obvious. For SOME people, some of the time.If it really worked, it would be for most people all of the time. Some people some of the time sounds consistent with placebo. I spelled out the cites right here, and as far as I could tell there were. Could you tell me precisely what I missed?That particular successful studies weren't replicated by independent researchers. Hey, I thought that Reilly et al. would qualify but I got set straight here by aPRO homeopathy poster.

Answer: I was the one who pointed out that your claim that the reproducibility had beentested by independent researchers was not valid. I said all three Reilly studieswere done by the same investigators, but it looks like the studies were actuallyconducted by three different investigators. I had missassumed the reason they werenot accepted as replications by the Linde meta analysis is that Reilly had doneall three trials.

So now I'm stumped. When I said there were no attempted replications I was usingthe definition of reproducibility in the Linde et al meta analysis: "No series ofstudies completely met our predefined criteria for reproducibility, which requiredat least three independent replications (different investigators) on the sameclinical condition, with the same model of homeopathy, remedy, outcomemeasurement, and a similar population.. For some reason, the Reilly study doesn'tmeet those conditions, but I don't know why.

Besides, the Reilly trials all had positive results. My point was and still is,that you have no examples to support the idea that anyone has tried and failed toreproduce any previous trials by any other investigators.

Now if I understand you correctly Syd, you are saying that in this particularinstance the criteria requiring the same interventionshould not require the exact same remedy, because the idea behind standardhomeopathic theory is that remedies should bedifferent for different people. That makes alot of sense, but it means adding awhole other layer of argument to the debate overreplication, which is why I guess the authors of the meta analysis went with thestandard definition of replication which requires exactly the same remedy. Eitherthat or for some inexplicable reason they didn't mention. they felt that "nearreplication" was simply not compelling enough to mean much. Personally , I don'tknow for sure why they didn't accept different remedies as the same intervention.I can't even venture a guess until I learn more about clinical trials in general.or at least hear Linde et al's reasons for their position.

This is a very complicated issue for anyone who really tries to understand allthe variables involved. How many, or what percentage of trials in general areallowed to ignore what criteria for reproducibility? Offhand it seems like a realcomparison would also depend at least partially on how effective homeopathy wasin those trials, which brings up the issue of rate ratios and percentage improvedand how high those figures are in other modalities before they are considereddramatic. If I understand him correctly , HD suggests that a trial of homeopathywould have to be nine times as effective before its considered dramatic. I'm notsure but I'd venture that an intervention which is three times as effective asplacebo is usually considered dramatic. I often see trials reported in the paperswhere an intervention which is twice as effective as placebo is considered abreak through.

I don't know if homeopathy works or not. I'm just trying to understand what theexisting evidence means. What I am looking for is information from someone who isfamiliar with normal, widely accepted medical standards and how clinical trials ofhomeopathy compare to the trials of other modalities. They should be able toquantify definitive claims where possible, or short of that give us some idea ofhow they arrived at their opinions. Something besides empty slogans which are alsobesides the point. Is that asking for too much ?