 |
 | Home | Contact Us | FAQ | Search & Site Map | Link to Us |
 | Sign In | Join | Other 45 Sites in Network |
Medical Forum / General / General / September 2005Homeopathy is of Value in the Treatment of Menopause
Even with the small sample size results were distinctive(p=.05). The results from the homeopathic drugs were mild but positive. "homeopathy may be of value in the treatment of menopausal symptoms" Homeopathy for Menopausal Symptoms in Breast Cancer Survivors: A Preliminary Randomized Controlled Trial Journal of Alternative and Complementary Medicine Feb 2005, Vol. 11, No. 1: 21-27 http://debunkbigpharma.blognation.us/blog/_archives/2005/8/27/1174694.html > Even with the small sample size results were distinctive(p=.05). The > results from the homeopathic drugs were mild but positive. [quoted text clipped - 7 lines] > > http://debunkbigpharma.blognation.us/blog/_archives/2005/8/27/1174694.html First trick: distinguish the homeopathic remedy from the solvent Second trick: Firgure out how companies like Boiron get a way with selling plain sugar pills. Since the P=.05 that means there is a 5% or less chance that the study is wrong. Prove the study wrong or accept the findings. If you are going to accuse J Alt CAM of promoting false results you better have some evidence. > Since the P=.05 that means there is a 5% or less chance that the study > is wrong. Prove the study wrong or accept the findings. If you are > going to accuse J Alt CAM of promoting false results you better have > some evidence. Why don't you apply the same standards you use to evaluate studies that conclude that pharmaceuticals are effective?: moo Guess what dork, using a P of .05 or lower IS WHAT THE PHARMAS use to get their drugs approved by the FDA. This is reality, go to school and learn how clinical studies are performed. If the P=.05 or less = significant finding and approved of by the peer reviewed scientific community. > Guess what dork, using a P of .05 or lower IS WHAT THE PHARMAS use to > get their drugs approved by the FDA. This is reality, go to school and > learn how clinical studies are performed. > > If the P=.05 or less = significant finding and approved of by the peer > reviewed scientific community. "RESULTS: There was no significant difference found in the primary outcome measure, the hot flash severity score..." If you'd read the abstracts instead of making up garbage (that it was of value in the treatment of menopause) and calling others 'dork', you'd contradict the results you reference less often. To quote you, "Prove the study wrong or accept the findings. If you are going to accuse J Alt CAM of promoting false results you better have some evidence." Of course this a 'comment' which they published, not subjected to serious review. PMID: 15750360 So you agree the Pharmas use a P=.05. What's you problem? Are you are saying the Journal is not to be trusted? If you are, you offer no proof. Prove it, or shut up. >So you agree the Pharmas use a P=.05. Yes, sort if -- with caveats. For simplicity these are general comments about testing, without getting into any details of this specific report. No drug would be approved based on one single test with p = 0.05 (or one single test, period). Multiple testing is done, better and better tests. By better, I mean that the tests are more focused (and they are also larger). It is the norm that many/most drugs drop out upon follow-up testing. A p value is a statistical calculation. It passes no judgment on the validity of the test itself. Was the placebo a proper placebo (not at all a trivial concern in many tests)? Was the study really properly blinded? Were the data properly objective? Etc. When there are contradictory results for something, or something "very odd" is reported, much time is spent analyzing these details. (And this point does not necessarily mean the authors did anything wrong; it may just mean that things are more complex than what they assumed.) Another way to say this is to realize that when one applies a statistical test, one makes a series of assumptions. The conclusion from the resulting p value is no more valid than the assumptions made. When many things are tested, or many outcomes are measured, it is likely that some spurious positives will show up. In fact, if one measures 20 different outcomes of a treatment that has absolutely no effect (good or bad in any way), it would be surprising if we did not see one that gives a p of 0.05 (or better). Apparently, that is an issue in this case. Did they do the statistics to take into account how many things they measured? Many don't bother (it really isnt conclusive one way or the other anyway). You may recall a controversy a couple years ago when a company reported that a drug (an AIDS vaccine?? don't remember) seemed to have some benefit for certain ethnic groups. The conclusion seemed "unreasonable", and there followed considerable discussion about the proper way to analyze the data. There is a maxim in science that extraordinary claims require extraordinary proof. Sometimes extraordinary claims turn out to the right, but a great weight of evidence is required to convince people. Statistical tests do not in isolation prove or disprove things. They merely provide evidence, with confidence levels. So it is one thing to suggest that the paper referred to has something in it that _might_ be worth following up with a larger, more focused test. But it is quite another to suggest that they have actually shown anything worth quoting, at this point, as true. bob dgillila@cox.net wrote: > If the P=.05 or less = significant finding and approved of by the peer > reviewed scientific community. Surely the issue here is the small sample size which renders the results dubious? I'd also question your thread title: the authors noted that there were no effects on the primary outcome measures of hot flush frequency and severity; the benefit they saw was in 'general health score one year later'. I wonder if this just demonstrates that if you keep measuring different things to a 5% significance level, one in twenty of them will give the result you want? Even the authors only describe it as a 'preliminary' study. I'd be interested in seeing the results of the full one. Actually, I'd be interested in seeing the full info on this study, given that all that appears to be available is a comment in a journal. Tom Surely the issue here is the small sample size which renders the results dubious? dubious? "homeopathy may be of value in the treatment of menopausal symptoms" is the claim that I'm and the study is sticking to. You are not going to get an absolute from this study. Bash it all you want, whatever. dgillila@cox.net wrote: > dubious? "homeopathy may be of value in the treatment of menopausal > symptoms" is the claim that I'm and the study is sticking to. Oh, you're trolling. How stupid of me to waste my time replying to you. Ah well, won't make that mistake again. *plonk* Tom > dgillila@cox.net wrote: > > If the P=.05 or less = significant finding and approved of by the peer > > reviewed scientific community. > > Surely the issue here is the small sample size which renders the results > dubious? The size was what it was. Its not dubious because of that, just weaker that if it had been larger. It says "There was no significant difference found in the primary outcome measure, the hot flash severity score". Its funny that they didn't give a separate P for that result. You wouldn't expect both results to be the same although thats possible. A separate P (.1) was stated for one of the treatments for the first three months before those patients' data returned to match the rest. Had it been a three month trial (ended it there) I suppose they would have claimed a victory, but they got greedy. > I'd also question your thread title: the authors noted that there were > no effects on the primary outcome measures of hot flush frequency and > severity; the benefit they saw was in 'general health score one year > later'. For the treatment of menopause in breast cancer survivors, there was no improvement over placebo. This was the title and purpose of the study. They didn't title it 'General Health Benefit ...". > I wonder if this just demonstrates that if you keep measuring > different things to a 5% significance level, one in twenty of them will [quoted text clipped - 6 lines] > > Tom Authors hope for money to do more of the same and try to choose their words optimisticly, short of lying. Maybe they will do more but I don't know who would pay: there's no sign that there're onto a salable result such as reducing hot flashes. You don't need to be 'seen by a homeopathic provider' to check on compliance, give out the remedies/placebos and record data. When I read that "Patients were seen by homeopathic providers every 2 months for 1 year." I immediately assumed that this was not blinded and that these were coaching sessions to bias the health score number. Why else? was he focusing the energy or giving the stuff fresh shakes? Breast cancer survivors may have been through a lot, a little counseling/handholding and positive affirmation can impact their view of their "general health". SF-36 has a significant psych. component: http://www.sf-36.org/tools/sf36.shtml#MODEL |
Reply to this Message |
 Sign In
 Join
 My Latest Posts My Monitored Threads My Blog My Photo Gallery My Profile My Homepage Start New Thread Enable EMail Alerts Rate this Thread
|
©2008 Advenet LLC Privacy Policy - Terms of Use
This website includes both content owned or controlled by Advenet as well as content owned or controlled by third parties.