werner@aecom.UUCP (Craig Werner) (05/15/85)
Some years ago in the United States it was argued that since Rheumatoid arthritis was caused by the body's immune system attacking itself that suppressing the immune system would relieve some of the symptoms of the disease. This was tried, and sure enough, about 30% of the people recovered from Rheumatoid arthritis during the course of the study. Now in England, they didn't like the way the study was done, so they got 60 people with Rheumatoid arthritis and immunosuppressed half of them. The other half got a placebo - a fake pill. As in the US, 30% treated got better. Unfortunately for the study, 30% of the placebo group got better as well. (This may help to explain why copper bracelets help about 1/3 of all arthritics -- about 1/3 will be helped by anything.) Therefore, immunosuppression was declared a worthless treatment -- why treat the patient and increase the risk of infection if just telling the patient that you're treating them will do just as well. Several years later, the evidence began to accumulate that treatment was some immunosuppressive drugs increased the risk of cancer. The physicians in England said, "Oh no, we exposed these patients to needless risk, etc." and then they went back to follow the cohort they had studied for 5 years. Much to their dismay, 5 of the 60 people had developed cancer within the 10 years post-study. This is much higher than would be expected. However, much to their suprise and relief, all of the 5 cancers were in the placebo group and none were in the group that actually got the drug that was suspect. In other words, all 5 cancers were caused by random chance and had nothing to do with the study. Had chance fallen the other way, the drug would most probably have been banned and lawsuits continuing to this day. Does placebo cause cancer? No, of course not. It cannot. If starch and gelatin caused cancer we'd all be shapeless masses of lymphoma by now. Thus ends part three of my didactic diatribe in partial defense of Western Allopathic medicine. Comments always welcome. -- Craig Werner !philabs!aecom!werner What do you expect? Watermelons are out of season!
seifert@hammer.UUCP (Snoopy) (05/20/85)
> Much to their dismay, 5 of the 60 people had developed cancer within >the 10 years post-study. This is much higher than would be expected. > > Does placebo cause cancer? No, of course not. It cannot. If starch >and gelatin caused cancer we'd all be shapeless masses of lymphoma by now. > > Craig Werner > !philabs!aecom!werner > What do you expect? Watermelons are out of season! Just EXACTLY what is in those "placebos", anyway? Are they 100.00000% pure? Are all placebos used in all experiments always the same? If 5-in-60 is much higher than would be expected, than what *is* the explaination? Is the small sample size sufficient to account for this? Perhaps we could take up a collection and ship poor Craig a watermelon before he posts any more of this nonsense. The Earth revolve around the Sun? No, of course not. It cannot. Snoopy tektronix!mako!seifert What do you expect? The Scientific Method is out of season!
ems@amdahl.UUCP (ems) (05/22/85)
> Does placebo cause cancer? No, of course not. It cannot. If starch > and gelatin caused cancer we'd all be shapeless masses of lymphoma by now. > > Thus ends part three of my didactic diatribe in partial defense of > Western Allopathic medicine. Comments always welcome. > Craig Werner Perhaps not so obviously an 'on course not.'. That sounds so final, so sure. Was there no coloring in the placebo capsule? Was there nothing in the packaging and delivery materials? Was this truely chance? I don't know. I would wager that you also cannot *really* know. We can only make a hypothesis and test it. It would be interesting to repeat the test with several different placebos... could explain a large number of anomalies in testing if the color in the gelatin of our 'placebos' was not inert... -- E. Michael Smith ...!{hplabs,ihnp4,amd,nsc}!amdahl!ems Tilapia Zilli is the way and the light. This is the obligatory disclaimer of everything.
ems@amdahl.UUCP (ems) (05/22/85)
> > Craig Werner > > !philabs!aecom!werner > > What do you expect? Watermelons are out of season! > > Perhaps we could take up a collection and ship poor Craig a watermelon > before he posts any more of this nonsense. > > The Earth revolve around the Sun? No, of course not. It cannot. > > Snoopy > tektronix!mako!seifert > What do you expect? The Scientific Method is out of season! Hmmm, this raises an interesting point. Where is Craig? Here in sunny California (The Land of Fruits and Nuts) watermelons *ARE* in season!!! Does this mean that a change of .signature is in the wind? -- E. Michael Smith ...!{hplabs,ihnp4,amd,nsc}!amdahl!ems Tilapia Zilli is the way and the light. This is the obligatory disclaimer of everything.
hollombe@ttidcc.UUCP (The Polymath) (05/22/85)
In article <1273@hammer.UUCP> seifert@hammer.UUCP (Snoopy) writes: > >Just EXACTLY what is in those "placebos", anyway? Are they 100.00000% pure? >Are all placebos used in all experiments always the same? No. They have to simulate different conditions, after all. Some are injected, some are taken orally, some are applied topically, etc. > >If 5-in-60 is much higher than would be expected, than what *is* >the explaination? Is the small sample size sufficient to account for this? ~sigh~ I wish they'd make courses in introductory statistical analysis requirements for graduation. Then maybe people would have some idea of how experimentation and scientific method works. The explanation for the higher than normal cancer rate among the controls in quoted experiment is probably normal statistical variance. It might be a good idea to check this out by replicating the experiment (that's what Master's theses are for (-: ). If the same thing happened again, _then_ it would be a good idea to start checking into the cancer causing properties of the placebos. An even easier method would be to research the literature and check out the cancer rate among subjects given the same placebos in other experiments. -- -_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_- The Polymath (aka: Jerry Hollombe) Citicorp TTI 3100 Ocean Park Blvd. Santa Monica, CA 90405 (213) 450-9111, ext. 2483 {philabs,randvax,trwrb,vortex}!ttidca!ttidcc!hollombe
wmartin@brl-tgr.ARPA (Will Martin ) (05/23/85)
A general query about controlled experiments, inspired by the subject discussion: If you are testing a truly effective medication, say for some serious disease or condition, and are conducting a blind or double-blind test where some participants get placebos and some the medication, and a number of the control group die because they got only placebos, while all the test subjects survive because this experimental medication was really effective, have you not violated the Hippocratic oath (and also laid yourself open for damages, no matter what you had the participants sign), by denying this (admittedly experimental) medication to people who could possibly have been saved by it? I recognize the necessity for testing before drugs or treatments are made available; there are both scientific and legal requirements for such testing. Yet I always wonder about those in these "control" groups; they aren't subject to the risks of the tested substance, but cannot get any possible benefits, either. I would think these considerations would make controlled testing for truly-life-saving treatments practically impossible to carry out. Am I right in this supposition? Will Martin USENET: seismo!brl-bmd!wmartin or ARPA/MILNET: wmartin@almsa-1.ARPA
hollombe@ttidcc.UUCP (The Polymath) (05/24/85)
In article <10970@brl-tgr.ARPA> wmartin@brl-bmd.UUCP writes: >If you are testing a truly effective medication, say for some serious >disease or condition, and are conducting a blind or double-blind test >where some participants get placebos and some the medication, and a >number of the control group die because they got only placebos, while >all the test subjects survive because this experimental medication was >really effective, have you not violated the Hippocratic oath (and also >laid yourself open for damages, no matter what you had the participants >sign), by denying this (admittedly experimental) medication to people >who could possibly have been saved by it? Usually if there is any effective treatment for a condition at all, the new treatment will be compared to the effects of the old treatment rather than to a placebo control. This is especially true with life-threatening diseases. It is still possible, of course, that some patients on the old treatment will die for lack of the new treatment. The rationale here is that there is no way to know if the new treatment is better, or even effective at all, until the experiment is performed. Preliminary studies with animals may well lead the experimenters to expect so, but the final proof must be with human subjects. This whole area of experimental ethics is very complex and there are abuses. Questions arise with regard to coercion and what is a volunteer. This is why prison inmates are seldom used as subjects even on a "volunteer" basis. They may be doing it to get brownie points on their records and this is interpreted by some to be coercion (if they don't volunteer, they may stay in prison longer). It's also why most Universities no longer allow graduate psychology students to experiment with freshman Intro. Psych. students (my alma mater still does, but the experiments have to meet very rigid standards and be approved by a faculty ethics committee). Finally, for an example of a case where placebos are totally inappropriate, I cite the infamous L.A.P.D. "Goat Study": Seems the L.A. Police wanted to test out some new-type bullet proof vests to see if they were really more effective. So, they got some goats, wrapped them in the new vests, and opened fire at them. Of course they had to have a control group, and they did -- goats without bullet proof vests. Not surprisingly, the results of the experiment were statistically significant (very). Obviously, the controls should have been wearing the old-type vests for the experiment to be meaningful. Fortunately, the P.D. didn't have the budget to do testing with human subjects. -- -_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_- The Polymath (aka: Jerry Hollombe) Citicorp TTI 3100 Ocean Park Blvd. Santa Monica, CA 90405 (213) 450-9111, ext. 2483 {philabs,randvax,trwrb,vortex}!ttidca!ttidcc!hollombe
seifert@hammer.UUCP (Snoopy) (05/25/85)
In article <10970@brl-tgr.ARPA> wmartin@brl-bmd.UUCP writes: >A general query about controlled experiments, inspired by the subject >discussion: > >If you are testing a truly effective medication, say for some serious >disease or condition, and are conducting a blind or double-blind test >where some participants get placebos and some the medication, and a >number of the control group die because they got only placebos, while >all the test subjects survive because this experimental medication was >really effective, have you not violated the Hippocratic oath (and also >laid yourself open for damages, no matter what you had the participants >sign), by denying this (admittedly experimental) medication to people >who could possibly have been saved by it? ... >I would think these considerations would make controlled testing >for truly-life-saving treatments practically impossible to carry out. >Am I right in this supposition? There have been cases where the results were quite dramatic, and the experiment cut short and the control group given the experimental treatment. Experiments don't have to give binary lived/died results to be useful. Snoopy tektronix!mako!seifert