 |
 |
arnie, sham surgery is not the only option. consider amputations and gangrene; do we need to do sham surgery to test whether amputations actually prevent the spread of gangrene? if the concern is really that investigators have bias, there are much more ethical approaches than sham surgery. for instance, simply don't tell the clinician evaluating patient condition which patients have received surgery and which haven't. if the concern is about a placebo effect on patients, comparison to historical placebo effects in PD patients seems like a fairly good first cut. statistically speaking, other approaches may not be as pure, but the whole point of medicine is to treat people, not perform elegant experiments. if the trials for something like DBS, just make that information available to the patients and physicians considering the treatment with language along the lines of "40% of PD patients showed improvement with DBS surgery, but there was no control group for ethical reasons. historically, 15% of patients with similar circumstances have shown improvement without any treatment." On 9/14/07, Arnie Kuzmack <[log in to unmask]> wrote: > > There were several comments made by respondents (most of whom are > apparently on > this list), that I would like to respond to. > > Comment: The statistical proof is not available in another way. I am no > expert > but can't we demonstrate that something is helpful just by a larger sample > size > or is a control group always required? Can't we pair people as close as > possible and not have sham? > > Response: A basic principle in statistics is that a larger sample size > does not > correct for bias. For example, in the classic case, if the investigators > are > emotionally committed to showing that the treatment works, and if they > know > which patients got the new treatment, they will have an unconscious > tendency to > evaluate patients in the treated group differently from those in the > controls. > This is particularly true with diseases like PD, where the evaluation is > not > strictly objective. It is also a problem using "historical controls" or > pairing > patients as suggested by the commenter, since the evaluations of the > treated > group would be subject to evaluation bias. > > This is separate from the "placebo effect", where patients getting a > placebo > actually do better than they would otherwise. > > Comment: There are alternatives to double blind trials, they are actually > > becoming quite common for diseases like cancer where it would be unethical > to > give a patient a placebo and let the patient die to prove that a new > medication > is effective. The studies aren't as simple, but these could and should be > > applied to Parkinson's surgeries as well. > > > Response: Standard practice for trials involving serious diseases is to > give > the control group whatever the current standard treatment is for that > disease. > The object is not to prove that the new treatment is better than nothing > but > that it is better (or not) than the current standard practice. In the > simplest > case, where both treatments are pills, the treated group will get two sets > of > pills, one of which is the new drug and one a placebo, while the control > group > also gets two sets of pills, one of which is the placebo and and one the > standard treatment. The total treatment of each of the groups looks the > same as > the other one. That way, the patients in the control group get the same > treatment as they would have if they were not in the trial. > > > ----- Original Message ----- > From: " [log in to unmask]" <[log in to unmask]> > To: <[log in to unmask]> > Sent: Friday, September 07, 2007 7:11 PM > Subject: Opinions on the use of sham surgery in PD research > > > Over the last month, some listmembers participated in the Parkinson > Pipeline > Project's " Opinion Survey on Sham Surgery in Parkinson's Clinical > Research." > The results are now in. The results of the opinion poll and respondents' > comments are on the PPP website at: > http://www.pdpipeline.org/whatsnew/shamsur_survey.htm > There is also a link from the homepage www.pdpipeline.org > Thanks to everyone who participated. > Linda Herman > > > ---------------------------------------------------------------------- > To sign-off Parkinsn send a message to: mailto: > [log in to unmask] > In the body of the message put: signoff parkinsn > > ---------------------------------------------------------------------- > To sign-off Parkinsn send a message to: mailto: > [log in to unmask] > In the body of the message put: signoff parkinsn > ---------------------------------------------------------------------- To sign-off Parkinsn send a message to: mailto:[log in to unmask] In the body of the message put: signoff parkinsn