tag:blogger.com,1999:blog-4519234397783312626.post3719630931813611224..comments2023-10-09T11:42:57.305-04:00Comments on Healthcare, etc.: Early radical prostatectomy trial: Does it mean what you think it means?Marya Zilberberghttp://www.blogger.com/profile/16080475886113209344noreply@blogger.comBlogger5125tag:blogger.com,1999:blog-4519234397783312626.post-49937821750140370862012-07-26T09:59:55.537-04:002012-07-26T09:59:55.537-04:00Shannon, you are absolutely right on both points! ...Shannon, you are absolutely right on both points! I have often thought that much of our medical treatment research is about detecting minute differences at the margins. And though these differences get magnified at the population level, it is difficult to know the real impact since treatments behave differently in the wilds of clinical practice than in RCTs. In my post yesterday (http://evimedgroup.blogspot.com/2012/07/medicine-as-trolley-problem.html) I referred to the Mandelblatt paper on mammography, where it looks like we have to screen 1,000 40-yo women for 30 years to avoid 8 deaths from breast CA. At the same time there are false alarms and lots of biopsies, and we don't even know the extent of complications from these biopsies.Marya Zilberberghttps://www.blogger.com/profile/16080475886113209344noreply@blogger.comtag:blogger.com,1999:blog-4519234397783312626.post-40319440994574445132012-07-26T07:12:08.763-04:002012-07-26T07:12:08.763-04:00There's another important take home message he...There's another important take home message here, which is rarely expressed this way, but may matter more to patients than the minor differences that matter to researchers. If prostatectomy offered a large, slam dunk mortality benefit for men with early stage prostate cancer, we would see the effect fairly easily. Sadly, it is not. If it does offer a mortality benefit (which is looking less and less likely) that benefit is a small one at best, and it must be weighed by the individual patient against the very significant side effects of the treatment. The very fact that we have to spend so much time and money and effort to find out what the benefit might be is a signal that it's not very big. <br /><br />There are people who will argue that even if the mortality benefit to the individual is small, the population based benefit could be large because so many men are diagnosed with prostate cancer. Leaving aside for moment the fact that prevalence is being driven in large measure by a lousy screening test, we should question this "population multiplier" idea. Yes, there may be many lives "saved" by a procedure that has a very small individual benefit, but populations don't die, and populations don't suffer treatment side effects. Individuals do. What matters to the patient is what is most likely to happen to him.Shannon Brownleehttps://www.blogger.com/profile/02830276421398983266noreply@blogger.comtag:blogger.com,1999:blog-4519234397783312626.post-73141850306479955202012-07-21T10:25:19.070-04:002012-07-21T10:25:19.070-04:00Kenny, yes, you are absolutely right about that --...Kenny, yes, you are absolutely right about that -- practice ahead of evidence is a recipe for never getting the answer. <br /><br />Dr. Skeptic, very interesting observation, thank you. In some ways the doctor or the medical center the patients use acts as an instrumental variable and essentially randomizes them to one treatment or another. This is an idea that I had the good fortune to discuss once with Jack Wennberg, and would be fascinating to test.<br />As for as treated analysis, that would be prone to insurmountable biases, I am afraid, and would be even more confusing than enlightening.<br /><br />Thanks to both of you for your comments.Marya Zilberberghttps://www.blogger.com/profile/16080475886113209344noreply@blogger.comtag:blogger.com,1999:blog-4519234397783312626.post-42551942067287663862012-07-21T04:21:21.180-04:002012-07-21T04:21:21.180-04:00Excellent post.
Re: Low participation rate.
Our gr...Excellent post.<br />Re: Low participation rate.<br />Our group is currently researching reasons why people decline participation in randomised trials of surgery versus non-operative treatment. While it depends on the specific trial (the disease, the intervention, the control etc) it appears that one of the main reasons is that people do not want their treatment to be decided by chance. This is despite them understanding that the two treatments are similar, and despite being happy to have either treatment.<br />I think that they don't realise that the treatments they normally receive depend on the (random) nature of which particular doctor they see.<br />Re: Crossover.<br />Even in real life, not all people offered a prostatectomy will go ahead, and patients who have been advised against an operation may shop around to get one. It might be informative to know why they crossed over. Do you think an "as treated" analysis would have helped or hindered?Dr Skeptichttps://www.blogger.com/profile/09376469049519802493noreply@blogger.comtag:blogger.com,1999:blog-4519234397783312626.post-32604738891490954522012-07-20T12:11:31.010-04:002012-07-20T12:11:31.010-04:00One thing that the trial points out, as did the mu...One thing that the trial points out, as did the much-criticized PLCO and ERSPC trials of PrCa screening, is that it's really difficult to recruit patients for and maintain randomized groups to test an intervention that is already in widespread practice. No doubt most of the men who decline participation in this study said: "So you want me to be in your study when there's a 50/50 chance I'll be given NO treatment for a potentially lethal cancer? The heck with that!" And unfortunately, by making screening and treatment so widespread before there was evidence, we physicians reinforced the notion (now very much refuted) that most if not all prostate cancers would eventually be lethal.kennylinhttps://www.blogger.com/profile/00240060576692353940noreply@blogger.com