How to read Kaplan-Meier plots
ฝัง
- เผยแพร่เมื่อ 22 ต.ค. 2023
- Vinay Prasad, MD MPH; Physician & Professor
Hematologist/ Oncologist
Professor of Epidemiology, Biostatistics and Medicine
Author of 450+ Peer Reviewed papers, 2 Books, 2 Podcasts, 100+ op-eds.
If you want to contact me, do it here: www.vinayakkprasad.com/contact
Google Scholar: scholar.google.com/citations?...
Substack: vinayprasadmdmph.substack.com/
Podcast: podcasts.apple.com/us/podcast...
Personal Website: www.vinayakkprasad.com
Laboratory Website: www.vkprasadlab.com
Podcast Website: www.plenarysessionpodcast.com
Academic Publications: www.vinayakkprasad.com/papers
Follow me on:
Twitter @vprasadmdmph
Through out this lecture I could hear, echoing in my mind, my basic statistics professor talking about how easily numbers can be manipulated to produce whatever story or result is desired. The most important thing I learned in that class was how to question and evaluate the results of any given study.
Like who wins the War gets to write the history..
This is the best video on these issues I have ever seen. All summed up with examples.
I am a fellow at a clinical trials unit.
Fascinating. I can understand and I have no medical training...just an interest. Thank you for explaining in words even people like me can understand
Medical training and biostatistical training are two different things.
@@kodowdus Oh, okay. thank you. What is the difference? I love to learn.
An MS in Biostatistics or MPH with a concentration in Biostatistics typically requires at least half a dozen courses in quantitative methods related to medical research, but the core of any MPH program would still provide more biostatistical training than a standard medical degree.
@kodowdus This was so eye opening to me, a general consumer. Not just the lecture by Dr. Prasad, but knowing how much dedication and effort are required to do what you do are so appreciated. My son is a second year resident in Neurology and I am amazed by all you folks do. Thank you for making the world a little better
A very important lectura imparted by a world-class expert!
And this, video techs, is why we synch the source to the camera and ideally to the lights as well. Your camera and projector are not getting along.
It’s amazing how the FDA didn’t pick this out. So much for the high quality of the regulators.
The FDA is just a rubber stamp agency
Thanks for this. These are the types of lectures that I’m interested in consuming. I want to learn study types/design and interpretation. The epistemology of medicine and pharmacology should be front and center. Thanks for all you do!
Thanks so much for this Vinay! Would love to see more of these to help read studies critically. Am working my way through all of your books too.
I learned more from this than any professor or tv pundit. Thank you Vinay!
Excellent, lots of bases covered. I will refer people to this if they question why I listen to Vinny.
This lecture packs ton of information on epidemiology and biostatistics!
i don't get this obsession with using aperture or camera zooms to make the person in the foreground sharp but the backgrounds blurry. it works well for your podcast but for a lecture, can't you just use conventional video specs? the slides are blurry. Look at how often the focus glitches and even the slides have weird flashing color bars
Amazing lecture. KM curve interpretation made so easy. Thanks Dr Vinay
thank you, dr. prasad.
Vinay, en tant que lambda c’est unique de pouvoir avoir accès à cela, merci de me donner les outils pour comprendre et être moins ignorante!
Bon pour Topol j’avais déjà remarqué toute seule les conneries qu’il raconte😅
You're an inspiration Dr. Vinay. I aspire to be a Hem-oncologist someday and do what you do.
Excellent lecture, thank you
Thanks Doc.
Great!
Thank you. It would be great if you could talk about the problems in the recent Harvard School of Public Health study claiming that red meat increases your risk of type 2 diabetes.
We need more of these tutorials please. So the drug companies can't perform statistical gymnastics on the public.
If over-simplified statistical analyses of medical research (for example, the presentation of Kaplan-Meier curves for purposes of analyzing observational data without any attempt at the use of proportional hazards models to control for potential confounding) is actually making its way into peer-reviewed publications, it behooves us to ask whether this is a phenomenon across the board or are some journals (e.g., the American Journal of Epidemiology) more dependable sources of such publications than others.
It is possible to have and 'immediate' effect, if the effect is related to 'fast acting' innate immune reactions, and not restricted to 'delayed' adaptive immune reactions. Excellent information overall and very useful!
As a thought experiment: if the immune response is so fast, why would you even need a vaccine? In other words, if it could respond so quickly, it could respond equally well to the real infection or the vaccine. The advantage of a vaccine is preparing your immune system in advance
*so helpful so good!* gr8 stuff. is it really harder to have the robot adjust every time someone is removed, rather than missing number tricks as if we still do these calculations by hand? _(it may not be that easy, exactly. i guess, it _*_may_*_ require some new math. id have to think about it w someone who only does math. but either way: surely this can be solved & coded in R & STATA etal, published in a math or technology journal: then we'd not have to deal w things like that misleading magnitude change)_
ill add, often it is actually as easy as one first thinks. like the reporting of censoring. _JC
I wonder if you do km curves readjusted for censoring, what do they show?
Vijay- how do you randomize in gene therapy trials of ultra rare diseases
*22:40** yes, this type of thing.* i guess, also, it is hard to enroll people & trials are expensive. but ive never liked the ways we deal w missing data. i get the argument, but it isnt true fx size to just take mean assumptions or similar of any kind. if it doesnt break the underlying math, why do it? unless it does. but thr has to be another way. _JC
Bro, this is a fabulous talk, but, for the Love of God, did your seven year old cousin film this?
Lol - Probably because they didn’t plan on videoing the lecture and / or last minute someone volunteered and didn’t adjust settings (frame rate etc.). Camera and TV are out of sync as the camera recording fewer images per second, thus the flicker. Try it, it’s not easy to get good quality video recording with TV background for the presentation. Easier/ better if screen wasn’t TV.
Double lol - screen casting to my large TV now. But great presentation still!
1rd
Stock exchange has their fingers in this! 🙃🤩👹🎭
Do you know Dr John Campbell?
HIGHLY RECOMMEND YOU SPEAK WITH HIM
YOUR ON THE SAME TONE ABOUT ALOT OF ISSUES
No. John doesn't understand science and statistics. He just reads it.
YOU NEED TO TALK TO DR JOHN CAMPBELL AND UNITE TRUST ME