The Official Cricketweb Science Thread!

[Teja.]

Originally Posted by Spooony

Far out in the uncharted backwaters of the unfashionable end of the Western Spiral arm of the Galaxy lies a small unregarded yellow sun. Orbiting this at a distance of roughly ninety-two million miles is an utterly insignificant little blue green planet whose ape-descended life forms are so amazingly primitive that they still think digital watches are a pretty neat idea.

<

Behind every man now alive stand thirty ghosts, for that is the ratio by which the dead outnumber the living. Since the dawn of time, roughly a hundred billion human beings have walked the planet Earth.

Now this is an interesting number, for by a curious coincidence there are approximately a hundred billion stars in our local universe, the Milky Way. So for every man who has ever lived, in this Universe there shines a star.

But every one of those stars is a sun, often far more brilliant and glorious than the small, nearby star we call the Sun. And many--perhaps most--of those alien suns have planets circling them. So almost certainly there is enough land in the sky to give every member of the human species, back to the first ape-man, his own private, world-sized heaven--or hell.

[Spooony]

Originally Posted by Teja.

<

Behind every man now alive stand thirty ghosts, for that is the ratio by which the dead outnumber the living. Since the dawn of time, roughly a hundred billion human beings have walked the planet Earth.

Now this is an interesting number, for by a curious coincidence there are approximately a hundred billion stars in our local universe, the Milky Way. So for every man who has ever lived, in this Universe there shines a star.

But every one of those stars is a sun, often far more brilliant and glorious than the small, nearby star we call the Sun. And many--perhaps most--of those alien suns have planets circling them. So almost certainly there is enough land in the sky to give every member of the human species, back to the first ape-man, his own private, world-sized heaven--or hell.

Well, basically, there was this little dot, right? And the dot went bang and the bang expanded. Energy formed into matter, matter cooled, matter lived, the amoeba to fish, to fish to fowl, to fowl to frog, to frog to mammal, the mammal to monkey, to monkey to man, amo amas amat, quid pro quo, memento mori, ad infinitum, sprinkle on a little bit of grated cheese and leave under the grill till Doomsday

[silentstriker]

Laser spotlight reveals machine 'climbing' DNA

[silentstriker]

Black widow's tango mortale in gamma-ray light: Scientists discover record-breaking millisecond pulsar with new analysis method

[Spark]

Researchers report major advance in using sunlight to produce steam without boiling water

Thing I noticed:

The potential societal benefits are staggering. They include more energy-efficient distillation of alcohol...

[wellAlbidarned]

**** solar power, we want our whiskey

[Top_Cat]

Interesting thing popped up in my inbox.

Ignoring the Science on Mammograms - NYTimes.com

I had a read of the paper and its conclusions fly in the face of Cancer Australia's position on the matter and, tbh, the available evidence not to mention our work in the area. But, I tell you what, this is a journal with a pretty hefty impact factor (got a paper with them right now which I reallyreallyreally hope gets up) so that paper automatically gets a heck of a lot of credibility. The NYT article above will probably be bloody influential too and certainly this has popped up elsewhere.

If you have a read of the paper (N Engl J Med 2012;367:1998-2005. DOI: 10.1056/NEJMoa1206809), prima facie it seems reasonable but that's part of the problem; a reasonable-sounding journal paper often leads to much stronger news articles like the above which is what a lot of people see in social media feeds, etc. which will, you'd think, directly impact on the screening rate. I just get very worried at a time when, the vaccination rate has been dropping, about the influence of something like this. As Andrew Wakefield showed, something like this can really take hold if you don't stomp on it quickly.

Also shows you should critically review anything, even **** published in peer-reviewed journals with awesome impact factors. Just quietly, one of the authors has a, shall we say, somewhat favourable relationship with the NEJM.

[uvelocity]

[silentstriker]

Originally Posted by Top_Cat

Interesting thing popped up in my inbox.

Ignoring the Science on Mammograms - NYTimes.com

I had a read of the paper and its conclusions fly in the face of Cancer Australia's position on the matter and, tbh, the available evidence not to mention our work in the area. But, I tell you what, this is a journal with a pretty hefty impact factor (got a paper with them right now which I reallyreallyreally hope gets up) so that paper automatically gets a heck of a lot of credibility. The NYT article above will probably be bloody influential too and certainly this has popped up elsewhere.

If you have a read of the paper (N Engl J Med 2012;367:1998-2005. DOI: 10.1056/NEJMoa1206809), prima facie it seems reasonable but that's part of the problem; a reasonable-sounding journal paper often leads to much stronger news articles like the above which is what a lot of people see in social media feeds, etc. which will, you'd think, directly impact on the screening rate. I just get very worried at a time when, the vaccination rate has been dropping, about the influence of something like this. As Andrew Wakefield showed, something like this can really take hold if you don't stomp on it quickly.

Also shows you should critically review anything, even **** published in peer-reviewed journals with awesome impact factors. Just quietly, one of the authors has a, shall we say, somewhat favourable relationship with the NEJM.

I haven't seen compelling data from the other side though, especially in the value of screening younger women. I mean, technically, if you put every person in a full body scan MRI machine every three months from age 18 onwards, you'll definitely save some lives. But it's about the cost/benefit analysis.

I could be wrong about the data though - I haven't done research on it.

[Top_Cat]

It's not a settled issue, sure. But it's not as simple as the paper would suggest either. Just to pick on one of the assumptions, their estimate for 'overdiagnosis' didn't distinguish between invasive vs insitu tumours which obviously have differing risk profiles and this is apart from the hormones any given tumour expresses which changes the risk profile again, something they didn't look at.

Plus, it's also based on an assumption that can't ever be tested, that if you pick up early breast cancers of, say, younger women with an insitu tumour and remove it, that's an overdiagnosis because it probably wouldn't have gone anywhere. Especially in younger women, there's clinical evidence enough that's a questionable assumption. But because they're being picked up earlier and removed, well, we'll never know will we? But this guy has no problem calling those overdiagnoses. That's plenty debatable, in my book.

That said, there are differences in how this stuff is dealt with between countries so I can understand why you'd mention risk/benefit. The American Cancer Society recommends breast exams from 20 every 3 years and yearly mammograms from 40 (plus MRIs for certain patients and the efficacy of MRIs is somewhat controversial) whereas Cancer Australias only recommends mammograms every two years from 50. Huge difference in costs and, unlike here where the screens are free post-40, all borne by the patient. That and the CA website also specifically mentions what it doesn't recommend (regular screens for women < 40), unlike the US equivalent.

I certainly do understand the resistance to over-screening, picking up every minor defect and treating it regardless of risk isn't smart medicine and, of course, screen yourself often enough and you might just give yourself a tumour.

[silentstriker]

Originally Posted by Top_Cat

It's not a settled issue, sure. But it's not as simple as the paper would suggest either. Just to pick on one of the assumptions, their estimate for 'overdiagnosis' didn't distinguish between invasive vs insitu tumours which obviously have differing risk profiles and this is apart from the hormones any given tumour expresses which changes the risk profile again, something they didn't look at.

Plus, it's also based on an assumption that can't ever be tested, that if you pick up early breast cancers of, say, younger women with an insitu tumour and remove it, that's an overdiagnosis because it probably wouldn't have gone anywhere. Especially in younger women, there's clinical evidence enough that's a questionable assumption. But because they're being picked up earlier and removed, well, we'll never know will we? But this guy has no problem calling those overdiagnoses. That's plenty debatable, in my book.

Surely you could figure out the mortality rates of women who don't have the checkups vs. women who do, and see what the difference in mortality is vs. the amount of invasive procedures that you did. And if that ratio is worth it.

Not defending this paper, btw.

[Top_Cat]

Like I said, it's not that simple. So many confounds and competing risks, especially > 60, raw mortality doesn't give enough info about the real risk to an individual.

Not just that, certain invasive surgeries are performed for reasons other than survival these days. For example, following a lumpectomy for a <20mm screen-detected invasive lobular tumour, you might ink the tumour and decide that the margins of the excision still contain residual cancer. Now, without screen-detection, that person would almost certainly have died but it's entirely possible you could leave those little bits of invasive tumours behind and nothing happen to that person at all. Or maybe you do nothing and they're dead within 5 years. But then, maybe you're cautious and decide to do a re-excision over a complete mastectomy for cosmetic reasons coupled with a metric ****-ton of radiotherapy to nuke those little bastards. Or maybe you're really cautious and decide, nope, take the whole lot and take the other breast while you're at it. Maybe you were really, really cautious and bullied your patient into neoadjuvant chemo (i.e. prior to any surgery) to shrink it followed by a lumpectomy + radio. And maybe they still had a recurrence and died anyway.

Depending on which pathway the cancer team took, the same person with the same cancer could fall in or out of the 'over-diagnosed' box of the study.

[Top_Cat]

Guest | Why Climate Deniers Have No Scientific Credibility - In One Pie Chart

[Top_Cat]

Noted botanist Ray Comfort takes on the science behind the theories of esteemed paleobiologist John Lennon and creates a documentary pretty much comprised of him yelling at clouds.

[benchmark00]

Ray 'Southern' Comfort.