Patterico's Pontifications

3/30/2005

$100,000 Challenge

Filed under: Schiavo — Patterico @ 6:26 pm

Interesting challenge from CodeBlueBlog. He will show any neurologist 100 CAT scans of brains, and ask them to determine whether the patients in question are PVS. If they get at least 60 of 100 right, they will win $100,000.

To play, you need only put up $25,000.

All of you commenters who say Terri Schiavo’s CAT scan was conclusive proof that she is PVS — round up your neurologist and go play CodeBlueBlog’s game. If you’re right, it’s easy money.

(Via Xrlq.)

15 Responses to “$100,000 Challenge”

  1. A measly 4-to-1 payoff? Are you kidding? This is a sucker’s bet – no one will ever collect.

    One of the issues at dispute here is whether Terri Schiavo is in PVS. Some say yes, others say no.

    There doesn’t seem to be a 100% objective standard as to what is and what isn’t a case of PVS – as a result, intelligent people who have lots of initials after their names can reach different conclusions looking at the same CAT scan (and being able to observe the patient, something unlikely to be made available to any of our contestants).

    As a result, the likelihood of any given neurologist – no matter how good – matching CodeBlueBlog’s assessment on at least 60 of the 100 scans is a real long shot… even more so if, as I would expect, no CAT scans of ‘normal/undamaged’ will be included in the test.

    Such a bet is akin to trying to match 60 numbers out of a 100 number Keno game – were such a game to exist, the payoff would be a lot higher than 4-to-1 to match the astronomical odds against winning. Not that I’d expect anyone to win.

    Which may be exactly the point he’s trying to make.

    steve sturm (0b7c1a)

  2. Actually Terri Schiavo’s CAT scan image could be conclusive proof of PVS but CodeBlue’s challenge a bad bet. This would be the case if some CAT scan images (like say Schiavo’s) are conclusive but others (obviously the ones that CodeBlue would pick for his challenge) are ambiguous.

    That said it would not surprise me if in fact some commentators are overstating the conclusions that can be drawn from this single image (note my understanding is that a CAT scan contains many such images).

    James B. Shearer (fc887e)

  3. CodeBlue says hers is ambiguous, and is being misrepresented as conclusive. Hence the challenge.

    Patterico (756436)

  4. Well–since CodeBlueBlog will obviously be taking a few challengers $ 25 K, the next conclusion is that upon autopsy, Terri S. will have been found to have died of heart failure associated with a very clear diagnosis of bulimia at the time of demise.

    Nanuk (cb61cc)

  5. steve,

    You have entirely misread the challenge, and stepped on my toes in the process (B.S. Mathematics). This is a true/false quiz, and you have to be right 60% of the time to collect. If you flip a coin, and count heads=PVS, tails=no PVS, you can expect to get 50% correct. By my admittedly rusty calculations, you have a 3% chance of winning by doing just that. If you could just be correct 57% of the time, you would expect to break even (ie. you would only need to be sure of about 1 out of 7, and just coin-flip the rest). If no neurologist in the Schiavo case can be sure of a PVS diagnosis/non-diagnosis in even 1 out of 7 cases, based purely on a CT slice, then the only reason to be trotting out that CT slice is for its prejudicial effect on the public, which I believe is CodeBlue’s point.

    (If someone checks my numbers and gets a different answer, post your reasoning, and I’ll post mine. Otherwise, I don’t want to clutter Patterico’s fine blog with a lot of hard to read calculations.)

    Ben Zeen (a pseudonym) (c6f653)

  6. Let’s see the same CT “cut” showing the same or worse damage as does Terri Schiavo’s publicized 1996 cut, with the subject demonstrating conscious functioning. CodeBlue’s challenge is irrelevant, imo. And with CodeBlue selecting the “cut”, no one should take the challenge otherwise. You’d basically have to read his mind without even looking at the cuts.

    J. Peden (ffccb8)

  7. OK, so CodeBlue is trying to pull a fast one and get rich on this scheme. There is a lot of judgement involved in reading a CAT scan, so nobody will be able to get 60% “right.”

    In Terri’s case though, we’ll just take the first interpretation (by a doctor hired by someone with conflicts of interest) and we’ll kill based on that.

    Mike (d3f5fd)

  8. Patterico you are missing my point, Terri Schiavo’s cat scan (or at least the single image floating around on the internet) may well be ambiguous but CodeBlue’s challenge won’t prove it. Some CAT scans are ambiguous (such as to be extreme when the scan is botched and you can’t see anything). CodeBlue could select 100 such scans and win his challenge but this would not show that Terri Schiavo’s scan is ambiguous.

    James B. Shearer (fc887e)

  9. What a silly, meaningless stunt.

    Let’s say we have a picture of a parrot whose head has been cut off. I would say that is a picture of a dead parrot.

    Suppose someone, call her CodeFuchsia, comes along and says, “Actually, it’s ambiguous. I’d say the parrot is sleeping. And if you think you’re such a smarty-pants, I’m going to show you 100 parrot pictures. You tell me whether the parrot is dead or sleeping.”

    Then CodeFuchsia shows you 100 pictures of parrots who are maybe sleeping or maybe dead. Or maybe they’re just stunned. Yes, you can’t tell which it is in many of those cases. Does that also make you incapable of recognizing a headless bird as an ex-parrot?

    So let’s stick with the original picture. 100 other random pics tell you nothing.

    m.croche (1c16ec)

  10. OK, Clod, let’s do that. Let’s stick with the original picture. As applied to parrots, it’s a crappy snapshot taken on low-speed film from a moving vehicle that was at least 100 feet from the parrot. Upon viewing that photo, one can barely make out the parrot’s body, but no one cares about that since everybody knows it has one. As to the head, the photo is much to fuzzy to tell, but one can be pretty sure that the head is either missing or unusually (though not implausibly) small. We have better cameras available, but hey, who needs those? Instead, let’s just starve the parrot to death. Now, almost two weeks later, you’re finally right. It is it’s an ex-parrot. Hope you’re happy.

    Xrlq (c51d0d)

  11. One thing some here are missing. Part of the point of Codeblue’s challenge is that you *cannot* (contra Shearer above) diagnose PVS from a CT scan. “PVS” is a clinical diagnosis. Yet the CT is being paraded out in MSM as proof (or something) of her PVS condition. Codeblue is calling foul, by noting that if this made sense, then one should be able to post-diagnose other PVS patients solely based on a CT. By saying this is a sucker’s bet etc., you’re just conceding his point. If, like Shearer above, you believed that a CT *could* show “conclusive proof of PVS”, you should put your money where your mouth is cuz you have an easy $100k just waiting for the asking.

    The other objection raised above is that Codeblue would get to selectively choose which CT slice from each patient to show. That is a valid point; Codeblue should allow the the challenger to choose which slice he wants. Otherwise he could pick all slices that showed only the top of the skull or whatever.

    But I’ll bet Codeblue would be happy to let you pick the slices. Ask him!

    Blixa (7a6725)

  12. Does that also make you incapable of recognizing a headless bird as an ex-parrot?

    Or are you claiming you’ve examined the CT scan and determined that? If not, what are you talking about?

    GeraldA (055b6e)

  13. Neither. It means that a CT scan doesn’t provide enough detail to tell. An MRI or PET scan would have come much closer, but Michael Schiavo didn’t allow either, lest the “fact” that Terri was “headless” be disturbed by any … um … facts?

    Xrlq (c51d0d)

  14. Blixa, you are missing my point. Perhaps you can’t ever properly diagnose PVS from a single CT slice as CodeBlue claims. But purely as a matter of logic, CodeBlue’s challenge can’t show this. Suppose hypothetically for example that when a CT image shows a normal brain the patient is never PVS, when the CT image shows moderate damage the patients is sometimes PVS and sometimes isn’t and when the image shows severe damage the patient is always PVS (or worse). CodeBlue could select 100 images which showed moderate damage for his challenge and demonstrate that you cannot always determine PVS status from a single image but this would not show that you cannot diagnose PVS (or worse) status from a single image showing severe damage. CodeBlue could better make his point by producing images showing worse (or comparable) damage than Schiavo’s from patients who are not PVS.

    Also of course as a practical matter it is not prudent to accept such bets unless there is a mutual trusted stakeholder to hold the stakes and arbitrate disputes.

    James B. Shearer (fc887e)

  15. Croche,

    Your parrot argument assumes that Terri Schiavo’s CT scan is conclusive. CodeBlueBlog says it ain’t. He says it’s like the CT scan of the brain of an old person going a little senile. And, he says, neurologists are unqualified to read these things anyway.

    What do I know? I’m not a doctor. But I’m guessing you aren’t either. You have your lefty doctors whom you believe, and to hell with the opinion of someone from the Mayo Clinic if he believes in Jesus. You, M.Croche, know better.

    Patterico (756436)


Powered by WordPress.

Page loaded in: 0.2285 secs.