Health progress

[kdh]

So how do those statistical methods allow for a large number of confounding variables?

Are we talking about longitudinal studies here?

AI techniques are my ready answer. We won't be driving cars in congested areas with them any time soon but medical diagnosis is an easier problem, as far as I can tell.

But I'm a statistician, not a physician.

[Orestes Munn]

Beau, my understanding is that there's a lot of confusion between a CT angiogram (CCTA), which uses an iodine-rich contrast injection and a plain old CT scan of the heart used to get a calcium score. The former involves significant radiation, that latter the same as a standard X-ray.

Heartflow uses a basic CT scan, no dye.

I haven't dug into the last 18 months of Heartflow's process, but when I was working there we definitely used Iodine contrast. My objection to the comment about radiation is that it's out of date. Modern CTs have much better receivers and can operate on far less power. They also have much higher resolution and yield far better results. I've dropped an email to the Clinical head to see if I'm out of date.

I don't think we ever use calcium score as the bases of the FFRct.

I guess I'm also confused.

I've heard that the preference for a standard, through-your-groin angiogram is from its being a setting that allows implanting a stent if necessary, another questionable practice.

Death is a certainty but in any given year it's not all that probable and as such if we look for something bad a false you-might-die-of-this is likely and believable because, well, we all die eventually.

Think of the likelihood of someone winning the lottery twice. Unlikely, right? No. The probability that I win the lottery twice is epsilon but anyone?

What's the probability that I win the lottery once? Still miniscule. That someone wins it is a certainty.

I'm with OM. Getting medical diagnoses and treatments right screams for large-sample statistical methods, not human experts.

And a rock solid corollary of that proposition is that medical decision making should be done by machines.

[Jamie]

Isn't that Watson's, (the computer on Jeopardy), day job?

https://en.wikipedia.org/wiki/Watson_(computer)

[LarryHoward]

Beau, my understanding is that there's a lot of confusion between a CT angiogram (CCTA), which uses an iodine-rich contrast injection and a plain old CT scan of the heart used to get a calcium score. The former involves significant radiation, that latter the same as a standard X-ray.

Heartflow uses a basic CT scan, no dye.

I haven't dug into the last 18 months of Heartflow's process, but when I was working there we definitely used Iodine contrast. My objection to the comment about radiation is that it's out of date. Modern CTs have much better receivers and can operate on far less power. They also have much higher resolution and yield far better results. I've dropped an email to the Clinical head to see if I'm out of date.

I don't think we ever use calcium score as the bases of the FFRct.

I guess I'm also confused.

I've heard that the preference for a standard, through-your-groin angiogram is from its being a setting that allows implanting a stent if necessary, another questionable practice.

Death is a certainty but in any given year it's not all that probable and as such if we look for something bad a false you-might-die-of-this is likely and believable because, well, we all die eventually.

Think of the likelihood of someone winning the lottery twice. Unlikely, right? No. The probability that I win the lottery twice is epsilon but anyone?

What's the probability that I win the lottery once? Still miniscule. That someone wins it is a certainty.

I'm with OM. Getting medical diagnoses and treatments right screams for large-sample statistical methods, not human experts.

And a rock solid corollary of that proposition is that medical decision making should be done by machines.

Interesting discussion. My clinical oncologist would argue that there is no formula for the treatment’s I’ve had. He and I banter about his assertatiin that “the body is not a machine and I have to consider how you are reacting at each step so we can’t really boil it down to A+B = C. I respond with Its a biochemical and mechanical machine and we can address things in processes and procedures based on trials, experience and studies and adjust based on feedback. I think at the bleeding edge it’s probably a bit different. My radiation treatments were denied by the insurance company ($62k) because there are no studies that prove effectiveness as a standard of care for the broader area of “soft tissue sarcoma of the retroperitonium” as there is no specific ICD code for a Ewing’s Sarcoma of the Retroperitonium. Still discussing that little gem of an insurance response. Same with the surgery in that most surgeons (and the chief at Georgetown (and the chief of Oncology Surgery at Georgetown told me I was inoperablej. Sloan-Kettering said “Tricky, but doable. Let’s do Chemo and maybe radiation first to try to shrink it and take out some of the surgical risk.” Given the somewhat subjective line between surgery or not, I wonder what Watson would say. I get the percentages. My undergrad is in Ops Research. Like the lottery question, some percentage will benefit from the “this expensive treatment is going to be effective in 3-5% of the patients.” How does Watson pick the 3-5%. What’s the decision criteria. Who chooses the criteria based on what data? I can guarantee that the small % that do benefit will take the “screw the numbers” position.

[Ajax]

This is incredible. You may very well be leading the way in how this condition is treated in the future, and it could result in insurance companies covering the radiation treatment for future patients.
I sure as hell hope you get it covered for you.

[kdh]

Larry, to me in your case the available sample size is way too small to make reasonable statistical inferences--without a large sample there's no benefit from statistical/machine learning methods.

[Orestes Munn]

Beau, my understanding is that there's a lot of confusion between a CT angiogram (CCTA), which uses an iodine-rich contrast injection and a plain old CT scan of the heart used to get a calcium score. The former involves significant radiation, that latter the same as a standard X-ray.

Heartflow uses a basic CT scan, no dye.

I haven't dug into the last 18 months of Heartflow's process, but when I was working there we definitely used Iodine contrast. My objection to the comment about radiation is that it's out of date. Modern CTs have much better receivers and can operate on far less power. They also have much higher resolution and yield far better results. I've dropped an email to the Clinical head to see if I'm out of date.

I don't think we ever use calcium score as the bases of the FFRct.

I guess I'm also confused.

I've heard that the preference for a standard, through-your-groin angiogram is from its being a setting that allows implanting a stent if necessary, another questionable practice.

Death is a certainty but in any given year it's not all that probable and as such if we look for something bad a false you-might-die-of-this is likely and believable because, well, we all die eventually.

Think of the likelihood of someone winning the lottery twice. Unlikely, right? No. The probability that I win the lottery twice is epsilon but anyone?

What's the probability that I win the lottery once? Still miniscule. That someone wins it is a certainty.

I'm with OM. Getting medical diagnoses and treatments right screams for large-sample statistical methods, not human experts.

And a rock solid corollary of that proposition is that medical decision making should be done by machines.

Interesting discussion. My clinical oncologist would argue that there is no formula for the treatment’s I’ve had. He and I banter about his assertatiin that “the body is not a machine and I have to consider how you are reacting at each step so we can’t really boil it down to A+B = C. I respond with Its a biochemical and mechanical machine and we can address things in processes and procedures based on trials, experience and studies and adjust based on feedback. I think at the bleeding edge it’s probably a bit different. My radiation treatments were denied by the insurance company ($62k) because there are no studies that prove effectiveness as a standard of care for the broader area of “soft tissue sarcoma of the retroperitonium” as there is no specific ICD code for a Ewing’s Sarcoma of the Retroperitonium. Still discussing that little gem of an insurance response. Same with the surgery in that most surgeons (and the chief at Georgetown (and the chief of Oncology Surgery at Georgetown told me I was inoperablej. Sloan-Kettering said “Tricky, but doable. Let’s do Chemo and maybe radiation first to try to shrink it and take out some of the surgical risk.” Given the somewhat subjective line between surgery or not, I wonder what Watson would say. I get the percentages. My undergrad is in Ops Research. Like the lottery question, some percentage will benefit from the “this expensive treatment is going to be effective in 3-5% of the patients.” How does Watson pick the 3-5%. What’s the decision criteria. Who chooses the criteria based on what data? I can guarantee that the small % that do benefit will take the “screw the numbers” position.

Exactly. You are in one of those relatively rare, but clinically important, situations where there are very few data and you and your docs are going on scant clinical reports, basic science, and the experience of a few individuals, because that's all there is. However, to take the easiest example, the rule-out strokes and heart attacks that present to emergency rooms every night still cause lots of head scratching, seat of the pants flying, second guessing, and decision making that looks bad the following day at rounds, on the hospital balance sheet, or next year in court. Simple weighted scoring algorithms can give you numerical probabilities for all kinds of things and would save lots of lives and money.

By the way, you don't need Watson to do it. When I was an intern, there was an algorithm for probability of MI that could be implemented on a programmable handheld calculator--remember those? In the paper describing it, it was more accurate than a panel of ER docs in real-life rule-outs. It was treated as a novelty. However, maybe in these days of autonomous vehicles...

[kdh]

When I was an intern, there was an algorithm for probability of MI that could be implemented on a programmable handheld calculator--remember those? In the paper describing it, it was more accurate than a panel of ER docs in real-life rule-outs. It was treated as a novelty. However, maybe in these days of autonomous vehicles...

Kahneman talks about these formulae in his book, and gives plain reasons for their superiority based on human flaws. One is that at least a formula provides the same conclusion given the same inputs. Not so for humans.

I work in automated stock picking. My old boss was laughed out of the room for proposing the idea at Fidelity in 1985. Peter Lynch is notable for liking the idea. Fast forward to 2017 practically all stock picking is done this way and the markets are more efficient.

[Orestes Munn]

When I was an intern, there was an algorithm for probability of MI that could be implemented on a programmable handheld calculator--remember those? In the paper describing it, it was more accurate than a panel of ER docs in real-life rule-outs. It was treated as a novelty. However, maybe in these days of autonomous vehicles...

Kahneman talks about these formulae in his book, and gives plain reasons for their superiority based on human flaws. One is that at least a formula provides the same conclusion given the same inputs. Not so for humans.

I work in automated stock picking. My old boss was laughed out of the room for proposing the idea at Fidelity in 1985. Peter Lynch is notable for liking the idea. Fast forward to 2017 practically all stock picking is done this way and the markets are more efficient.

People have inordinate faith in the probity of physicians and seem highly swayed by the factual, but unhelpful, idea that "every patient is different." Precision, genomic-based, medicine is going to be a reality soon for everyone, but that's just going to erode the doctor's cognitive role further.

[LarryHoward]

Well, just to prove I don’t have a clue. Had a CT and met with the surgeon today. No changes in the scan but he wants to go ahead with surgery and clinical thinks we may be done with chemo as they have hit my reserves pretty hard.

Basically, the thinking seems to be that the tumor may be dead but they can’t prove it and it’s aggressive enough that only a small remnant is required for it to regrow. I may not have the reserves to deal with a repeat of the high dose course of chemo again so based on that, the surgeon is planning to take it out 27 Nov. In the next couple of weeks I will meet with the vascular surgeon from Cornell (I think) who will be in the OR to do the possible/probable aorta to arterial graft. Surgeons comment was that it’s a tricky surgery but not overly risky other than normal risks of major abdominal and vascular surgery on a 63 YO. They will evaluate what they see when I’m on the table and tell me what happened when I wake up.

So as of tonight, it appears I’m done with chemo and will be up and about from the surgery and back home by mid December. Might even have some hair by Christmas. I need to confirm that with the Clinical tomorrow or Thursday and have lots to ponder and clarify over the next week or so.

Well, finally connected with the clinical oncologist. He still plans on chemo after surgery so once I recover from surgery, I’ll have 4 more rounds of surgery if my body can handle it, pushing the end of treatment until late March or so.

[Orestes Munn]

Well, just to prove I don’t have a clue. Had a CT and met with the surgeon today. No changes in the scan but he wants to go ahead with surgery and clinical thinks we may be done with chemo as they have hit my reserves pretty hard.

Basically, the thinking seems to be that the tumor may be dead but they can’t prove it and it’s aggressive enough that only a small remnant is required for it to regrow. I may not have the reserves to deal with a repeat of the high dose course of chemo again so based on that, the surgeon is planning to take it out 27 Nov. In the next couple of weeks I will meet with the vascular surgeon from Cornell (I think) who will be in the OR to do the possible/probable aorta to arterial graft. Surgeons comment was that it’s a tricky surgery but not overly risky other than normal risks of major abdominal and vascular surgery on a 63 YO. They will evaluate what they see when I’m on the table and tell me what happened when I wake up.

So as of tonight, it appears I’m done with chemo and will be up and about from the surgery and back home by mid December. Might even have some hair by Christmas. I need to confirm that with the Clinical tomorrow or Thursday and have lots to ponder and clarify over the next week or so.

Well, finally connected with the clinical oncologist. He still plans on chemo after surgery so once I recover from surgery, I’ll have 4 more rounds of surgery if my body can handle it, pushing the end of treatment until late March or so.

Oof! What a long road. Presumably, they'll want you pretty recovered by the time of surgery, so maybe it'll be more like starting from scratch.

[LarryHoward]

Well, just to prove I don’t have a clue. Had a CT and met with the surgeon today. No changes in the scan but he wants to go ahead with surgery and clinical thinks we may be done with chemo as they have hit my reserves pretty hard.

Basically, the thinking seems to be that the tumor may be dead but they can’t prove it and it’s aggressive enough that only a small remnant is required for it to regrow. I may not have the reserves to deal with a repeat of the high dose course of chemo again so based on that, the surgeon is planning to take it out 27 Nov. In the next couple of weeks I will meet with the vascular surgeon from Cornell (I think) who will be in the OR to do the possible/probable aorta to arterial graft. Surgeons comment was that it’s a tricky surgery but not overly risky other than normal risks of major abdominal and vascular surgery on a 63 YO. They will evaluate what they see when I’m on the table and tell me what happened when I wake up.

So as of tonight, it appears I’m done with chemo and will be up and about from the surgery and back home by mid December. Might even have some hair by Christmas. I need to confirm that with the Clinical tomorrow or Thursday and have lots to ponder and clarify over the next week or so.

Well, finally connected with the clinical oncologist. He still plans on chemo after surgery so once I recover from surgery, I’ll have 4 more rounds of surgery if my body can handle it, pushing the end of treatment until late March or so.

Oof! What a long road. Presumably, they'll want you pretty recovered by the time of surgery, so maybe it'll be more like starting from scratch.

Yeah, just ready to get things done and get back to my life. It will end up as a year lost to treatment and I’m having symptoms that suggest I have a knee replacement waiting in the wings, possible as soon as I recover from the chemo. Sucks for work and personal life. Tired of being tired and low energy. All of that pissing and whining stuff. I’ve had it pretty well overall, but now understand those who choose to forgo treatment and seek palliative care only as it totally consumes you life. This shit isn’t for weenies as they say. I’m eternally thankful for a solid medical team, supportive friends and my family.

[BeauV]

Larry,

Sorry to hear you'll have an extended sentence. But it sure seems that you're getting good care. My old man checked out after learning that all he efforts would only extend his life a few months. He went sailing with us and died a happy guy. It was hard for me to accept, but you've made the case for choice and I'm glad you're in control the way he was.

Having watched, what you're going through is NOT for the faint folks. I think you're holding up darn well and you're certainly a inspiration to the rest of us.

Hang in there,

B

[kimbottles]

Larry,

Sorry to hear you'll have an extended sentence. But it sure seems that you're getting good care. My old man checked out after learning that all he efforts would only extend his life a few months. He went sailing with us and died a happy guy. It was hard for me to accept, but you've made the case for choice and I'm glad you're in control the way he was.

Having watched, what you're going through is NOT for the faint folks. I think you're holding up darn well and you're certainly a inspiration to the rest of us.

Hang in there,

B

+1

[Ish]

Larry,

Sorry to hear you'll have an extended sentence. But it sure seems that you're getting good care. My old man checked out after learning that all he efforts would only extend his life a few months. He went sailing with us and died a happy guy. It was hard for me to accept, but you've made the case for choice and I'm glad you're in control the way he was.

Having watched, what you're going through is NOT for the faint folks. I think you're holding up darn well and you're certainly a inspiration to the rest of us.

Hang in there,

B

+1

+2

[Olaf Hart]

Larry,

Sorry to hear you'll have an extended sentence. But it sure seems that you're getting good care. My old man checked out after learning that all he efforts would only extend his life a few months. He went sailing with us and died a happy guy. It was hard for me to accept, but you've made the case for choice and I'm glad you're in control the way he was.

Having watched, what you're going through is NOT for the faint folks. I think you're holding up darn well and you're certainly a inspiration to the rest of us.

Hang in there,

B

+1

+2

+3

I really like that everyone on the team is playing for a win ...

Was in Ho Chi Min who said “in every battle the side that wins is the side that can absorb the most pain”?

[Ish]

Larry,

Sorry to hear you'll have an extended sentence. But it sure seems that you're getting good care. My old man checked out after learning that all he efforts would only extend his life a few months. He went sailing with us and died a happy guy. It was hard for me to accept, but you've made the case for choice and I'm glad you're in control the way he was.

Having watched, what you're going through is NOT for the faint folks. I think you're holding up darn well and you're certainly a inspiration to the rest of us.

Hang in there,

B

+1

+2

+3

I really like that everyone on the team is playing for a win ...

Was in Ho Chi Min who said “in every battle the side that wins is the side that can absorb the most pain”?

Oh, that's cheerful. Do you have any happy quotes from the Marquis de Sade to accompany that?

[Jamie]

Larry,

Sorry to hear you'll have an extended sentence. But it sure seems that you're getting good care. My old man checked out after learning that all he efforts would only extend his life a few months. He went sailing with us and died a happy guy. It was hard for me to accept, but you've made the case for choice and I'm glad you're in control the way he was.

Having watched, what you're going through is NOT for the faint folks. I think you're holding up darn well and you're certainly a inspiration to the rest of us.

Hang in there,

B

+1

+2

+3

I really like that everyone on the team is playing for a win ...

Was in Ho Chi Min who said “in every battle the side that wins is the side that can absorb the most pain”?

Oh, that's cheerful. Do you have any happy quotes from the Marquis de Sade to accompany that?

How about, "Your body is the church where Nature asks to be reverenced."

But more seriously: Larry you're getting the aggressive treatment because you can take it and it will lead to a more assured outcome.

[Olaf Hart]

Larry,

Sorry to hear you'll have an extended sentence. But it sure seems that you're getting good care. My old man checked out after learning that all he efforts would only extend his life a few months. He went sailing with us and died a happy guy. It was hard for me to accept, but you've made the case for choice and I'm glad you're in control the way he was.

Having watched, what you're going through is NOT for the faint folks. I think you're holding up darn well and you're certainly a inspiration to the rest of us.

Hang in there,

B

+1

+2

+3

I really like that everyone on the team is playing for a win ...

Was in Ho Chi Min who said “in every battle the side that wins is the side that can absorb the most pain”?

Oh, that's cheerful. Do you have any happy quotes from the Marquis de Sade to accompany that?

Seems to me Larry has already absorbed more suffering than most could bear, and deserves a win, sorry if my comments dont convey that message.

[LarryHoward]

Thanks guys. Just having a bad day. Between a swollen knee and being reduced to a hobbling with a cane and having a “miscommunication” between the clinical and the surgeon that gave the surgeon and me the impression that the clinicL felt I’d had all the chemo I’d be getting reversed mid day yesterday to “up to 4 sessions of chemo after you recover from surgery” pushing back the end of treatment 3-4 months, I was a pretty unhappy camper. That and paying a diver $200 to scrub the bottom and change zincs on a boat that didn’t leave the dock this year.

Was kinda planning a healthy spring to get ready for next season.

[BeauV]

Beau, my understanding is that there's a lot of confusion between a CT angiogram (CCTA), which uses an iodine-rich contrast injection and a plain old CT scan of the heart used to get a calcium score. The former involves significant radiation, that latter the same as a standard X-ray.

Heartflow uses a basic CT scan, no dye.

I haven't dug into the last 18 months of Heartflow's process, but when I was working there we definitely used Iodine contrast. My objection to the comment about radiation is that it's out of date. Modern CTs have much better receivers and can operate on far less power. They also have much higher resolution and yield far better results. I've dropped an email to the Clinical head to see if I'm out of date.

I don't think we ever use calcium score as the bases of the FFRct.

I guess I'm also confused.

I've heard that the preference for a standard, through-your-groin angiogram is from its being a setting that allows implanting a stent if necessary, another questionable practice.

Death is a certainty but in any given year it's not all that probable and as such if we look for something bad a false you-might-die-of-this is likely and believable because, well, we all die eventually.

Think of the likelihood of someone winning the lottery twice. Unlikely, right? No. The probability that I win the lottery twice is epsilon but anyone?

What's the probability that I win the lottery once? Still miniscule. That someone wins it is a certainty.

I'm with OM. Getting medical diagnoses and treatments right screams for large-sample statistical methods, not human experts.

Keith, I heard from the clinical head of HeartFlow, they use Iodine Contrast but the power of the X-Ray for the CT is now the same for Calcium Scoring and for CTs for HeartFlow.

I completely agree with your assessment that "...through-your-groin angiogram is from its being a setting that allows implanting a stent if necessary, another questionable practice." When the two technical/medical folks approached me about HeartFlow, the primary motivation was to limit the number of questionable stents which were installed. It seems that some doctors will "take a look" at the renal arteries as they go by and perform what my friends call a "Drive By Stent" if it's even close.

As you can imagine, there is a significant argument going on in cardiology as HeartFlows FFRct is showing as good or better results without the side effects of an invasive procedure; it's also about $4,000 less per patient. My understanding is that it used to be that the only argument in favor of the invasive angiography was that you could do the stents while you were there, sadly that seems to have encouraged stents in marginal situations.

[LarryHoward]

Well, almost time. Sitting in NY and enjoying my presurgical liquid diet. Checking in tomorrow morning at 0730 for 0915 in the OR. Team play with the top Sarcoma surgeon in the US at Memorial Sloan-Kettering and his sidekick vascular surgeon from Cornell to get this thing out and do an arterial/aorta graft of necessary (betting is that it is).

SWMBO, the kids and my sister are out for dinner and promised me some pho broth when they return. Lynne will let Mrs OM and Mrs Ajax know when I’m out and in recovery.

Thanks for all the support to date.

Larry

75D35C45-EE82-48F0-97E7-F47BAFFC7420.jpeg

[kimbottles]

Kick ass Captain!

[Ajax]

I'm pulling for you, Larry. Looking forward to getting you home and recuperating.

[Slick470]

Get that bitch out of there. Kick it's ass Larry!

[BeauV]

All fingers and toes are crossed. Sending all my best thoughts and wishes your way.

[Orestes Munn]

Text just received:

Yay! From Lynne: Doc just left - surgery went great, able to remove tumor without need for grafts or bypasses. Tumor appears dead but being sent for testing - should be able to see him in about 90 minutes.

[Tim Ford]

Great news! Thanks for the update, Mr. OM

[BeauV]

Eric, that's great news!! Thank you so much for posting here so we can all cheer!!! B

[Orestes Munn]

The news will occasion a celebratory foray into the cache of good whiskey I am sort-of saving for the old boy.