8 ms·

Which can entirely be explained by the fact that we have gotten way way better at detecting cancers since 1993.

by sunbum 5y ago

cl42 5y ago
Beautiful point! I remember reading about this a few years ago, and it is a fantastic confounding factor. I wonder -- do cancer studies take cancer detection rates into account?
- OJFord 5y ago
  Not GP and I don't know, and this suffers the same problem (but in a way it doesn't matter) but for what you're interested in here you can just look at mortality (per general population, not patients): https://www.cancerresearchuk.org/health-professional/cancer-statistics/statistics-by-cancer-type/head-and-neck-cancers/mortality#heading-Two https://www.cancerresearchuk.org/health-professional/cancer-... which is roughly stable over the same period. Which must mean treatment (or early detection) got better, and you can say 'oh it would be even better if it weren't for ~5G~ detecting it in more people', but I say in a way it doesn't matter because it shows we're sort of.. 'coping with it', and wanting it to be better is a different conversation than whether we're making it worse.
  - cl42 5y ago
    Thanks, this is really helpful and I like how you're looking at it.
newbamboo 5y ago
I’m reminded of the history of the science establishing the link between cigarettes and cancer. All the same arguments and similar conflicted interests. Took many decades before the mainstream accepted the linkage and of course the industry still denies the science.
- dekhn 5y ago
  It gets even better. After the linkage was accepted one of the founders of modern statistics continued to fight against the conclusion of causality: https://priceonomics.com/why-the-father-of-modern-statistics-didnt-believe/ https://priceonomics.com/why-the-father-of-modern-statistics... The similarities to modern public health calculus (around covid), as well as the point that a single smart person can create significant doubt where should not be any (IE, he wasn't operating in good faith), are quite extraordinary.
pessimizer 5y ago
Which might entirely be explained by us getting better at detecting cancers, but I would have to see evidence. The purpose of cancer detection is to be as early as possible, allowing more time for treatment. It is not to count cancers. I would assume the vast majority of cancers are eventually detected, unless you die of something else before they become a problem. So, unless I'm missing something, your claim is that at least 33% to 50% of head and neck cancers would not have ever become identifiable as cancers without improvements made since 1993, which I think is insanely high. edit: and seeming high doesn't make it impossible, I know that there are a lot of slow growing prostate cancers where people die of old age before ever noticing them. But we know this because we checked for them after people were dead. Is there anything similar connected to particular head and neck cancers that you can point out?
- true_religion 5y ago
  Most of the time, if someone is old and dies in their sleep, then no one goes looking for an autopsy. The spread of cancer can be really dramatic after it metastasizes so you can go from looking just a bit unwell, to dead in a few months. And even if cancer is present, and an autopsy is done… will it get listed as cause of death when there are other conditions present that don’t need lab reports to verify them?
  - pessimizer 5y ago
    That's my point. Are at least a quarter to a third of head and neck cancers going undiagnosed because the sufferer dies of something else before they are noticed? In order for the count of head and neck cancers to increase by a third or a half due to earlier diagnosis, without an underlying increase in the actual number of head and neck cancers, this has to be true.

Which can entirely be explained by the fact that we have gotten way way better at detecting cancers since 1993.

by sunbum 5y ago

Which can entirely be explained by the fact that we have gotten way way better at detecting cancers since 1993.

cl42 5y ago

Beautiful point! I remember reading about this a few years ago, and it is a fantastic confounding factor. I wonder -- do cancer studies take cancer detection rates into account?

OJFord 5y ago
Not GP and I don't know, and this suffers the same problem (but in a way it doesn't matter) but for what you're interested in here you can just look at mortality (per general population, not patients): https://www.cancerresearchuk.org/health-professional/cancer-statistics/statistics-by-cancer-type/head-and-neck-cancers/mortality#heading-Two https://www.cancerresearchuk.org/health-professional/cancer-... which is roughly stable over the same period. Which must mean treatment (or early detection) got better, and you can say 'oh it would be even better if it weren't for ~5G~ detecting it in more people', but I say in a way it doesn't matter because it shows we're sort of.. 'coping with it', and wanting it to be better is a different conversation than whether we're making it worse.
- cl42 5y ago
  Thanks, this is really helpful and I like how you're looking at it.

newbamboo 5y ago

I’m reminded of the history of the science establishing the link between cigarettes and cancer. All the same arguments and similar conflicted interests. Took many decades before the mainstream accepted the linkage and of course the industry still denies the science.

dekhn 5y ago
It gets even better. After the linkage was accepted one of the founders of modern statistics continued to fight against the conclusion of causality: https://priceonomics.com/why-the-father-of-modern-statistics-didnt-believe/ https://priceonomics.com/why-the-father-of-modern-statistics... The similarities to modern public health calculus (around covid), as well as the point that a single smart person can create significant doubt where should not be any (IE, he wasn't operating in good faith), are quite extraordinary.

pessimizer 5y ago

Which might entirely be explained by us getting better at detecting cancers, but I would have to see evidence. The purpose of cancer detection is to be as early as possible, allowing more time for treatment. It is not to count cancers. I would assume the vast majority of cancers are eventually detected, unless you die of something else before they become a problem. So, unless I'm missing something, your claim is that at least 33% to 50% of head and neck cancers would not have ever become identifiable as cancers without improvements made since 1993, which I think is insanely high. edit: and seeming high doesn't make it impossible, I know that there are a lot of slow growing prostate cancers where people die of old age before ever noticing them. But we know this because we checked for them after people were dead. Is there anything similar connected to particular head and neck cancers that you can point out?

true_religion 5y ago
Most of the time, if someone is old and dies in their sleep, then no one goes looking for an autopsy. The spread of cancer can be really dramatic after it metastasizes so you can go from looking just a bit unwell, to dead in a few months. And even if cancer is present, and an autopsy is done… will it get listed as cause of death when there are other conditions present that don’t need lab reports to verify them?
- pessimizer 5y ago
  That's my point. Are at least a quarter to a third of head and neck cancers going undiagnosed because the sufferer dies of something else before they are noticed? In order for the count of head and neck cancers to increase by a third or a half due to earlier diagnosis, without an underlying increase in the actual number of head and neck cancers, this has to be true.