Bad calculations also tend to be distributed widely, with people saying things like “one pint of blood can save up to three lives” when the expected marginal lives saved is actually tiny.
Just in the past week I was trying to figure out the math behind that statistic. I couldn’t find actual studies on the topic that would let me calculate the expected utility of donating blood. Do you happen to know said information?
I bet it’s that 1⁄3 of a pint is pretty much the minimum amount of blood needed for a lifesaving transfusion (The Red Cross says the average transfusion size is three pints).
The expected utility of giving blood is currently very small, at least in the USA, because people are not being refused transfusions due to lack of blood. If that started happening blood drives would expand hugely and you’d know about it.
We do have shortages of other organs, such as kidneys, and with those you have around maybe a 50% chance of saving someone’s life if after you offer to donate one they find a match for you. If you’re able to start a chain of kidney swaps that wouldn’t happen otherwise, you may be able to get above one expected life saved per kidney.
Upon further thought, 50% may be too high for kidney donation. I was estimating that you’d only be giving your kidney to someone who would die otherwise (there are many more people who need kidney transplants than are available, so the replaceability effect should be absent.), and it had a 50% chance of working. While people who get a donated kidney do live longer than ones who stay on dialysis (and get to stop having to go in for dialysis) they tend to only live an extra 10-15 years.
As the lives of people with major kidney problems are worse than those without, maybe count each year as only 75% when quality weighting? So 50% chance of working 10-15 years 75% gives a very rough estimate of 4-6 QALYs per kidney donation. Still much better than blood, but with deworming at around $100/QALY donating $1K does more good than donating a kidney.
The expected utility of giving blood is currently very small, at least in the USA, because people are not being refused transfusions due to lack of blood. If that started happening blood drives would expand hugely and you’d know about it.
I have heard statistics (from sources that aren’t the red cross) that the supply of blood in the US is nearly always quite low. Running out rarely happens, although I did hear that lack of adequate blood reserves did cause a problem on 9/11. My impression is that the Red Cross typically has enough reserves for a few average days, but if something major happens, that supply can get used up pretty quickly.
Somewhere I heard that a single donation unit of blood can be made into three distinct products that could be given to three distinct patients. I assume that this possibility is the source of the statistic. Wikipedia is suggestive but not definitive.
Even if that’s right, it’s not a very useful statistic because it ignores the fairly common occurrence than more than one unit of medical thing will be needed for a single patient.
I hate to go off on a tangent, but:
Just in the past week I was trying to figure out the math behind that statistic. I couldn’t find actual studies on the topic that would let me calculate the expected utility of donating blood. Do you happen to know said information?
I bet it’s that 1⁄3 of a pint is pretty much the minimum amount of blood needed for a lifesaving transfusion (The Red Cross says the average transfusion size is three pints).
The expected utility of giving blood is currently very small, at least in the USA, because people are not being refused transfusions due to lack of blood. If that started happening blood drives would expand hugely and you’d know about it.
We do have shortages of other organs, such as kidneys, and with those you have around maybe a 50% chance of saving someone’s life if after you offer to donate one they find a match for you. If you’re able to start a chain of kidney swaps that wouldn’t happen otherwise, you may be able to get above one expected life saved per kidney.
Upon further thought, 50% may be too high for kidney donation. I was estimating that you’d only be giving your kidney to someone who would die otherwise (there are many more people who need kidney transplants than are available, so the replaceability effect should be absent.), and it had a 50% chance of working. While people who get a donated kidney do live longer than ones who stay on dialysis (and get to stop having to go in for dialysis) they tend to only live an extra 10-15 years.
As the lives of people with major kidney problems are worse than those without, maybe count each year as only 75% when quality weighting? So 50% chance of working 10-15 years 75% gives a very rough estimate of 4-6 QALYs per kidney donation. Still much better than blood, but with deworming at around $100/QALY donating $1K does more good than donating a kidney.
This is wrong: the 10-15 year estimate already takes into account rejections and other transplant failures. So I’m off by 50% and we get 8-12 QALYs.
The data is also from a study on kidneys from cadavers; live donated ones might be better.
Live ones apparently are better; I heard this before recently, and this seems to be right according to a few pages I checked, although they don’t cite specific studies: http://kidney-beans.blogspot.com/2009/08/living-kidney-donation-vs-cadaver.html or http://kidney.niddk.nih.gov/kudiseases/pubs/transplant/
I have heard statistics (from sources that aren’t the red cross) that the supply of blood in the US is nearly always quite low. Running out rarely happens, although I did hear that lack of adequate blood reserves did cause a problem on 9/11. My impression is that the Red Cross typically has enough reserves for a few average days, but if something major happens, that supply can get used up pretty quickly.
Somewhere I heard that a single donation unit of blood can be made into three distinct products that could be given to three distinct patients. I assume that this possibility is the source of the statistic. Wikipedia is suggestive but not definitive.
Even if that’s right, it’s not a very useful statistic because it ignores the fairly common occurrence than more than one unit of medical thing will be needed for a single patient.