Effectiveness of Giving Blood
I believe past discussion about giving blood within EA has undervalues doing so and hopefully the reasoning laid out here will show why.
Some caveats first; I think that some of the reasoning used here may only apply to certain blood groups: i.e. universal donors, additionally past discussion has focused on QALYs and reaches a different conclusion to me - it may be that using different metrics leads to different conclusions or that my method is flawed however due to the limited information about the stats of giving blood and limited discussion within EA (2 posts one against and one for blood donation being effective as far as I can tell) means having a discussion at al could be useful.
My reasoning
- Each whole blood donation produces 1 unit of blood (~about 1 pint although the exact amount is not important for this argument)[1]
- About 1/3 of blood donations are used for emergencies[2]
- I am assuming that in such an emergency the patient would die if they do not receive a blood transfusion
- When used in an emergency 8-10 units of universal donor blood is used initially[3] (I will use the conservative 10 units going forward)
- This means a 1 unit donation makes up 10% of the blood used/needed to save a life
Overall this means we have: 1/3 chance of a donation being used in an emergency * a 1/10 contribution to saving a life in the emergency equating to ~1/30 average life saved per donation.
This does not take into account that the 10 units is for the intial 24 hours of transfusions, after which other blood types can be used. This also does not take into account the other 2/3 times where if not saving a life the donation goes on to improve someone's well-being/health. Additionally the concrete information about blood donation and how it is used is lacking and claims about each donation saving up to 3 lives are inaccurate at best and misleading at worst. Furthermore statements of "1/3 of blood is used for emergencies" may be because 1/3 of blood donated is universal and thus 100% of a universal blood donator's blood may be used in an emergency - greatly improving the impact for them; this seems unlikely but these points hopefully show that the 1/30 estimate is likely to be conservative rather than optimistic.
This would make each donation roughly equivalent to a (1/30*5000) $166 donation to the Against Malaria Foundation for maybe an hour of time.
one point of criticism if this cost-effectiveness estimate: in high-income countries there is no substantial shortage of blood. In case of acute shortage, blood banks can easily recruit donors. So if you don't donate and that results in a blood shortage of one unit, another donor is likely to step in. If you donate blood, you simply replace the donation: the other donor who would have donated in your place, will not be recruited. Your donation will not be an extra donation. Or in other words: blood donation has low additivity and blood supply is inelastic. The case for plasma might be different, as there is a global plasma undersupply. A plasma donation will not simply be a replacement of someone else's donation. But I don't know how many QALY's you save by donating plasma.
1: But how do you know when there is an acute blood shortage, when it is time for you to donate with high impact? You only know it when a blood bank actively communicates about it in order to increase donations, but then other potential donors will also be informed and become motivated to donate blood. Cfr blood shortage in New York after 9/11, donors were recruited by mass advertisement, and quickly there was an oversupply of blood. It is like on the stock market: if you don't buy a share, someone else will, and it is difficult to know the good time... (read more)