Blood Donation: (Generally) Not That Effective on the Margin

[Note: This origi­nally came out of a post on the Face­book group, which was then re­versed in light of Gre­gory Lewis’s ex­pert com­ments. The sub­stance of this ar­ti­cle is his; any er­rors are mine.]

If no one donated blood, a lot of trauma/​hem­or­rhage vic­tims would die, and the world would be a lot worse off. The av­er­age unit of blood donated goes pretty far, in terms of the ex­pected value of the good it does. How­ever, when con­sid­er­ing whether or not to donate, we need to eval­u­ate the coun­ter­fac­tual differ­ence our ac­tions make. That is, rather than look­ing at the av­er­age dona­tion, the rele­vant mea­sure­ment is of the marginal dona­tion. This brings up the fol­low­ing con­sid­er­a­tions:

1) As it stands, a sub­stan­tial num­ber of peo­ple already donate reg­u­larly, and will con­tinue to do so whether or not the (com­par­a­tively tiny) EA com­mu­nity does too.

2) Truly life-or-death situ­a­tions are a minor­ity of trans­fu­sions, and these are pretty much already cov­ered by the ex­ist­ing sup­ply. In fact, hos­pi­tals al­most always keep an emer­gency re­serve of O- speci­fi­cally for these cases, so it’s very rare that some­one di­rectly dies for lack of com­pat­i­ble blood. Be­cause a large num­ber of trans­fu­sions/​dona­tions hap­pen each day, and blood product can of­ten be trans­ported to differ­ent hos­pi­tals to meet lo­cal short­ages, pro­jected sup­plies are rel­a­tively easy to fore­cast within a given mar­gin of er­ror, so it is pos­si­ble for hos­pi­tals to main­tain this emer­gency sup­ply to han­dle ur­gent cases.

Thus, the effect of an ad­di­tional dona­tion to the ex­ist­ing sup­ply is to help cases where the pa­tient wouldn’t be di­rectly saved from death, but a trans­fu­sion would im­prove the qual­ity of their re­cov­ery. Nailing down ex­actly how many QALYs this typ­i­cally adds is very difficult to track, and prob­a­bly hasn’t been done in a rigor­ous way. How­ever, there is rea­son to be­lieve this num­ber is not that high.

In the UK, a unit of red blood cells (RBCs) costs about 120 pounds. While fi­nan­cial in­cen­tives don’t trans­late seam­lessly into ex­tra dona­tions, this is roughly this price at which more sup­ply can be ob­tained, so it roughly re­flects the med­i­cal field’s im­pres­sion of how valuable it would be to do more out­reach per unit. Fur­ther­more, the typ­i­cal cut­off for whether to fund treat­ment is ~20,000 pounds/​QALY, which is much less effi­cient than the ~130 pounds/​QALY one can get by donat­ing to the AMF. (For more de­tail on these num­bers, see this guessti­mate and this ex­pla­na­tion of it.)

Thus, for blood dona­tion to be any­where near as effec­tive as the AMF (in terms of pay­ing 120 pounds/​unit for more product), the med­i­cal field would have to be un­der­valu­ing the effec­tive­ness of blood dona­tions by 2 or­ders of mag­ni­tude. De­spite the lack of rigor­ous calcu­la­tions done in the liter­a­ture, a col­lec­tive mis­calcu­la­tion of this mag­ni­tude seems im­plau­si­ble given the feed­back mechanisms which ex­ist in medicine, not to men­tion the tacit knowl­edge hema­tol­o­gists have de­vel­oped from mak­ing these trade­offs.

The role of effec­tive al­tru­ism is to look for, and seize upon, moral op­por­tu­ni­ties that have been un­fairly passed over by so­ciety at large. GiveWell-recom­mended char­i­ties, for in­stance, may some­times get pos­i­tive com­ments from economists, but re­ceive in­suffi­cient fund­ing to fully ex­ploit the eth­i­cal gold mine that is their cause area. In the case of blood dona­tions, the med­i­cal field gen­er­ally has ways of spot­ting and filling in the cheap and ob­vi­ous ways to save more lives, so our time is bet­ter spent on causes that aren’t be­ing watched over as care­fully.

That said, there are oc­ca­sional cases where emer­gency sup­plies dwin­dle. When this hap­pens, spe­cific ap­peals are made, and in these cases it prob­a­bly is effec­tive to lend some helping hemoglobin. Less cru­cially, reg­u­lar donors of­ten drop out on holi­days and dur­ing the win­ter (due to colds/​flu), so if one is in­clined to donate, those are the best times to do so.