Wednesday, March 7, 2007

Iowa relaxing blood donation ban?

The largest blood donor service in Iowa is recommending that a ban on gay-men donating blood be re-examined. As we all know, the ban on men who have sex with men from donating blood dates back to the 1970's when the height of the HIV scare was just beginning. However, since that time our ability to test for the HIV virus has improved dramatically.

Each year during this week, the fraternities and sororities of Iowa State University hold a contest to see who can collect the most blood donations. However, one fraternity (Delta Lambda Phi - an LGBT fraternity) complained that they would be unable to participate as a result of the ban. So the entire greek community has decided to not hold the competition. They estimate this will amount to a loss of 500 units of blood.

While as a physician-to-be I cannot support boycotting blood donation, I do applaud the campus community for supporting their LGBT brothers and sisters who are being unfairly discriminated against. Maybe someone with authority will finally listen and re-think this unnecessary and discriminatory policy.


Anonymous said...

Isn't the fact that HIV testing on blood products is as good as it is (1:1,000,000 false negative rate I've heard) because high risk populations are excluded? It seems that increasing the prevalence of HIV in the donating population will also increase the false negative rate of HIV testing on the donated units.

Garrett said...

That's a pretty ignorant assumption to say that including homosexuals would significantly increase the prevalence in the donating population. In 1987, maybe. But in 2007, the prevalence of HIV in the heterosexual and homosexual populations aren't SO disparate that the above argument holds much weight.

Blood transfusions always carry just above minimal risk, as does not having blood available in proper serotypes, etc. I'd be hard pressed to believe that, with our current quality of screening, in addition to the benefit of an additional 3-5% of the population being eligible to donate blood, that the sum benefit didn't drastically outweigh the sum cost.

I mean, what would be worse, one additional case of HIV, or one additional risk of an immediate death due to a lack of compatible blood? I'd imagine the risk of the latter is much greater than the risk of the former.

And HIV is a chronic infection, not a death sentence. It sucks, sure, but having to choose between dying in an ER, and dying of AIDS 20 years later with ever-improving care, I would imagine that living with a chronic illness at age 50 beats dying of an acute one at age 30.

Philip said...

"Men who have sex with men (MSM) account for nearly half of all newly reported HIV/AIDS diagnoses, and young men are at highest risk. A 2005 survey of MSM in several large U.S. cities (CDC, 2005) found that 1 in 4 of those surveyed was HIV-positive and nearly half of them were unaware of their HIV status. Prevalence of HIV/AIDS is higher among MSM from racial and ethnic minorities than among white MSM."


1% prevalence is considered a high risk population worthy of universal screening (USPSTF). 25% vs. less than 1% in the general population seems like a pretty big difference to me. If you have a different source to prove that the prevalence in the two populations is more equivalent I'd like to know. There may well be subsets of the MSM population with a very low HIV prevalence that would not affect the prevalence of the donating population. But the blood donation screening process is not known for being very good at identifying subsets. If you've ever been in certain African countries you can't donate blood for life regardless of your length of stay (I think malaria risk, not HIV, is the issue here). But the point is that you could ask more questions to tease out the safe subsets from the risker ones.

Blood products are wasted left and right in this country. I think we could have a bigger impact on the blood supply by cutting down on waste rather than increasing the supply by 3-5%. Of course you'll obviously have the biggest impact by doing both at the same time. Sorry I don't have a source for that right now, that wasn't the point of my first post, but proper use of blood products is a huge issue all on it's own.

Also sorry for the last post being anonymous, didn't have a blogger ID and didn't realize I could use the "other" option till I was proofing this one.

Garrett said...

Assuming that 5% of the population was gay males (which is, of course, a pretty significant overestimate) and assuming that 25% of MSMs have HIV, then we're faced with this math:

(0.95)(0.01) + (0.05)(0.25) = 0.0149.

So, we'd suddenly be increasing our total risk pool from 0.01 to 0.015.

In our old population, we'd expect 20,000 cases of HIV per 2 million people. We would miss two of them in screening.

In our new population, we'd expect 29,800 cases of HIV per 2 million people. We would miss three of them in screening. And that's assuming that A) none of them know they are infected, and B) they are equally predisposed to donate blood as the rest of the population, which I would argue are pretty faulty assumptions.

So, by adding MSM to our pool of potential blood donors, we'd miss one extra case per 2 million units of blood.

And while I certainly agree that we're wasteful of our blood in equilibrium circumstances, circumstances do exist when shortages form. I've already reviewed the advantages of an expanded pool of possible donors. An additional 1 case of HIV per 2 million units of blood (which is frankly a daft overestimate of the risk as it is, based on the extremely "liberal" assumptions we made above) seems to more than offset the value of having the increased donor pool.

I agree that it sounds like adding a group with a supposed prevalence of 25% HIV into the donor pool would muck things up. But throw the numbers through the screening tool, and the extremely modest increase in overall prevalence leads to an extremely low increase of risk from a transfusion, with less modest benefits to be gained. I wish I could more easily quantify those gains, but I don't have the software or programming skills to run those models. But I think it's safe to say that, in a Katrina-like situation, where blood products might be in short supply, the ratio of blood products donated to the lives saved immediately would far exceed 1 in 2 million. More like 1 in 200 for a unit of O+ MSM blood.

Overall, gains >> costs. That simple. If our screens weren't so excellent, you'd have a great point. But they are, and the addition of MSM into the donor pool, as I originally stated, does not significantly increase the prevalence in the donating population.