The World Health Organization (WHO) and the United States government are having a fascinating debate about the number of H1N1 flu vaccinations that children under age 10 should receive. Both agree that two shots are better than one. The vaccine is scarce, however, so public health officials have to make hard choices about how to prioritize its use.
As NPR reported a few days ago, the WHO favors giving a single dose to each child because of the vaccine shortage:
“The priority is to give…one dose rather than vaccinate half the number of children with two doses,” says Dr. Marie-Paule Kieny, director of WHO’s Initiative for Vaccine Research.
The National Institute of Allergies and Infectious Diseases (NIAID), however, favors giving two doses to each child:
“One of the dangers in jumping ahead and saying you want to stretch out the supply by giving a single dose to these young kids,” [NIAID Director Dr. Anthony] Fauci says, “is that you’ll be under-protecting them.” If that happened, he says, we wouldn’t be saving a dose for each vaccinated child. We’d be wasting a dose.
Intrigued by this difference in opinions, I took a quick look at some of the newest vaccination data that the NIAID reported last week. Researchers examined the immune response of 583 children who received two doses of vaccine and found the following:
To use these data in developing vaccination strategies it is helpful (as I repeatedly tell my students) to consider the marginal benefits and costs associated with each shot:
The two columns on the left are the same as before. The two columns on the right measure the marginal benefits of each shot. For example, the marginal benefit of a second shot for children aged 6 to 35 months is that it increases by 75 percentage points the fraction having a strong immune response. For older children, that marginal benefit is only a 39 percentage point increase.
The figures suggest three things about the difference in opinions between the WHO and the NIAID.
1. If you are willing to give one dose to a child aged 6-35 months, then you really ought to be willing to give that child a second dose. The second dose provides much more incremental benefit (a 75% increase vs. a 25% increase) than the first dose. In short, score one for Dr. Fauci and the NIAID.
2. For children age 3-9 years, a second dose provides much less incremental benefit than the first dose (39% for the second versus 55% for the first). If vaccine is in short supply, you are better off using it to vaccinate many 3-9 year olds once, than half as many twice. In short, score one for Dr. Kieny and the WHO.
3. A possible compromise between the WHO and NIAID positions would be to vaccinate the youngest children twice and older children once.
Needless to say (I hope) this is not intended as any recommendation of what our vaccine policy should be. A full-fledged analysis of vaccination strategies would have to consider many other issues including differences in flu vulnerability (boosting the immune response of vulnerable individuals may yield higher public health benefits), any benefits of weak immune response (which would increase the marginal benefit of the first shot), the benefit of vaccinating older people, and the amount of vaccine available. Not to mention various measurement issues about how the immune response figures were collected (e.g., the effects of the first shot were measured after 21 days, while the effects of the second were measures after 8-10 days).
Instead, my point is simpler: that a little bit of marginal analysis can help shed light on a quite striking disagreement among public health experts.