[matteotom]

I’d read that as saying the data are insufficient to prove either way. Likely because there are so many fewer deaths and only 10% of the children used booster seats, so the error bars are too big.