In randomized controlled trials, 35,000 women have now taken aspirin. There is evidence that there's reduction in preterm births and in perinatal mortality. It is statistically significant but a questionably clinically significant finding. One of the possibilities is that there are subsets of patients. The doses were not the same in all the studies. The hypothesis that guided this study was prostocylcin thromboxane imbalance, and the idea was that if you give a very low dose of aspirin you'll only affect the production of prostaglandin in some of the studies but you lose a lot of other aspirin's effect. So it could be that asprin is effective for reasons that we didn't know of. There's now an attempt being done at Oxford to get as much of the primary data as they can for these 35,000 people so they can identify subsets of people for which there was value. From a clinician's perspective it's certainly the low risk patients. This is not a significant enough association to justify treating probably 250 women.