The evidence is statistical and not direct so it may take years to get something that is strong enough evidence but the suspicion might be there early on. I know nothing about the methods but I assume they aim for at least 3 sigma or a 99.7% chance that he used doping. Still, methods have flaws...