Martin Landry: "People will argue about the need for earlier use (possibly) or that with even more data there may be a benefit (possibly - but where are those trials going to come from).

But the effect is modest at best - and so far there is no strong evidence it is there at all."

But it doesn't follow at all that because there is no effect when administered late, the effect, if administered early, is modest as best. In fact, there are many medication that are very effective if administered before the first symptoms appear and are completely useless after. If you get exposed to rabies, you can get a rabies vaccine and have close to no chance of contracting it, whereas if you get a vaccine after you develop symptoms, you're almost certain to die.

Similarly, the flu antivirals are somewhat helpful if taken early but have no effect if taken too late.