Maybe ask him when was the right time for doctors to start having an opinion about whether smoking is unhealthy?

Not sure if this answers, but the book Superforecasting explains, among other things, that probabilistic thinkers tend to make better forecasts.

I think one reason doctors want such highly credible evidence is that they're defending themself against attacks by snake oil salesmen. Historically, there have been many attempts to sell "treatments" that have absolutely no benefit at all, and for an individual doctor, it is very hard to tell whether they're being sold something that actually works or not. 

Another related reason for wanting highly credible evidence is that doctors tend to be conservative, out of a desire to not prescribe things that might harm their patient (they don't necessarily believe harm coming from a prescription they make is the same sort of thing as harm coming from a disease they didn't stop).  

Now I don't know why he actually reasons the way he does, but these are some possibilities. Perhaps you should make more efforts to understand his reasoning before coming up with a strategy to convince him.