I'd rather them disregard their bias than disregard their patients. If someone insists they don't drink when challenged with the BAC data (especially with corroboration like in this case), do a test to determine if they're lying or not. Don't just assume they're lying.
But doctors often are arrogant and unwilling to admit error, even to other doctors. The best you can sometimes do in those situations is to get bossy and try to take control of the situation (e.g. I know you're assuming I'm a lying alcoholic, but I don't drink, so explain to me how I could be intoxicated without drinking...what tests would you need to prove that...), but not everyone has the confidence to do that or to do it only when it's needed.