MCDA swing weighting and discrete choice experiments for elicitation of patient benefit‐risk preferences: a critical assessment