The importance of rigorous methods in observational comparative effectiveness studies of rare diseases: comment on the article by Ruperto et al