Propensity score methods and regression adjustment for analysis of nonrandomized studies with health‐related quality of life outcomes