For nonrandomized comparative studies, the Methods Guide suggests taking into account mainly the eligibility criteria, the similarity of the comparative groups in terms of baseline characteristics, methods for measuring exposure variables, and any adjustment for confounders.