The first author identified relevant abstracts through review of citations obtained from this
search strategy. Each abstract was assigned to a team member for independent review to
confirm relevance to the research question. We developed a standardized form to facilitate the
review of abstracts and articles to determine relevance to the study question, document study
characteristics, extract data, and assess the quality of evidence. Full articles were obtained for
those abstracts appearing eligible and in cases in which determinations could not be made from
the abstracts alone. Three team members (TLF, DLB, and KAC) reviewed articles
independently and then compared findings; each article was reviewed by at least two reviewers,
and differences were adjudicated by team consensus.