Figure 1
The flowchart is headed identification of studies via databases and registers. The identification stage lists records identified from databases, N equals 480, and other sources, N equals 43. Before screening, 120 records are removed. These include 120 duplicate records, 0 records marked as ineligible by automation tools, and 0 records removed for other reasons. The screening stage lists 403 records screened and 275 records excluded. Reports sought for retrieval are 128, with 11 reports not retrieved. Reports assessed for eligibility are 117. Reports excluded are 51: opinion-only, n equals 9; wrong population, n equals 14; no relevant exposure or outcome, n equals 12; duplicate or overlapping sample or superseded version, n equals 6; and irretrievable full text, n equals 10. The included stage lists studies or reports included in review, n equals 66.PRISMA flow diagram: identification and selection of studies and literature