"Brimful of STARLITE": Toward standards for reporting literature searches

Abstract

Context: Systematic reviews of qualitative research studies extend understanding of health care beyond effectiveness to acceptability and user views. Objective: The paper surveys reports of qualitative systematic reviews and, by characterizing techniques used to identify articles for inclusion, proposes standards for reporting of literature searches. Data Sources and Study Selection: A search of MEDLINE was performed for qualitative systematic reviews published from 1988 to December 2004, supported by searches of CINAHL, Web of Knowledge (including the Science and Social Sciences Citation Index), and the Cochrane Methodology Register, and Internet searches using the Copernic Agent Professional meta-search agent. Studies were included if they used techniques of qualitative synthesis in reviewing research studies in health care. Narrative reviews were excluded. Data Extraction: Authors, year of publication, sampling strategy, databases, keywords, and other approaches used were extracted. Data Synthesis: Sixty-four studies were identified, and forty-three met inclusion criteria for this review. A summary of searching methods was produced and used to construct the STARLITE mnemonic (sampling strategy, type of study, approaches, range of years, limits, inclusion and exclusions, terms used, electronic sources). Conclusions: Considerable variation exists in search methods for qualitative systematic reviews. While diversity in methods is appropriate during the development of review methodology, major concerns remain about the absence of an accepted standard and the consequent poor quality of reporting.