Development and validation of LGBTQIA+ search filters: Report on process and pilot filter for queer women

Abstract

Introduction: A search filter for studies involving lesbian, gay, bisexual, transgender, queer, intersex, asexual, and additional sexual minority and gender identities (LGBTQIA+) populations has been developed and validated; however, the filter contained very small gold standard sets for some populations, and terminology, controlled vocabulary, and database functionality has subsequently evolved. We therefore sought to update and re-test the search filters for these selected subgroups using larger gold standard sets. We report on the development and validation of two versions of a sensitivity maximizing search filter for queer women, including but not limited to lesbians and women who have sex with women (WSW).

Methods: We developed a PubMed search filter for queer women using the relative recall approach and incorporating input from queer women. We tested different search combinations against the gold standard set; combinations were tested until a search with 100 percent sensitivity was identified.

Results: We developed and tested variations of the search, and present two versions of the strategy with 99% and 100% sensitivity. The strategies included additional terms to improve sensitivity and proximity searching to improve recall and precision.

Conclusions: The queer women search filters balance sensitivity and precision to facilitate comprehensive retrieval of studies involving queer women. The filters will require ongoing updates to adapt to evolving language and search platform functionalities. Strengths of the study include the involvement of the population of interest at each stage of the project. Future research will include development and testing of search filters for other LGBTQIA+ subgroups such as bisexual and transgender people.