![[ Visit Client Website ]](images/banner.gif)
Background: The purpose of this study was to evaluate the reliability and factorial validity of a scale designed to measure trans- and gender nonconforming people's perceptions of discrimination when using healthcare services, and to identify potential subscales. The 52-item Trans Perceptions of Discrimination in Health Care (TransPDHC) was constructed, predicating discrimination on four domains based partially on findings from the National Transgender Discrimination Survey Report on Health and Health Care (Grant, Mottet, & Tanis, 2010). The hypothesized domains were harassment/violence; refusal of services; lack of provider knowledge; and visual cues in the healthcare setting. Review of psychometric literature revealed no validated instrument exclusively assessing trans people's perceptions of discrimination in healthcare settings.
Methods: Data were collected using a cross-sectional survey design. Convenience and snowball sampling were used to recruit participants through known LGBTQ advocates, agencies serving trans populations, and social media. The final sample was 97 people aged 18+ who identified in some way as trans. 52.6% of respondents identified as transgender; ~34% as transmen and ~26% as transwomen; 20% as women and ~26% as men; ~15% as gender nonconforming, ~18% as genderqueer, and ~6% as gender variant (many respondents [61%] reported >/= 2 gender descriptors).
Approximately 79% of participants were white, with non-negligible multiracial and Latino/a subgroups. Their mean age was 35.62 years (SD = 12.26). Median annual income was $24,000. Most respondents (~88%) had health insurance, most frequently employer-based (~26%) or public (~23%).
Due to the lack of previous scale development on this topic and a small sample size, TransPDHC items were assessed using exploratory factor analysis employing maximum likelihood extraction and Promax rotation. Thirteen factors were originally extracted, however, rotation of these failed to converge onto interpretable factor loadings. Based on examination of scree plots, eigenvalue patterns, Field's (2005) recommendation of retaining only factors with >/= 4 pattern coefficients > .6, and assessment of item communalities, subsequent EFAs constrained to three, four, and five factors were compared. The 4-factor result was most parsimonious, explaining 50.4% of total variance with the cleanest factor loadings. The cutoff for item retention was a pattern coefficient >/= .4.
Results: The 4-factor constrained EFA revealed a reduced 34-item TransPDHC scale with four preliminary subscales. The data suggest a factor structure slightly different than originally anticipated. Items loading onto Factor 1 were discernible as provider stereotyping behaviors, such as providers making assumptions about a trans person's sexual behaviors, orientation, or values. Factor-2 content contained items predominately concerning harassment/bullying by providers or staff. Factor-3 items mostly concerned competency (e.g., awareness of trans-relevant resources), and Factor 4 retained the items concerning visual communication of trans inclusivity in healthcare settings. The scale demonstrated significant convergent validity with single-item indicators related to postponing healthcare due to trans identity.
Implications: Further analyses of larger samples are needed to clarify the TransPDHC's factor- and higher-order structure. However, preliminary findings support its emergent domains as potentially capturing trans-based discrimination in healthcare. The measure has utility for aiding social workers in increased research and advocacy towards more trans-affirming/inclusive health services.