Patterns of Health Care Barriers Among LGBTQ Adults: A Latent Class Approach

Background: Quality health care is critical to address LGBTQ health disparities. However, LGBTQ care-seekers consistently face significant barriers to care. The COVID-19 pandemic may have only exacerbated barriers, widening existing disparities. Latent class analysis (LCA) was used to explore patterns of health care barriers among LGBTQ adults since the start of the COVID-19 pandemic (March 2020). Multiple barriers are examined simultaneously, aligning with LGBTQ care-seekers’ likely experiences. Investigating demographic and psychosocial predictors of class membership can identify priority populations for targeted intervention.

Methods: We analyzed data from the Wellness with Pride study. Six indicators representing health care barriers were included as latent class indicators. LCA was performed in MPlus to identify latent subgroups with varying combined levels of health care barriers. A 3-step approach to modeling was utilized so that the measurement model remained fixed when completing the latent class regression.

Results: The final sample included 719 respondents. A six-class model fit best: Class 1’s membership (n= 239; 33.2%) experienced no barriers; Class 2’s membership (n= 72; 10%) experienced fear-driven barriers, i.e., high rates of being “afraid” to seek care with minimal additional barriers; Class 3’s membership (n= 100; 13.9%) experienced multiple barriers and high fear, i.e., moderate rates of cost- and provider-driven barriers, and high rates of deciding care could wait and being afraid to seek care; Class 4’s membership (n= 67; 9.3%) experienced individual-driven barriers, i.e., highest rates of deciding care could wait with no additional barriers; Class 5’s membership (n= 178; 24.8%) experienced multiple barriers and low fear, i.e., moderate cost- and individual-driven barriers, highest rates of provider cancellation, but low rates of being afraid to seek care, and; Class 6’s membership (n=63; 8.8%) experienced cost-driven barriers, i.e., highest rates of difficulties affording care, with few additional barriers. Key regression findings are reported with Class 1 (no barriers) as a reference. Members of Class 2 (fear-driven barriers) were more likely poverty-impacted and to have poor physical health. Members of Class 3 (multiple barriers and high fear) were more likely queer (than lesbian/gay (LG)), women, and to have poor physical and mental health. Members of Class 4 (individual decision barriers) were more likely women, while members of Class 5 (multiple barriers and low fear) were more likely sexually diverse. Members of Class 6 (cost-driven barriers) were more likely poverty impacted and women, and less likely to have health insurance or regular providers.

Conclusions/Implications: Results may inform leveraging resources toward priority populations. Members of Class 6 only endorsed cost as a barrier. This barrier may be mitigated through efforts assisting group members—who were likely poverty-impacted and without insurance or regular providers—through case management and health care navigation support. Alleviating barriers experienced by queer women must be considered. Queer women had higher odds of multiple barriers and high fear in care-seeking, which was associated with poor health and mental health. Though uncertainty remains if fear stemmed from potential COVID-19 exposures or anticipated discrimination, research shows that digital health services may address both factors, promoting increased access to care.