Table 3. Multivariable Logistic Regression in All Transgender Women^a.

Characteristic	OR (95% CI)	P value
Age group, y
40-54	0.65 (0.17-2.37)	.52
55-69	1 [Reference]	NA
≥70	1.83 (1.02-5.49)	<.001
Race and ethnicity
American Indian or Alaska Native	0.22 (0.01-3.11)	.30
Asian	2.97 (0.28-21.6)	.31
Hispanic	2.23 (0.21-15.9)	.43
Non-Hispanic Black	2.11 (0.45-9.98)	.33
Non-Hispanic White	1 [Reference]	NA
Otherb	0.62 (0.08-4.33)	.62
Annual income, US$c
0-25 000	1 [Reference]	NA
25 000-50 000	0.76 (0.22-2.54)	.74
50 000-75 000	1.05 (0.25-4.35)	>.91
≥75 000	0.67 (0.16-2.65)	.63
Employment status
Employed	1 [Reference]	NA
Unemployed	1.04 (0.25-4.05)	>.92
Homemaker	0	>.92
Student	0	>.93
Retired	0.14 (0.03-1.54)	.71
Highest educational level
No high school	1 [Reference]	NA
High school or GED	1.55 (0.33-7.88)	.60
Some college	1.99 (0.38-11.4)	.42
College	2.55 (1.08-14.60)	.03
Cost barrier to care
Yes	1 [Reference]	NA
No	1.60 (0.36-7.61)	.52
Time since last primary care visitc
Past year	1 [Reference]	NA
Past 1-2 y	0.47 (0.03-4.42)	.54
Past 2-5 y	0.40 (0.01-6.27)	.52
>5 y	0.88 (0.04-11.80)	>.90
Had a clinician recommendation for a PSA test
Yes	12.40 (4.47-37.80)	<.001
No	1 [Reference]	NA
Had a clinician-led discussion of PSA advantages
Yes	7.51 (2.49-24.90)	<.001
No	1 [Reference]	NA
Had a clinician-led discussion of PSA disadvantages
Yes	1.14 (0.42-3.00)	.82
No	1 [Reference]	NA

Abbreviations: GED, General Educational Development; NA, not applicable; OR, odds ratio; PSA, prostate-specific antigen.

^a This multivariable logistic regression model assesses the association of the sociodemographic and access-to-care variables with the odds of recent PSA screening in the cohort of transgender women patients from the matched cohort (n = 255).

^b Not specified in the Behavioral Risk Factor Surveillance System.

^c The upper bound of the range is noninclusive.