Skip to main content
. 2018 Oct 12;34(10):879–888. doi: 10.1089/aid.2018.0039

Table 3.

Characteristics Included in the Multivariable Logistic Regression Model of Being in a Cluster, Among 1,027 Persons with Sequences Sampled at Vanderbilt Comprehensive Care Clinic, 2011–2015

Characteristic Total, n (%) In cluster, n (%) Multivariable, OR (95% CI)a
Total, row% 1,027 400 (38.9)  
Sex and risk, n = 966
 Female–WSM 179 (17) 50 (13) Ref
 PWID 42 (4) 12 (3) 1.15 (0.52–2.55)
 Male–HET 126 (12) 44 (11) 1.57 (0.92–2.66)
 Male–MSM 619 (60) 270 (68) 1.79 (1.21–2.65)
Race/ethnicity
 White 471 (46) 195 (49) Ref
 Black/AA 473 (46) 176 (44) 0.71 (0.53–0.97)
 Latino 53 (5) 19 (5) 0.82 (0.44–1.54)
 Other/unknown 30 (3) 10 (3) 0.64 (0.28–1.50)
Age
 ≥30 years 663 (65) 210 (53) Ref
 <30 years 364 (35) 190 (48) 2.48 (1.85–3.31)
Sequence year, median (IQR) 2013 (2011–2014) 2013 (2011–2014) 0.89 (0.80–0.98)
Area of residence
 Nashville metro 469 (46) 184 (46) Ref
 Middle Tennessee 463 (45) 190 (48) 1.05 (0.79–1.40)
 Other 95 (9) 26 (7) 0.49 (0.29–0.84)
CD4 lymphocyte >500, cells/μL 301 (30) 131 (33) 1.40 (1.03–1.90)
HIV RNA viral load, median log10 copies/mL (IQR) 4.5 (3.7–5.0) 4.5 (3.9–5.1) 1.19 (1.03–1.38)
a

All listed variables were retained in the multivariable model: sex/transmission risk, race/ethnicity, age, sequence year, area of residence, CD4 lymphocyte count, and HIV RNA viral load. Bold estimates are statistically significant (p < .05).

AA, African-American; CI, confidence interval; OR, odds ratio; WSM, women who have sex with men.