Table 3.
Univariate and multivariable binary Logistic Regression Models for predicting PrEP linkage plus adequate (Model 2) and near perfect PrEP adherence (Model 3) at the last post-trial visit where levels were measured among participants at all four participating centers.
| Variables for Predicting Post-Trial PrEP Adherence (n=122) | OR | 95% CI | p value | aOR | 95% CI | p value |
|---|---|---|---|---|---|---|
| Univariate Model | Multivariable Model | |||||
| Model 2: Variables for Predicting Adequate PrEP Adherence at the last post-trial visit | ||||||
| Intervention Arm (i.e. receiving daily text messages for PrEP adherence | 0.897 | 0.435 – 1.851 | 0.769 | |||
| Study Site other than UCSD | 0.768 | 0.304 – 1.937 | 0.576 | |||
| Age (per year) | 1.016 | 0.979 – 1.053 | 0.401 | |||
| Non-Hispanic ethnicity | 1.333 | 0.528 – 3.368 | 0.543 | |||
| Higher Education category | 0.959 | 0.716 – 1.285 | 0.779 | |||
| Higher Income category | 1.059 | 0.861 – 1.304 | 0.587 | |||
| Duration on PrEP trial (per day) | 1.003 | 1.000 – 1.006 | 0.025 | - | - | n.s. |
| Adequate Adherence Week 48 PrEP trial Study visit | 4.063 | 1.019 – 16.20 | 0.047 | - | - | n.s. |
| Drug Abuse Screening Test (DAST10; per score point) | 0.707 | 0.548 – 0.914 | 0.008 | 0.717 | 0.550 – 0.933 | 0.013 |
| Stimulant Substance Use | 0.800 | 0.363 – 1.759 | 0.578 | |||
| Non-stimulant Substance Use (alcohol, marijuana and poppers excluded) | 1.179 | 0.571 – 2.437 | 0.656 | |||
| Popper use | 1.110 | 0.537 – 2.296 | 0.778 | |||
| Alcohol Use Disorders Identification Test (AUDIT; per score point) | 0.974 | 0.909 – 1.043 | 0.447 | |||
| Calculated HIV risk score (CalcR) Month | 1.821 | 0.072 – 45.83 | 0.716 | |||
| Model 3: Variables for Predicting Near Perfect PrEP Adherence at the last post-trial visit# | ||||||
| Intervention Arm (i.e. receiving daily text messages for PrEP adherence | 1.427 | 0.654 – 3.316 | 0.351 | |||
| Study Site other than UCSD | 1.053 | 0.372 – 2.978 | 0.923 | |||
| Age (per year) | 1.039 | 0.998 – 1.082 | 0.060 | - | - | n.s. |
| Hispanic Ethnicity | 0.963 | 0.340 – 2.724 | 0.943 | |||
| Higher Education category | 0.854 | 0.609 – 1.198 | 0.361 | |||
| Higher Income category | 1.133 | 0.888 – 1.446 | 0.316 | |||
| Duration on PrEP trial (per day) | 1.004 | 1.000 – 1.008 | 0.073 | - | - | n.s. |
| Near perfect Adherence Week 48 PrEP trial Study visit | 6.221 | 2.412 – 16.04 | <0.001 | 5.711 | 2.160 – 15.10 | <0.001 |
| Drug Abuse Screening Test (DAST10; per score point) | 0.830 | 0.622 – 1.108 | 0.206 | |||
| Stimulant Substance Use | 0.765 | 0.305 – 1.919 | 0.569 | |||
| Non-stimulant Substance Use (alcohol, marijuana and poppers excluded) | 0.877 | 0.386 – 1.992 | 0.754 | |||
| Popper use | 0.769 | 0.336 – 1.763 | 0.535 | |||
| Alcohol Use Disorders Identification Test (AUDIT; per score point) | 0.976 | 0.898 – 1.061 | 0.569 | |||
| Calculated HIV risk score (CalcR) Month | 21.283 | 0.708 – 639.5 | 0.078 | 98.444 | 1.261 – 7687 | 0.039 |
#
Multivariable model: Chi square 7.780; p=0.255 Hosmer Lemeshow; Forward Wald Binary Logistic Regression
Abbreviation: OR=odds ratio; aOR=adjusted odds ratio