Predictors of HIV Screening among Florida Students: An Application of Multilevel Logistic Regression

Abstract

Objectives: To explore the predictors of Human Immune Virus (HIV) screening among Florida high school students using a multilevel logical regression.

Methods: We used data of 5,394 high school students in grade level 8 to 12 from the Youth Risk Behavior Survey (YRBS) 2019. In addition to univariate analysis, multilevel logistic regression models were used to examine the association of selected predictors and HIV screening.

Results: Students aged 17 years and above were more likely to test for HIV than students 16 years or less (OR=1.928, 95% CI 1.413 – 2.630). Compared to students who did not have A’s or B’s, students who had A’s and B’s were 47% less likely to screen for HIV (OR =0.530, 95% CI 0.381 - 0.736). Discussing HIV with adults and parents (OR= 1.417, 95% CI 1.029 -1.952) and healthcare experts (OR=3.923, 95% CI 2.838 -5.423) was associated with increased odds of screening.

Conclusions: Multilevel approaches to examine HIV screening predictors have tremendous potential to provide more insight into class level factors that influence public health programs. This is important especially in situations where there may not be enough state or departmental funding to implement a comprehensive screening program.