Predictive Models to Identify Factors Associated with a High One-Time Sexual Partnership Rate Among U.S. Men Who Have...

Area of Science:

• Epidemiology
• Network Analysis
• Public Health

Background:

• One-time sexual partnerships contribute to densely connected sexual networks.
• A right-skewed distribution of partnerships exists among men who have sex with men (MSM), with a few MSM having numerous partners.
• Understanding predictors of high partnership counts can inform sexually transmitted infection (STI) prevention strategies.

Purpose of the Study:

• To identify covariates that predict high rates of one-time sexual partnerships among US MSM.
• To evaluate the predictive performance of different statistical models for identifying individuals with high partnership counts.

Main Methods:

• Negative binomial regression models were used to estimate the count of one-time partnerships.
• Penalized logistic regression models were employed to identify predictors of high partnership counts at various thresholds.
• Data were collected from a web-based egocentric network survey of US MSM.

Main Results:

• Significant predictors of high one-time partnership rates included the number of main and casual partners, exchange sex, using mobile apps or the internet to meet men, and non-injection drug use.
• Having one main partner was associated with a lower rate of one-time partnerships compared to having none, while more than one main partner or any casual partners increased the rate.
• Logistic models with a threshold of 6 partners per year demonstrated the best predictive performance for identifying high partnership counts.

Conclusions:

• A consistent set of covariates predicts high partnership rates among MSM.
• These findings can serve as indicators for targeted HIV/STI prevention interventions, including partner-based services.
• The identified predictors can aid in the parameterization of STI transmission models for improved public health strategies.