Keith Horvath, Ph.D.
Pronouns: he, him, his
Professor of Psychology
College of Sciences
Department of Psychology
San Diego
- khorvath@sdsu.edu
- Phone
- 619-594-3346
- Office Hours
-
By Appointment
- Location
-
ARPC-63
250
6363-250 Alvarado Court
San Diego, CA 92120 - Mail Code
- 4611
- Fax
- Links
- Website
- Accounts
Areas of Expertise
Digital HIV Prevention, Treatment Interventions, Sexual and Gender Minority Health, Adolescent and Young Adult Health, Machine learning, Multimodal Data Collection.
Student Opportunities
Bio
Dr. Keith Horvath is a Professor of Psychology who has been involved with designing and testing digital HIV prevention and treatment interventions for over 20 years. The focus of his interventions has been on communities most impacted by HIV in the US, including sexual and gender minority persons. He currently leads several NIH-funded studies to improve pre-exposure prophylaxis (PrEP) adherence among communities at high risk for HIV and other health outcomes. The first is the Linking Youth to PrEP Services (LYPS) study, which is being conducted within the Adolescent Medicine Trials for HIV Interventions Network (ATN). The purpose of this study is to provide PrEP services in communities where youth (13-24 year olds) live rather than at traditional healthcare clinics, and support youth between PrEP care visits through a mobile app. Another study Dr. Horvath leads is the PrEPStep digital health intervention, the goal of which is to improve PrEP adherence among 18-29 year old sexual minority men in the US through a comprehensive mobile app. We also aim to provide timely information and tools to increase effective use of doxycycline post-exposure prophylaxis (DoxyPEP) to compliment the primary objective of optimizing PrEP adherence.
A more recent line of NIH-funded research is using machine learning algorithms to identify novel predictors of unsuppressed HIV viral load to develop novel intervention approaches to improving HIV treatment outcomes among people with HIV. In this study, the study team will collect multi-model data – including survey, electronic momentary analysis, biometric, electronic health record, and real-time adherence data – from people with HIV over a 12 month period to predict viral load status in the following year after data collection. Machine learning algorithms will be used to identify unique individual, social, and structural-level predictors of viral load status, and also inform the development and revision of knowledge graphs.
In addition to these and other funded research projects, Dr. Horvath is the instructor for a Ph.D.-level course on behavioral medicine interventions and research proposal writing.

