Obtaining Population-Based Estimates of the Timing of Antiretroviral Therapy Initiation Using HIV Surveillance Data: Change-point Model Development, Validation, and Dissemination

Statement of Work for Dr. Hongbin Zhang (MPI for NIAID R21AI147933) Background: The grant has the title “Obtaining population-based estimates of the timing of antiretroviral therapy initiation using HIV surveillance data: Change-point model development, validation, and dissemination.” It was initially awarded for the period 02/01/2020 to 01/31/2022. During that time, we have encountered COVID-related delays and staffing challenges replacing our postdoctoral fellow. Therefore, we requested No Cost Extension (NCE), and now the project is in the NCE period to January 31, 2023. The previous PI, Dr. Zhang, accepted a faculty position at the University of Kentucky (UKY) in Aug 2022. To proceed and complete the project, the proposal is now converted to a two Principal Investigators (MPIs) leadership structure with Dr. Denis Nash (co- investigator in the grant) as the contact PI at CUNY and Dr. Hongbin Zhang in the UKY. Dr. Zhang is a trained statistician whose research has focused on modeling complex longitudinal and cohort data. His research and experience in joint modeling for longitudinal and survival analysis for data complexities had prepared him well to propose the original application and obtain the grant as PI. With the leadership from Dr. Zhang, prior to the single PI to MPI conversion, we have completed the majority of work for aim 1, i.e., model development. In this current NCE period, i.e., from Sept 1, 2022 to Jan 31, 2023, Dr. Zhang will • Development of statistical models (aim 1). Specifically, he will explore a more computationally efficient method, namely Stochastic approximation Expectation and Maximization (SAEM) method for the random change-point model we proposed to estimate the antiretroviral therapy initiation time for people living with HIV. • Jointly with Dr. Nash, Dr. Zhang will make an executable plan for model validation (e.g., with ERC- CFAR’s data) and software development • Jointly with Dr. Nash, prepare the hiring for two RAs (one at CUNY and one at UKY).