Client preferences for HIV Care Coordination Program features in New York City: latent class analysis of a discrete choice experiment

INTRODUCTION: The PROMISE study, launched in 2018, evaluates the implementation of revisions to the HIV Care Coordination Program (CCP) designed to minimize persistent disparities in HIV outcomes among high‐need persons living with HIV in New York City. We conducted a discrete choice experiment (DCE) assessing the preferences of CCP clients to inform improvements to the program's design. METHODS: Clients chose between two hypothetical CCP options that varied across four program attributes: help with antiretroviral therapy (ART) adherence (directly observed therapy [DOT] vs. remind via phone/text vs. adherence assessment), help with primary care appointments (remind and accompany vs. remind and transport vs. remind only), help with issues other than primary care (coverage and benefits vs. housing and food vs. mental health vs. specialty medical care) and visit location (meet at home vs. via phone/video vs. program visit 30 or 60 minutes away). The latent class analysis identified different preference patterns. A choice simulation was performed to model client preferences for hypothetical CCPs as a whole. RESULTS: One hundred and eighty‐one CCP clients from six sites implementing the revised CCP completed the DCE January 2020–March 2021. Most clients had stable housing (68.5%), reported no problem substance use in the last 3 months (72.4%) and achieved viral suppression (78.5) with only 26.5% receiving DOT within a CCP. 77.3% of responses were obtained before the COVID‐19 pandemic. Preferences clustered into three groups. Visit location and ART adherence support were the most important attributes. Group 1 (40%) endorsed telehealth for visit location; telehealth for ART adherence support; and help with securing housing/food; Group 2 (37%) endorsed telehealth for visit location; telehealth for ART adherence support; and staff reminding/arranging appointment transportation; Group 3 (23%) endorsed staff meeting clients at program location and staff working with clients for medication adherence. In the choice simulation, Basic and Medium hypothetical CCPs were endorsed more than Intensive CCPs. CONCLUSIONS: This DCE revealed a strong preference for telehealth and a relatively low preference for intensive services, such as DOT and home visits; preferences were heterogeneous. The findings support differentiated care and remote service delivery options in the NYC CCP, and can inform improvements to CCP design.