Using Natural Language Processing on Expert Panel Discussions to Gain Insights for Recruitment, Retention and Intervention Adherence for Online Social Support Interventions on a Stage II-III Clinical Trial Among Hispanic and African American Dementia Caregivers.

We applied natural language processing (NLP) to a corpus extracted from 4 hours of expert panel discussion transcripts to determine the sustainability of a Stage II-III clinical trial of online social support interventions for Hispanic and African American dementia caregivers. Prominent topics included Technology/hard to reach populations, Training younger populations, Building trust, Privacy and security issues, Simplification of screening questions and recruitment procedures, Understanding participants' needs, Planning strategies and logistics, Potential recruitment places, Adjusting intervention size downwards to engage elderly participants, Targeting different generations, Internet-based interventions by age range, and Providing step-by-step instructions and an overview of the entire research process during recruitment. The application of NLP to qualitative data on a dementia caregiving clinical trial provides useful insights for recruitment, retention, and adherence to guidelines for such interventions serving Hispanic and African American dementia caregivers.