(988) Coincidence of Early Graft Rejection and Replication of Human Herpesvirus 6 in the Donor Heart Associated with a CD38+ Lineage of Negative T Cells

and cost-effective manner.However, well established, "optimal" standardized methodologies to deliver and monitor personalized health education in an equitable manner are still lacking.Methods: This was a 3-month, prospective, cluster randomized trial at ZCTA level to evaluate the engagement performance of optimal and nonoptimal SNIs using Facebook and a "home-brewed" adaptive content tuning algorithm used to deliver personalized health education to clustered ZCTAs in California.Demographic environments were derived from clustered geocoded data engineered with publicly available datasets.The intervention was delivered in pre-and post-optimization phases.The main study outcome was the number of times personalized organ donation educational content was accessed by the intended population measured in clicks per 1,000 impressions (C/I).Results: Four main clusters were uncovered (Figure).Cluster 1, 602 ZCTAs, is predominantly rural, white, and Hispanic, and has a high rate of public coverage Medicare.Cluster 2, 239 ZCTAs, is rural, white, advanced-age, low-employment, and low-females.Cluster 3, 481 ZCTAs, is predominantly urban, Asian and White, and has a high rate of employment, income, educational level, and internet subscription.Cluster 4, 418 ZCTAs, is urban, Hispanic and Black, average-family-size, and highemployment but low-income.The hyper personalization developed in this study enabled equitable content allocation in the optimal intervention arm, resulting in greater C/I level (beta=.2187;P