Health Ahead: Sequential Comparative-Effectiveness Studies Toward Automated, Universally Deployable Preventive Health Screening
Health Ahead: Sequential Comparative-Effectiveness Studies Toward Automated, Universally Deployable Preventive Health Screening
The Health Ahead Comparative Effectiveness Study is a pragmatic, parallel-arm interventional platform that systematically compares successive changes to preventive health screening on the path toward a fully automated screening system deployable in any environment, including the most isolated and resource-limited communities. Each change is isolated as a single variable against current practice, and every comparison is evaluated with a common set of engagement, behavior-change, experience, cost, and longitudinal outcome measures, so results accumulate on a consistent yardstick across the life of the platform.
The first comparison evaluates AI-assisted versus human-only interpretation ("reads") of screening data. Subsequent pre-planned comparisons, added by protocol amendment, evaluate static versus interactive personalized report delivery; mobile community versus fixed laboratory screening; and a hybrid medical-droid-plus-human delivery model versus human-only screening.
Access to comprehensive preventive health screening is profoundly unequal. Geographic, economic, and systemic barriers leave medically underserved populations, including rural, frontier, and isolated communities, with little or no access to screening deep enough to identify disease before symptoms occur. The long-term aim is a fully automated preventive screening system deployable anywhere people live. Reaching it safely requires testing each change to the screening model one variable at a time, against current practice.
The study is therefore a standing sequential comparative-effectiveness platform. Each comparison isolates a single change, holds every other element of the screening experience constant, and is evaluated against the same core outcome set. As one comparison concludes, the next opens by protocol amendment.
Two principles govern every comparison. First, allocation is randomized wherever participant-level random assignment is feasible, non-randomized only where it is not, as in comparisons of screening location. Second, every comparison is registered with non-inferiority as the primary hypothesis: each step toward a more automated, more broadly deployable model must show it does not degrade outcomes relative to current practice by more than a pre-specified margin. Superiority is a pre-specified secondary in a hierarchical, gatekeeping sequence. No numeric margins are fixed at registration; each is set per comparison by a rule anchored to a validated minimal clinically important difference where one exists, and locked before that comparison's data are analyzed. Each non-inferiority comparison reports intention-to-treat and per-protocol analyses, and a conclusion requires agreement between them.
The active comparison and the planned comparison sequence are set out below. Comparison 1 (active): AI-assisted versus human-only interpretation ("reads") of screening data. Arms complete identical screening for their visit type and differ only in whether the acquired data receive an AI-generated read. A physician of record reviews and signs every released interpretation in all arms. The comparison runs in two strata by screening complexity, randomized 1:1 within each. The primary contrast combines the strata; each stratum is also analyzed separately. The primary hypothesis is that AI-assisted interpretation is non-inferior to human-only interpretation on health activation.
Planned comparison - static versus interactive personalized health report delivery. Both arms complete identical screening and differ only in the report received. The interactive report lets participants adjust their own behaviors and diagnostic inputs in real time and see the projected effect on composite scores, estimated biological age, and aging trajectory. Randomized.
Planned comparison - mobile community screening versus fixed laboratory screening in medically underserved populations. Because assignment follows the site at which a participant presents, this comparison is not randomized at the participant level.
Planned comparison - a hybrid delivery model combining medical droids, one registered nurse, and remote physician oversight, versus human-only screening.
Future comparison - fully automated screening performed by medical droids with remote physician review, versus the hybrid model.
Future comparison - reduced-capture screening versus the full screening battery, testing how far the acquired input can be reduced while preserving what the screening delivers.
This study is one of four that compound into one system. The 100-Year Human Aging Study (NCT07563777) supplies the clinical data and validates what it means for health, disease, disability, and death. The Human Observatory Study (NCT07646782) does the same with sociodemographic and environmental data, and receives each model's geographic residuals. The Longevity Metrics AI/ML Development Study (NCT pending approval) builds the models that make automation, prediction, and broad utilization possible. This study moves the screening toward increasing automation and mobility while maintaining quality.
Inclusion Criteria:
Exclusion Criteria:
- Age under 18 years.
info@longevitymetrics.org13035010016