Health Disparities
Time-Varying Effects of Health Technology on Racial Disparities: Why When We Measure Matters Chloe R. Bennett* Chloe Bennett Bennett Uniformed Services University of the Health Sciences
Evaluations of health technology interventions increasingly consider effects on health disparities, yet typically rely on summary measures, like hazard ratios, that may obscure how intervention effects on disparities evolve over time. Health technologies do not confer benefits passively—they require users to interpret data, modify behaviors, and access follow-up care to translate information into health improvements. Health technologies present unique evaluation challenges: user engagement evolves as individuals learn from and adapt to devices, and the technologies themselves change rapidly through software updates. Well-established critiques note that collapsing effects over time obscures meaningful variation; such concerns are amplified when both engagement patterns and structural barriers change over follow-up. We used target trial emulation to examine continuous glucose monitor (CGM) effects on White-Black disparities in hospitalization among adults with type 2 diabetes using All of Us Research Program data. The White-Black risk difference among CGM users was -0.10 at 12 months (White lower risk), 0.01 at 24 months (no difference), and 0.11 at 72 months (Black lower risk)—a reversal that a summary hazard ratio of 1.88 (Black lower risk) would entirely mask. Early CGM use may reveal the need for clinical intervention, while sustained benefits depend on determinants that are socially and structurally shaped (e.g., access to healthy foods). While findings require cautious interpretation given wide confidence intervals and small samples of Black CGM users, results illustrate how the equity implications of a technology can shift across follow-up. Substantial investment in health technology as a tool for reducing disparities requires rigorous evidence, yet single-time-point evaluations risk premature conclusions about whether technologies narrow or widen inequities.