Answers are grounded in this paper and cite their sources.
Artificial intelligence (AI) now allows health organizations and commercial platforms to sort populations into segments and to direct interventions at each one, yet this capability is usually judged on predictive accuracy or engagement. This narrative review argues that segmentation in the digital health economy is a resource allocation decision made computationally, and that it should be evaluated on economic and distributional criteria. Literature on segmentation, consumerization, reimbursement and equity was purposively selected and interpreted. Economic evaluations of digital health technologies remain scarce, and engagement, the outcome most often reported, is a commercial measure rather than an economic one. Germany’s digital health application programme shows that structured reimbursement is achievable, although prices rest on negotiation because the framework does not specify how benefit should translate into price. Digital health use is patterned by socioeconomic position, so an intervention may be cost-effective in aggregate while widening inequality; whether it does depends on effectiveness, adherence, care pathways, financing and compensatory measures. Distributional cost-effectiveness analysis can assess this risk, but completed applications in digital health were not identified in the literature reviewed. Consumerization can shift device, connectivity and self-management costs to households, and algorithms trained on expenditure can reproduce unequal access. Evidence that noise-based differential privacy distorts small-area measures of inequity is mixed. A major constraint on AI-enabled health transformation is therefore the limited evidence on whether deployed capability delivers value, and for whom. Routine cost-per-outcome reporting, distributional evaluation, economic evaluation of segmentation itself and lifecycle costing are the next steps.
No takes yet. Share an insight, caveat, or question.
Yu et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: