Series: 'Medical desertification and care wandering' — Article 5, Part 2/5
This article addresses the central question: how to unlock a seed budget to move from proof-of-concept to accelerated deployment of a predictive preventive medicine infrastructure? The system is not poor — it is misaligned. French healthcare spending is high, but concentrated downstream: complications, hospitalisations, dependency.
Funding predictive prevention means shifting existing expenditure from downstream to upstream, not creating new spending. Predictive prevention at scale is an infrastructure characterised by upfront investment, recurring costs, returns to scale, and governance requirements.
Four building blocks compose the infrastructure: precedence (IT interoperability, algorithms, protocols, compliance, training), recurring OPEX (where marginal cost is the discriminating variable — €80–150/patient/year with AI versus €650 in a “human-only” model), structural incentives (as long as prevention remains micro-economically penalising under fee-for-service, nothing changes), and measurement with accountability. The seed budget for 50 territories and 500,000 patients over 3 years amounts to approximately €160M, i.e. 0.02 % of ONDAM — the equivalent of 32 latest-generation MRI machines.
The credible strategy combines five instruments in sequence: FIR and Article 51 for bootstrapping and field learning, a multi-year ONDAM allocation as the central lever for sustainability, a disciplined PPP with anti-capture clauses for transitional working-capital coverage, SIB/OIS for 2–3 “hard proof” pilots locking in acceptability, and incentive reform as a non-negotiable condition — extracting longitudinal coordination from the acute-care funding framework and establishing standardised remuneration for primary-care coordination.
The keystone of the system is routine medico-economic validation through territorial digital twins, which transform funding from a cheque based on a promise into a commitment conditioned by proof in production. Article originally published on LinkedIn (in French).