Series: 'Medical desertification and care wandering' — Article 4/5
If the clinical evidence is ready and the technology is mature, why don’t predictive prevention approaches scale? The answer is neither medical nor technological. It is economic and institutional. This article explicitly distinguishes three levels of reasoning: [ESTABLISHED] for documented results, [ORDER OF MAGNITUDE] for plausible modelling, [ILLUSTRATION] for stylised scenarios.
Medical desertification is neither an accident nor a marginal anomaly. It is the predictable consequence of a system not designed to orchestrate longitudinal pathways. Recent scientific literature converges: RPM programmes significantly reduce hospitalisations. The problem is no longer demonstrating that predictive prevention works — it is understanding why it does not scale despite accumulated evidence.
A well-designed predictive prevention programme creates considerable systemic value — avoided hospitalisations, extended autonomy, freed physician time. But this value is fragmented across actors that are neither aligned nor economically coordinated. The actor who must invest is not the one capturing most of the gains: this is the heart of the economic lock.
Under fee-for-service, a hospital launching an RPM programme for 2,000 heart failure patients could avoid 240 hospitalisations per year (~€1.2M in collective savings), but those 240 avoided stays mechanically become 240 fewer billed stays. The rational decision for the director is not to invest. This is not a failure of will — it is a mechanical property of the funding system.
For 12 million chronic patients, the discriminating variable is marginal cost per patient per year. Highly human-intensive models (€650/patient/year) are effective but structurally unscalable. Automated infrastructure models (€100–150/patient/year) are the only ones sustainable at national scale. Even a profitable programme faces a wall: the working capital requirement. For 18 to 24 months, cash flow is negative. Only public authorities can absorb this valley.
The amount needed to initiate predictive prevention at national scale represents approximately 0.06 % of ONDAM. This is not a money problem — it is a decision problem. Article originally published on LinkedIn (in French).