You are Dr Livia Conti, head of the health technology assessment unit at San Gerardo, a public teaching hospital in Lombardy. A manufacturer wants its €22,000 heart device adopted, and the rest of the region tends to follow your unit's verdicts. Before it models anything, your committee must decide which reading of the trial's benefit it will use. The regional deadline is thirty days away.
The situation
On 6 May 2026 CardioNova's dossier reached your office: 240 pages asking for the adoption of VentriGuard, an implantable monitoring-and-therapy device for advanced heart failure, at €22,000 per patient. Your unit reports to the medical directorate, not to cardiology, precisely so that its verdicts answer to evidence and budget. Its habit is to set the bar before it knows whether the device clears it.
The whole of the evidence is one trial: 480 patients, twelve months of follow-up, funded by the manufacturer. It claims a gain of 0.60 quality-adjusted life years (QALYs) per patient and 22% fewer readmissions. The confidence intervals are wide: the true benefit could be half that, 0.30, and the centre of the interval sits around 0.47. The extra cost over standard care is about €18,000 per patient. The commonly cited willingness-to-pay band runs from €25,000 to €40,000 per QALY.
San Gerardo would have about 150 eligible patients a year; on the same criteria Lombardy would have about 1,400, and at full price the regional bill would pass €30 million a year. The hospital's uncommitted device allowance this year is €2.4 million. The cardiology chief wants approval this quarter with no restrictions, and has warned that without a ruling he will take the matter to the medical board and the press.