Cost per QALY Threshold

The maximum cost per quality-adjusted life year a health system is willing to pay for a health gain, used as the decision criterion in cost-utility analysis.

Source: Claxton et al. 2015

Two quite different conceptions compete. One treats the threshold as what society is willing to pay for a unit of health, derived from stated or revealed preference evidence, which makes it a statement about value and implies the budget should expand to fund everything below it. The other treats it as the health forgone when resources are displaced to fund something new, which makes it a property of the existing budget and its productivity. The two can differ substantially, and which is intended determines what the resulting decisions actually achieve.

Source: Culyer 2016

Estimating a threshold on the displacement conception means measuring how much health the system currently produces with its last unit of spending, which is not a quantity any dataset records directly. The published attempts work backwards from mortality and morbidity outcomes against expenditure across programme areas, using variation in local budgets that is arguably unrelated to need, and converting the resulting relation into a cost per unit of health. The estimates produced this way have generally come in below the thresholds systems were applying at the time, which is the finding that made the exercise contentious rather than technical.

Source: Claxton et al. 2015

Every adoption at the threshold removes more health than it adds, because the resources displaced were producing more than the new intervention delivers. The system therefore loses health while each individual decision appears justified, and the loss is invisible because the patients who bear it are unidentified and spread across services that quietly do slightly less. The effect compounds with each adoption, so a threshold above the true displacement value produces a steady erosion rather than a single error.

Source: Claxton et al. 2015

Ranking every candidate use of resources and funding downward until the budget is exhausted is the theoretically correct procedure and is impossible in practice, since candidates arrive at different times, the full set is never known, and existing commitments cannot be reopened continuously. A threshold approximates the outcome of that ranking with a single test that can be applied to each candidate as it arrives. The approximation is only as good as the correspondence between the threshold and the displacement actually occurring.

Source: Culyer 2016

That a single figure cannot reflect a system whose marginal productivity varies between services and over time. That published thresholds are frequently historical conventions rather than estimates of anything, and are rarely revised as budgets and productivity change. That applying one uniformly ignores considerations most systems regard as relevant, including severity and equity, which is why frameworks applying a threshold usually attach modifiers to it. And that the displacement it assumes may not occur where budgets are soft.

Source: healtheconomics.wiki