New OHE Research Paper: Uncertainty and Risk in HTA Decision Making

A new OHE Research Paper has just been published on the topic of uncertainty and risk in health technology assessment (HTA) decision making. The report is concerned with uncertainty as it affects the cost effectiveness aspects of HTA.

A new OHE Research Paper has just been published on the topic of uncertainty and risk in health technology assessment (HTA) decision making.

The quality of decision-making in key public sector bodies dealing with resource allocation is a major determinant of their efficiency. One of the most difficult and contentious areas of decision-making is the way that uncertainty is dealt with.

This report is concerned with uncertainty as it affects the cost effectiveness aspects of HTA.

It includes a review of the policy documents governing cost benefit analysis and cost effectiveness analysis in respect of HTA in England and Wales, and the results of interviews with former decision makers at the National Institute for Health and Care Excellence (NICE).

Uncertainty has a number of definitions, and components, which are explored. The first interpretation derives from risk aversion. We explore several factors that contribute to risk aversion and find limited reasons for a NICE Appraisal Committee to be risk averse.

The second view of uncertainty is described in terms of variability in the probabilities and outcomes associated with people receiving each treatment option. This is the main way that a NICE Appraisal Committee considers uncertainty in the context of economic evaluation.

The paper finds that:

1. Members of the NICE Appraisal Committees and Guidelines Committees should take a risk neutral approach to handling uncertainty;
2. Sampling variance may be ignored in the consideration of cost effectiveness except insofar as it helps assess the potential for efficiently collecting more information;
3. The Committee should look for the most plausible ICER. The choice of the relevant ICER should reflect concern about potential biases arising from key modelling assumptions or data sources. The potential for efficiently collecting more information should be explored.

Download the full paper here.

For more information please contact Adrian Towse at OHE.