Cost-effectiveness analysis of genomic tests: what are the methods challenges?

The development of genomic tests has been rapid; however, the link between the identification of gene-mutations and long-term patient health is still developing. [1, 2] As a result, the clinical value of genomic tests is not always apparent, and very few tests have demonstrated cost-effectiveness. Cost-effectiveness analysis is used to inform resource allocation by assessing the health benefits and cost implications of new interventions. This approach is well established for considering investment in new technologies through the National Institute for Health and Care Excellence’s (NICE) technology appraisals, and diagnostics assessment programme. However, the appropriateness of the current NICE approach for genomic tests has been questioned.

The main objective of this report is to determine whether the current principles and methods of cost-effectiveness analysis are appropriate for the assessment of genomic tests. First, a systematic literature review is undertaken to understand the challenges associated with applying the methods of economic evaluation to genomic tests, together with potential solutions previously identified. Second, the defining features, from an economic perspective, of genomic tests are established from the literature and used to define the key challenges associated with economic evaluation in genomic testing. Finally, the methods challenges associated with these defining features are identified. There are many practical issues remaining, however. Most are similar to problems faced in applying economic evaluation to other health technologies. We have discussed some practical issues commonly considered in the literature, such as the dearth of relevant evidence, but other general practical issues in economic evaluation have not been discussed in this report when they are covered in current methods guides.

Our review highlights two main differences between cost-effectiveness analysis of genomic tests and that of other technologies: the evaluation of tests for multiple disorders and the potential for infinite time horizons. In common with other diagnostic technologies, there is also a question of whether assessing health gain alone is sufficient to capture all the benefits of the intervention. Current methods for combining diagnostic evidence with treatment patterns and health consequences are well established. However, this could be difficult to implement with genomic tests given the large number of possible treatment disorders and treatment patterns. This report suggests three alternative approaches to economic evaluation to address this issue: an iterative approach, an aggregate approach or a pragmatic approach. The iterative approach suggests that the use of a genomic test would be considered independently in each disease area of potential use, this is the process used by the NICE Diagnostic Assessment Programme. The aggregate approach suggests that all diseases for which the genomic test could be useful would be assessed in a combined analysis, where the costs and consequences of all diseases would be assessed at once. The pragmatic approach requires some upfront clinical decision guidance to determine which disease areas are most likely to drive the cost-effectiveness of the test, and then to undertake a more qualitative approach for determining the direction of bias for those disease areas not fully included in the analysis.

A further potential challenge of assessing the cost-effectiveness of genomic tests is the potential for an infinite time horizon. This is because it depends on whether the information could and would be shared in the future. Currently there is no national system for sharing genomic information so information sharing ultimately depends on the individual. The principle of economic evaluation is that all future costs and consequences should be considered. Current methods capture costs and consequences over the lifetime of a patient. However, there is a potential for the genomic information collected to have consequences for future generations. This issue is the same for other types of diagnostics, but is not currently considered by the NICE Diagnostic Assessment Programme in terms of formal analysis. This challenge suggests the need to understand the cost-effectiveness of a national system that could store and share genomic information.

As with other technologies, genomic tests are associated with many non-health consequences. This is a broader issue that requires decision makers to determine, firstly, whether non-health consequences should be considered and paid for by the NHS. Subsequently it would be important to determine how these non-health consequences should be traded-off against health and how to take account of non-health benefits in the opportunity costs.

Many of the issues in the genomics literature have been identified for other types of technologies including the addition of non-health benefits such as information value, the lack of relevant evidence, the quickly changing environment and the benefits to non-patient populations. Many of the features of conducting economic evaluation of genomic tests are the same as for more general diagnostic testing. The diagnostic nature of genomic tests implies the same common challenges, such as test error and the difficulty in establishing the added value of a test to the decision maker and patient populations. Solutions to many of these problems are well established, and the assessment of the cost-effectiveness of diagnostic tests is increasingly well practiced.

Many of the sections in this report refer to a lack of relevant evidence. This is invariably the case in economic evaluation and does not preclude the need to follow the methods given the available information. Many clinical questions remain regarding the use of genomic testing and particularly whole genome sequencing. Economic evaluation will not provide additional clinical evidence but provides a framework in which to consider the implications of available clinical information and recommended treatment patterns. Economic evaluations that use model-based analyses also provide a framework for assessing the value of additional research to generate further evidence.

The commonalities and shared challenges with other health technologies suggest that the principles and methods of economic evaluation, in general, are appropriate for genomic tests. In terms of cost-effectiveness analysis, further methods research would be valuable on approaches for trading-off health and non-health consequences of tests, assessing the value of sharing genomic information across generations and for choosing among multiple disorders. This methodological research is needed to understand more fully whether current standard methods of cost-effectiveness analysis could be made more directly relevant to health system resource allocation decisions for genomic based diagnostics.