Although genomic medicine is still a fairly new clinical area, the history of health economics involvement in genomics has a longer history than might be anticipated. Some of the earliest health economics input into genomics was in areas such as neonatal and newborn screening, where health economists contributed to decisions about adding new conditions into newborn screening programmes worldwide. More recently, the first human genome was only sequenced in 2003, costing between US$500 million and US$1 billion. However, by 2008 costs had fallen to a level where so called ‘next-generation sequencing (NGS)’ approaches started to enter clinical research. NGS approaches allow either the whole genome using methods such as whole-genome sequencing (WGS) or parts of it using whole-exome sequencing (WES) or targeted panels to be sequenced in hours with increased sensitivity compared to older less advanced genetic testing approaches. These sequencing approaches provide information that can inform diagnosis, prognosis and clinical management for a variety of disorders, such as rare diseases and some cancers. However, the current costs are still too expensive for some health care providers and the benefit of the tests is largely unknown. Indeed, a lack of evidence on the cost-effectiveness of novel genomic technologies such as WGS is considered a key translational challenge. This is partly because economic evaluations of genomic technologies often fall outside the remit of health technology assessment (HTA) agencies, such as NICE and PBAC. Where they are undertaken (in a HTA context), the methods used for the assessment sometimes differ from those recommended by HTA agencies for cost-effectiveness analysis. This is against a background of uncertainty as to whether the terms precision medicine, personalised medicine or genomic medicine best capture this space in health care.
Some applications of genomic sequencing generate information that may not improve quality of life (as measured using preference-based health-related quality of life [HRQoL] instruments such as the EuroQol-five dimensions questionnaire) or extend life expectancy. One example is the use of WGS and WES to guide diagnosis in autism spectrum disorder. However, genomic sequencing results may influence patient wellbeing via non-clinical routes, generating ‘personal utility’. This is a particular issue for individuals with rare diseases, who often have lengthy diagnostic journeys but few (if any) treatment options available once they receive a diagnosis. This could also be an issue if individuals without known health problems (healthy cohorts) undergo genomic sequencing and find out that they have an elevated risk of a disease, but no preventive action can be taken to manage this risk.
With respect to costs, the costs of undertaking genomic tests are only one component of the cost of the overall genomic testing process. The costs that are incurred beyond those associated with the production of genomic information (so probably beyond the scope of any national tariffs that might be generated) include the costs of bioinformatics analysis, interpretation of results in multidisciplinary team (MDT) meetings and genetic counselling services.
Such issues have raised questions about whether or not genomics is exceptional for health economists – possibly not, but the combined issues perhaps lead to it often requiring additional attention. There is also a consideration of the importance of accounting for the ‘personal’ when evaluating personalized medicine and considers the extent to which extra-welfarist and welfarist approaches to economic evaluation achieve this objective. Extra-welfarist approaches are currently used by many health technology assessment agencies but may not capture all of the outcomes that are important to patients in this context. Extensions to the extra-welfarist approach that might better capture the ‘personal’ are outlined, including multi-criteria decision analysis and the capability approach.
A recent literature review identified only 36 economic evaluations of either WGS or WES, six of which were cost-effectiveness analyses using diagnostic yield as the outcome measure. Only two publications presented cost-utility analyses using quality-adjusted life-years (QALYs) as the measure of health outcomes. HTA agencies generally require data on survival and quality of life when evaluating new healthcare interventions, which, when combined, allow clinical utility to be quantified using QALYs. However, existing studies have primarily quantified the clinical utility of genomic tests in terms of changes in diagnostic yield. Methodological uncertainty among health economists is one potential explanation for the lack of evidence on the health outcomes associated with genomic sequencing. Over the past decade, health economists have repeatedly questioned whether metrics such as the QALY in genomic medicine, which focuses on clinical utility, can fully quantify the outcomes that are important to patients when they undergo genomic testing.
There are high-level discussions in several countries, including the UK, about extending the use of genomic sequencing into newborning screening, so effectively screening everyone at birth for a large range of conditions, far more than those currently being screened for and which there might not be treatments for yet. This is in addition to long term epidemiological and health economic discussions on using newborn screening for conditions such as hereditary hemochromatosis. A further area of uncertainty is the use of genomic sequencing in ‘healthy populations’, including direct to consumer testing (private genetic tests). In a public health care system setting, the UK Department of health is exploring the value of establishing a healthy cohort of volunteer. Furthermore, research studies are assessing the costs and effects of polygenetic risk scores in the context of primary care as an opportunistic ‘health check’ approach, which could incorporate risks for cardiovascular disease, diabetes, different cancers and conditions such as chrohn’s disease etc. Clearly, there are health economic questions to be asked about the downstream costs and consequences of genomic tests in these newborn and ‘healthy’ populations.
In cancer, there are discussions about how to handle the new invention of agnostic cancer drugs (which essentially target the mutation rather than the cancer, so the same drug can treat several cancers). This is an area where assessments are going through HTA agencies who are unsure about the best approaches to adopt to these assessments where drug companies are putting forward a drug for assessment that can potentially treat different cancers with very different disease profiles. These developments require careful consideration from many perspectives, including health economics.
Besides highlighting some of the challenges in assessing the economic impact of genomic medicine and the use of advanced (and less advanced) technologies, the book will propose potential solutions to these key challenges. For example, in terms of data availability, one obstacle to translating genomic sequencing into routine health care has been a lack of large randomised controlled clinical trials data for health economists and others to use to populate cost-effectiveness analyses (CEAs). Arguably, in response, reimbursement decisions have moved towards lower evidentiary standards, with the development of managed access programs that hope to balance the intense pressure for patient access with the need to consider the sustainability objectives of health care systems. Single arm trials are common for assessing clinical utility of precision medicine. By excluding a counterfactual, these trials introduce outcomes uncertainty through their inability to establish causal treatment effects. In this section of the book, we illustrate the application of quasi-experimental methods for evaluating precision medicine in case studies linking real-world big data and single arm trials.
A further potential option here might be provided by ‘big data’ can be used to partially support CEAs in genomics. Advanced genomic sequencing is considered to be a prominent example of "big data" because of the quantity and complexity of data it produces and because it presents an opportunity to use powerful information sources that could reduce clinical and health economic uncertainty at a patient level. The creation of large national sequencing initiatives with sequencing data linked to clinical data (including health outcomes) and resource use data such as hospital episode statistics data and claims data. Large-scale sequencing projects such as the 100,000 Genome Project in the UK and the All of Us Program in the US are collecting an unprecedented amount of genomic, clinical and healthcare resource use data on individuals with cancer or rare diseases, as well as healthy individuals. Some of these large-scale projects are now approaching completion, and national health services are deciding whether WGS and WES should be translated into clinical practice for specific disorders.
Describes many synergies with genomics and complex interventions such as obesity.
Discusses state of the art methodologies in outcomes and costs, especially in the context of big data.
Written by health economists for health economists.
Sarah Wordsworth, Oxford University, UK; Dean Regier, UBC, Canada