Terry Flynn recently blogged on how treatment tailored to genes will kill economic evaluation. It’s a catchy title that I hope will draw health economists working outside of genetics into a growing debate on the best way to do economic evaluation in genetics and genomics. However, I don’t entirely agree with everything that Terry said and wanted to respond on a few points:

1. Terry states that the assumptions underlying standard approaches are “about to collapse quite spectacularly due to the rapidly falling cost of complete genetic sequencing”. I don’t think this is entirely true, and it’s a bit of a myth that is becoming quite pervasive. There are three key drivers of the cost of sequencing: the assay itself, the bioinformatics analysis required to make sense of the basic test results, and the costs of generating the clinical evidence base, conveying results to patients and making decisions. Of these, only the assay cost is currently falling. The other cost drivers are either static or increasing (I blogged previously about a paper which quite nicely set out the real costs of sequencing).
2. Terry notes that the numerator (the difference in costs) in the standard ICER will change quite radically in economic evaluations in genetics. While I agree with this statement, I think Terry is inferring that it will become smaller (happy to be corrected here) whereas I would expect cost differences between genetic and non-genetic interventions to increase on a per patient basis. In particular, the costs associated with the actions taken on the basis of test results are likely to increase quite significantly, especially in cancer, as expensive new biological therapies are used in place of older established treatments (I blogged previously on Ibrutinib, which is a great example of this effect).
3. My reading of Terry’s blog post is that he is mainly talking about using genetic information to enable more precise cancer treatment (again, happy to be corrected). This is fine, but obviously there are a whole bunch of other clinical areas in which we currently use economic evaluation techniques, and in which genetic information currently has little or no impact (and won’t do for years, maybe decades, as clinical and basic science research focuses on the low-hanging fruit in cancer). There are also several clinical areas where genetic and genomic information is already useful, but standard approaches are likely to cope just fine. Here, I’m thinking in particular of sequencing pathogens to guide infection control decisions in hospitals. So while I can accept that there is an argument to be made about whether we should continue to use current economic evaluation techniques to make decisions in cancer, it’s a bit of leap to say that this will kill economic evaluation entirely as standard approaches will likely continue to be used in many clinical areas.
4. I do agree with Terry on several other points. The denominator of the cost-effectiveness ratio may indeed change radically. There are several applications of genomic testing that may result in very small changes in health outcomes (e.g. testing to diagnose learning disability and other developmental disorders), leading to a very small denominator (with an obvious knock on effect on ICERs). Such interventions may instead lead to larger changes in personal utility (‘benefits or harms that are manifested primarily outside medical contexts’), and this is an area where we may need new techniques to tease out the benefits of genetic and genomic testing (or at least the more widespread application of current techniques that decision-makers are yet to totally fall in love with, like the best-worst scaling approach developed by Terry and others). Terry also notes that uncertainty in genetic economic evaluations will be a growing problem, and I would add that this will have a knock-on effect on the use of the value of information approaches that are becoming somewhat more pervasive.
5. Finally, economic evaluation won’t be killed off because that just isn’t how health economics works, for better or worse. You only have to take a look at the life-cycle of the big value-based pricing idea to appreciate how difficult it can be to move the field forward at times! As an alternative example, look at how few cost-benefit analyses are conducted, even in areas where they may enable better decision-making (e.g. public health). Economic evaluation will evolve slowly but surely, as it always has done, incorporating new ideas as and when the evidence base is there to support their use. The impetus for change only tends to increase when there are outside pressures (e.g. from pharma or government).

So will treatment tailored to genes kill economic evaluation? Probably not – current approaches will probably still work in most cases, the evidence base isn’t changing quickly enough to drive rapid change, and there are no major outside pressures that could force the issue (at the moment). Cost-effectiveness analysis is the zombie technique that will never die.