What are people willing to pay for whole genome sequencing information?

Given the wide variety of health and non-health outcomes associated with genomic tests, it is perhaps particularly important that the preferences of key stakeholders are considered within the health technology assessment process for these interventions. Indeed, in a paper published last year, Rogowski et al. highlight the importance of ‘preference-based personalization’ in this context. To date, few studies have generated data on preferences for genomic tests. However, a recent publication in Genetics in Medicine by Deborah Marshall and colleagues has attempted to address this gap in the literature.

The focus in this study was on whole genome sequencing (WGS), with the stakeholders in question being members of the general public. The key aim was to estimate their willingness to pay for WGS using a contingent valuation approach. A survey was conducted to collect this data, with 410 respondents. The first part of the survey asked respondents to complete a ranking exercise to elicit preferences for who decides which WGS results are included in a genomic report. The second part contained the contingent valuation questions, with respondents asked to estimate their willingness to pay for basic genomic reports (actionable information only) and for information of unclear clinical utility that could be provided on top of a basic report. These estimates were generated by asking respondents a series of 2-3 choice questions in which prices were varied sequentially, depending on previous responses.

I’m not going to replicate all of the results here but instead I want to focus on three findings that warrant wider discussion. First, 43% of respondents wanted to decide for themselves what information is included in their basic genomic report. This is a really important finding because the current focus (both in the UK and internationally) appears to be on a panel of experts making this decision, prior to rolling out a genomic test. This finding suggests that such an approach could have a notable effect on test uptake.

Second, 38% of respondents were not willing to pay anything for the basic genomic report, while 55% were not willing to pay anything for additional information of unclear clinical utility. These respondents represent a significant proportion of the general population and this is a worrying finding as genomic tests are not cheap. Unfortunately the authors did not dig any deeper into the responses of these respondents so we do not know why their willingness to pay was zero, which seems like a missed opportunity.

Third, average willingness to pay for the basic genomic test was $299. This is low, especially when compared to the average cost of WGS. On top of this, the authors could find no respondent characteristics that were significantly associated with willingness to pay for this report which begs the question: what could possibly be driving this result? I’m also left wondering how trustworthy this finding really is? The average willingness to pay for additional genomic information was also low at $180, although the willingness to pay regression had more explanatory power here (e.g. those who had previously had a genetic test had a higher likelihood of being willing to pay).

I hope that these findings prompt a flurry of further research into stakeholder preferences in this area because they have important implications for the design of future genomic testing services. In particular, I hope that either the authors or other interested researchers build on this contingent valuation study by applying stated preference techniques to dig more deeply into stakeholder preferences. As I read through the paper I couldn’t shake the feeling that a discrete choice experiment or a best-worst scaling study would have more clearly delineated all of the different trade-offs being made in this decision context, better explaining why people were making different willingness to pay decisions. Such studies might also have generated information on willingness to pay (or willingness to accept, depending on the design) for those who valued WGS at $0 in this study. To the authors credit, they do note this limitation in their discussion.

I would also take issue with two further points that the authors make in their discussion. First, they note that the characteristics of the genomic information provided in the report did not influence willingness to pay for this information, suggesting that it is the information per se that is valued, not the outcomes associated with this information. However, it was always extremely unlikely that the authors would find that the characteristics of genomic information influenced willingness to pay due to the design of the study: respondents were not forced to trade these characteristics off against one another.

Second, the authors note that their results provide evidence that individuals in the general population value actionable genetic information more than genomic information of unclear clinical utility. If my understanding of the study methods is correct, this statement is completely incorrect. Responses to the willingness to pay questions were constrained entirely by the starting and sequential costs presented to respondents. Non-zero willingness to pay values for the basic genomic report were allowed to range from $250 to $2,000. Non-zero willingness to pay values for the additional information were only allowed to range from $100 to $800. It was therefore somewhat inevitable that willingness to pay for the latter information would be lower than that for the basic report.

That said, I cannot argue with the final conclusion of the authors:

“…patient preferences should be used to inform policies and consent processes about WGS testing and how results are reported in the future because negative value may be associated with information generated from these results”

More than anything, this should be the take home message from the paper.

2 thoughts on “What are people willing to pay for whole genome sequencing information?

  1. “I hope that either the authors or other interested researchers build on this contingent valuation study by applying stated preference techniques to dig more deeply into stakeholder preferences. As I read through the paper I couldn’t shake the feeling that a discrete choice experiment or a best-worst scaling study would have more clearly delineated all of the different trade-offs being made in this decision context”

    Researchers at the BC Cancer Agency in Vancouver have touched on this exactly: http://www.cmaj.ca/content/early/2015/03/09/cmaj.140697. This work is ongoing.

    Like

    • Many thanks. I was aware of Dean’s paper and should have mentioned it. In fact it’s cited by the above study! Thanks for bringing it to everyone’s attention.

      Like

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s