TCGC: Clinical Genome ConferenceTCGC: Clinical Genome ConferenceTCGC: Clinical Genome Conference

Returning Genomic Incidental Findings to Patients: Cost Effectiveness and Decision-Analytic Models

Click Here to Return to Download Page  

Ann Nguyen:

Greetings, and welcome to this podcast from Cambridge Healthtech Institute for 2015's TCGC: The Clinical Genome Conference, running June 22-24 in San Francisco, California. I'm Ann Nguyen, Associate Conference Producer. Dr. Caroline Bennette is joining us today. She's the K12 Patient-Centered Outcomes Research Scholar with the Group Health Research Institute. Carrie, thank so much for spending some time with us.

Caroline Bennette:

Thank you for spending time with me as well.

Ann Nguyen:

Can you describe how you came to do patient-centered outcomes research and what it currently involves at the Group Health Research Institute?

Caroline Bennette:

Sure. I started my research career primarily as a biostatistician and I work primarily on the evaluation of prediction models and novel biomarkers, actually primarily in prostate cancer screening. When I went back to graduate school I was really interested in learning more about decision-analytic modelling and medical decision-making because I wanted to focus in on research that was really relevant to problems that patients and providers and policymakers were facing when they needed to make a treatment or technology adoption decision.

After I finished graduate school, I'm now a Postdoctoral Fellow at the Group Health Research Institute, and I really wanted to focus on patient-centered outcomes research, because for me this was a natural extension of the work in medical decision-making. Because really in order to provide evidence, data that can really help patients make medical decisions, we have to understand and incorporate the outcomes that are most important to them, rather than just what we as researchers think is important.

Now, I do think there's a lot of overlap between what researchers think is important and what patients say is important, but it's interesting to learn sometimes when there isn't perfect correlation, and that's where I'm focusing my research efforts now.

Ann Nguyen:

How do you generally determine the cost of effectiveness of returning incidental findings from next-generation sequencing to patients?

Caroline Bennette:

The first step in setting up any kind of cost effectiveness model is really setting up the specific decision or treatment option that you want to be able to compare. And so when we started the work that was recently published that I worked on with David Veenstra, one of the hardest things to think about was thinking about which incidental findings are you going to be returning, and to whom?

It was a really difficult question for us to try and answer or think about when we were trying to set up our cost effectiveness model. But then these policy guidelines came out from the American College of Medical Genetics, and that was kind of a clear policy guideline that we could evaluate, and it was kind of like, "Oh, this came out. Great. Well, how did they come up with this list, and did they really think about the long-term benefits and costs?"

It turns out I don't think they really did in any formal way. They obviously thought about it quite carefully informally, but in particular they didn't really think about the costs of returning incidental findings and the downstream costs and the long-term outcomes. And I think it was interesting that a lot of people started coming out saying, "This policy is going to bankrupt society," while others were arguing that it was actually going to save money.

We really wanted to kind of carefully evaluate the long-term healthcare outcomes and costs. So we started looking at the lists that the American College of Medical Genetics put together, and it was really interesting because of the 56 different pathogenic variants that they identified. Most of the variants that are going to be returned could be mapped to maybe 1 of 7 different clinical conditions. When we built our cost effectiveness model, we really focus in on those specific conditions.

Ann Nguyen:

When you develop analytic tools and policy models to aid decision-making in oncology, what challenges do you face, and how have you and your collaborators addressed them?

Caroline Bennette:

I think in the space of oncology, one of the key challenges going forward in terms of building these decision-analytic models is incorporating this new idea, or not necessarily new, this idea of personalized medicine. It's really becoming one of the most important drivers of new treatments in cancer and how patients are being treated going forward, and it holds tremendous promise, but it also holds a lot of challenges, not just for decision-analytic modelling but for a lot of other areas.

But in terms of building decision-analytic models, one of the things we actually usually do is talk about population averages. When we build our models, we're really talking about a population of patients to build models that are really individualized to an individual patient are much more complicated, and it's going to require a lot more effort to build these methods and incorporate them in a more timely manner.

On a second really related point is data, and timely data. So to build a good decision model, you really need good data, primarily from a clinical trial or a really well-done epidemiological study, and one of the challenges I think we as researchers face is that by the time these data are often published and available to be incorporated into a new decision model that we might build, oftentimes they're obsolete or not relevant to clinical practice because clinical practice has just advanced so fast. There's just so much research being done in this space right now that it's difficult to keep up with it.

Ann Nguyen:

Makes sense. Well, Carrie, thank you for revealing some of your research and experiences today.

That was Caroline Bennette of the Group Health Research Institute. She'll be speaking during the session, Clinical Sequencing: A Good Investment?, at TCGC: The Clinical Genome Conference, happening June 22-24 in San Francisco. To learn more from her in person, visit www.clinicalgenomeconference.com for registration info, and enter the keycode “Podcast”. This is Ann Nguyen. Thanks for listening.