Some people choose to take part in genetic research, while others don't, and researchers question whether there is a difference between these groups. As part of the multi-center TRIUMPH study of heart attack patients, David Lanfear and his colleagues at the Henry Ford Hospital in Detroit, Mich., examined a variety of factors — from sociological information to recruitment site — to address this question. Knowing the answer, the researchers say, will clarify the applicability of study results to the wider population and inform the implementation of personalized medicine. Genome Technology's Ciara Curtin recently spoke with Lanfear — what follows is an excerpt of their conversation, edited for space.
Genome Technology: Was there anything in particular that prompted you to do this study?
David Lanfear: Not really. I thought that our data set was a good way to approach it because it was a multi-center data set, so it's not all from one place. And then the data set collected a lot of patient-level information, so we could look at a lot of different factors. Unlike a lot of large genetic registries or other approaches to this — asking people if they would or wouldn't [participate] — people had actually said yes or no. So, I thought it was a good real-world example. I thought it might be different also because it was in the setting of a heart attack.
In theory, you could have self-selection that makes your data set less generalizable or less representative of the larger population, so we also wanted to look at that.
GT: You found that enrollment site especially had an effect on participation rate. Was that surprising?
DL: It was. There's been a lot of previous research on patient-level factors — things like race — that have been consistently shown to be an issue, and we found that also. Some of the previous data had disagreements, but sometimes gender or education or age could impact whether people participate in research or not. And then genetic research seemed to be one step slightly more challenging to get people to recruit into.
GT: How might researchers account for the enrollment effect?
DL: The simplest and most straightforward thing is to try to make the process as uniform as possible. Some of that is possible and some of that may be a challenge. In our study, we attempted to try to do that, too, but maybe we have to do a better job of making sure people are all approached for genetic research — make sure the presentation is the same. That's one thing. And then the other thing is figuring out why it's so different. That's a little bit of a challenge. That would require additional work to figure out why it's high at some places and low at some places. You'd almost have to go observe a certain number of enrollments at each place to try to sort out what are people doing differently that makes it successful or unsuccessful.
GT: What are the main challenges to using results from genetic studies in the clinic for personalized medicine?
DL: Using it clinically, I think there [are] several different challenges. One of them is just having enough information and a strong enough association to act on. It's not necessarily one variant that is so strong that we're going to find that we can actually use it clinically. It will probably be some sort of combination of genetic variants. I think that's something that we're still learning more about.
The other thing is getting people, both physicians and patients, used to using genetic information clinically. There's not necessarily any reason it would be any different than using any other piece of medical information we have currently. You check somebody's kidney function and you take that into account when you are designing their regimen or what you are going to do with them. So, the genetics would be the same. People have to stop regarding it as something different than any other biomarker, any other marker of biologic process.