Originally published Nov. 6.
By Turna Ray
Even among highly educated early adopters who have ordered gene scans from online companies — mostly to appease their curiosity about disease risk and genetic ancestry — some still have trouble understanding test results, according to a recent study.
A survey presented this week at the American Society for Human Genetics annual meeting of more than 1,000 customers of gene scans from 23andMe, Navigenics, and Decode Genetics found that although 88 percent felt their genetic test results were easy to understand, 38 percent indicated the reports from these companies were "too vague."
Additionally, within the random sampling of US customers of DTC genomics firms who received their results between June 2009 and March 2010, between 4 percent and 7 percent misinterpreted test results when they were shown two examples of risk data. In addition, surveyed customers were twice as likely to be confounded by a result showing the disease-protective impact of a gene, compared to a result showing that a gene confers increased risk of a disease.
The survey, conducted by Johns Hopkins University's David Kaufman and others, also found that consumers tended to feel they understood more than they actually did.
At a press conference at ASHG this week, Kaufman, director of research and statistics at Johns Hopkins University's Genetics and Public Policy Center, posited that participants might be more likely to grasp information about heightened risk than lower risk because most were seeking information about the types of diseases they should watch out for and paid more attention to that kind of information.
Overall, the survey revealed that most customers of DTC genomics services were satisfied with their experience and more than half of the study participants said they learned something new about improving their health by taking the tests.
“However, many of the study participants found the information from their genetic test reports to be vague, and one in 12 was not able to correctly interpret the sample results that we showed them, suggesting that there is room to improve the clarity of the information provided to customers in their personal genetic test result reports and the way the information is being delivered to them," Kaufman said.
The survey was conducted with the participation of the three DTC firms, which sent out e-mails to their customers about the study. Kaufman noted that the study did not look at the accuracy or clinical utility of the genetic test results provided through the services. Additionally, the study results were provided as an aggregate as not to compare the services provided by the three companies.
PGx Reporter reached out to the three DTC genomics firms in order to find out whether they were planning any changes to the way in which they present genetic risk data as a result of the study findings.
“We are always updating our product to ensure that it is understandable and correctly interpreted. We pride our product on being informative and easy to use," a 23andMe spokesperson told PGx Reporter, without including any specifics about whether the company planned to affect changes on its website based on the JHU survey results.
"Navigenics recognizes that, as with any novel technology, some information related to the service may not be readily understood," Elissa Levin, director of Navigenics' genetic counseling program, told PGx Reporter. "That is why, since its inception, Navigenics has not only provided user-friendly, comprehensive reports but has also been dedicated to providing access to qualified genetic counselors. Navigenics proactively reaches out to each individual to ensure that they have the opportunity to discuss and understand their results."
Levin also pointed out that Navigenics is the only company of the three firms in the study that provides genetic counseling support to its customers.
PGx Reporter was not able speak to Decode ahead of press time.
Meanwhile, the study also explored why consumers are purchasing the tests and what they did with the results. Around 94 percent said curiosity was their main driver for buying a gene scan; 91 percent said they wanted to learn about their elevated risks for diseases; and 90 percent of the customers of 23andMe and deCode, which offer ancestry testing, said learning about their genetic ancestry was a major factor for getting tested.
One in three participants purchased a test because a first-degree relative was affected with a tested genetic health condition, and 42 percent expressed interest in learning their risk for one or more specific conditions.
After taking the test, around 58 percent indicated they learned something new that would help them lead a healthier lifestyle, while 9 percent felt they could not change their health risks after receiving their test results.
Kaufman noted that the people most likely to make lifestyle changes were those with information showing they were at a higher likelihood for a particular condition.
[ pagebreak ]
Three in 10 study participants said they shared their gene scan reports with a doctor or had gotten a follow-up test, while 18 percent said that they intended to show their tests to their physicians at a future date.
As a result of testing, 34 percent of the participants said they watching what they ate, 15 percent made changes to their medication or dietary supplement regimens, and 14 percent were exercising more. Kaufman noted that long-term follow-up of DTC test users is necessary in order to draw conclusions on the extent to which genetic data impacts healthcare decisions and behavioral changes.
"The benefits and risk of these tests haven't really been well established yet," Kaufman said. "On the one hand consumers may get access and control to genetic data that can be used to improve their health, prevent disease, behave better, and perhaps choose treatments. The concerns are about the limitations, potential harms, and lack of clear benefits … Our research has started to scratch the surface of some of these concerns."
Another study, also presented at ASHG, drew similar conclusions about why people take genetic tests, when they have trouble understanding the results, and what they do with the data.
Barbara Bernhardt, clinical professor of medicine and co-director of the Penn Center for the Integration of Genetic Health Care Technologies at the University of Pennsylvania, interviewed 60 individuals who received personalized genetic risk results for seven health conditions through the Coriell Personalized Medicine Collaborative.
CPMC was launched in December 2007 with the aim of studying the impact of genome-informed treatment on medical care. The study, which has so far enrolled around 5,000 participants, has research partnerships with Cooper University Hospital, Virtua Health, Fox Chase Cancer Center, Helix Health, and others, and plans to ultimately enroll 100,000 participants. The project only reports genetic risk information on eight conditions deemed by the CPMC Informed Cohort Oversight Board as "potentially medically actionable."
Based on the interviews conducted by Bernhardt, most of the study participants "appeared to have a good understanding" of their genetic risk reports, but "some" had trouble interpreting their relative risk figures.
Most of those interviewed "seemed to discuss their risks in terms of binary thinking, in that they seemed to discuss their risks as being elevated risk or being at average risk, or they reinterpreted that to mean no risk," Bernhardt said at ASHG this week. "They would say, 'Yes, I have a high risk for this … and I have no risk for this.'"
Study participants also tended to view their genetic risk data in the context of their own medical and family history. "I think they did a very nice job of incorporating this information into their own perception of risk," Bernhardt said, adding that those she interviewed generally tended to understand that genetics is only one factor contributing to their risk for disease.
"When people learned about their risk for a disorder that wasn't present in their family already, they did express some surprise, but we really didn't notice that anyone seemed to be overly worried about any of their risk," she added.
While in her estimation, the people she interviewed didn’t appear to display an in-depth understanding of genetic risk, they did "well enough so they could figure out what the risk meant to them … and what to do with that information."
Around 30 percent of those she interviewed changed their behavior based on their genetic test results. These were usually small changes to their diet or exercise regimen. A few people made major modification: one person stopped smoking and another changed to a vegan diet.
Those who made no changes based on their CPMC reports thought they were already doing enough to take care of their health, or felt that their genetic risk probabilities for a disease weren't high enough to warrant any lifestyle changes.
Around half of those interviewed discussed their results with a doctor, while the other half planned to talk about it with their physicians. A quarter of the doctors made "reasonable" recommendations based on the genetic information submitted by their patients, according to Bernhardt, such as recommending a patient with a heightened risk of melanoma get a skin exam by a dermatologist.
One-fourth of study participants said they felt their doctors did not understand the test results or did not know what to do with the data. However, on the whole, people Bernhardt spoke to said they had positive interactions with their doctors about their genetic risk information.
[ pagebreak ]
Bernhardt said that in her interviews she didn't get the sense that people were overly concerned about privacy. They seemed open to sharing their genetic information and not worried about insurance discrimination. This might be because CPMC educates study participants about existing legal restrictions against using genetic information to make insurance or employment decisions, Bernhardt pointed out.
"Time will tell how people react to this information from their provider in the healthcare setting, when they're getting masses of information like they are getting from 23andMe," Bernhardt said, noting that those she interviewed were early adopters who decided on their own to learn about their genetic risk and that CPMC returns genetic risk data for only a handful of diseases deemed to be medically actionable.
Although the above studies looked at whether early adopters of DTC genomics understand the test results and what consumers are doing with the information, their investigations did not attempt to figure out whether these tests were returning accurate or useful results.
Questions regarding the clinical validity and utility of genetic tests sold by DTC services have been the subject of controversy in recent months.
Earlier this year, when DTC firm Pathway Genomics tried to sell saliva collection kits for its testing service at Walgreens, the US Food and Drug Administration quickly moved to inform all test providers in this burgeoning industry that their tests needed to be cleared by the agency (PGx Reporter 05/19/10).
Then, a congressional investigation into 15 DTC genetics firms — which included purchasing tests from 23andMe, Pathway, Navigenics, and Decode — found that the results of these reports were "misleading and of little or no practical use" (PGx Reporter 07/28/10).
In the study led by JHU's Kaufman, participants were asked how they felt about federal regulation of DTC genetic tests. Two-thirds of the participants felt that these tests should be available free of government oversight, but more than 70 percent thought that an organization such as the Federal Trade Commission or Consumer Reports should keep an eye on the scientific accuracy of claims made by DTC. According to Kaufman, more people tended to favor Consumer Reports, than the FTC, as an oversight body for DTC genomics firms.
Various efforts are underway to ensure that the data reported directly to consumers by these firms are accurate and do not lead to rash medical decision-making, including laboratory guidelines issued by the Centers for Disease Control and Prevention. Additionally, the National Institutes of Health is developing a Genetic Test Registry to keep consumers informed about the available evidence on marketed tests.
Speaking at the same press conference as Kaufman and Bernhardt, Andrew Faucett, director of the Genomics and Public Health Program at Emory University School of Medicine, recommended some questions that consumers and their doctors should be asking before purchasing gene scans directly off the web.
According to Faucett, although healthcare providers evaluate other types of medical tests on a daily basis, most doctors are not experienced when it comes to evaluating genetic tests. Before making healthcare decisions based on a DTC gene scan result, doctors should consider what the test is saying about a patient's health and what it cannot reveal; whether the test will be able to provide information to guide treatment decisions; whether it is a single-gene or a multi-gene test; and if the test has been studied in the right population so that results are meaningful to the patient.
Faucett advised consumers interested in using DTC testing services to discuss their results with a genetics expert. He further suggested that consumers consider before taking the tests whether the results will help them make healthcare decisions, and whether there is evidence backing up the clinical validity and utility of the tests.
“It is important for both healthcare providers ordering clinical genetic tests for their patients and individuals who are considering purchasing direct-to-consumer genetic testing services to evaluate these tests carefully before making decisions,” Faucett said. “Consumers, in particular, must be wary of reading too much into their test results, and they may even want to consult with a certified genetics professional to make sure they’re receiving a test that will provide appropriate information that could help them improve their health and assist doctors in making better, more informed decisions about their medical care.”
Overall, Faucett believes that most US labs are doing a good job of ensuring the analytical validity of genetic tests. “Most US laboratories performing genetic testing do an excellent job of analysis – however, that said, sometimes the test results may not provide the appropriate information to address the questions that the healthcare providers or patients are asking,” he said.
Have topics you'd like to see covered in Pharmacogenomics Reporter? Contact the editor at tray [at] genomeweb [.] com