Risky business: risk perception and the use of medical services among customers of DTC personal genetic testing.

Direct-to-consumer genetic testing has generated speculation about how customers will interpret results and how these interpretations will influence healthcare use and behavior; however, few empirical data on these topics exist. We conducted an online survey of DTC customers of 23andMe, deCODEme, and Navigenics to begin to address these questions. Random samples of U.S. DTC customers were invited to participate. Survey topics included demographics, perceptions of two sample DTC results, and health behaviors following DTC testing. Of 3,167 DTC customers invited, 33% (n = 1,048) completed the survey. Forty-three percent of respondents had sought additional information about a health condition tested; 28% had discussed their results with a healthcare professional; and 9% had followed up with additional lab tests. Sixteen percent of respondents had changed a medication or supplement regimen, and one-third said they were being more careful about their diet. Many of these health-related behaviors were significantly associated with responses to a question that asked how participants would perceive their colon cancer risk (as low, moderate, or high) if they received a test result showing an 11% lifetime risk, as compared to 5% risk in the general population. Respondents who would consider themselves to be at high risk for colon cancer were significantly more likely to have sought information about a disease (p = 0.03), discussed results with a physician (p = 0.05), changed their diet (p = 0.02), and started exercising more (p = 0.01). Participants' personal health contexts--including personal and family history of disease and quality of self-perceived health--were also associated with health-related behaviors after testing. Subjective interpretations of genetic risk data and personal context appear to be related to health behaviors among DTC customers. Sharing DTC test results with healthcare professionals may add perceived utility to the tests.