Materials & Methods
A national survey was fielded in the US to gauge familiarity with attitudes and behaviors related to the coming shift from general population-based therapies to personalized medicine approaches, with a focus on the specific application of personalized medicine within oncology (seep Supplementary Material: Appendix A; https://www.futuremedicine.com/doi/suppl/10.2217/pme.14.74/suppl_file/suppl_appendix_a.doc see online at: https://www.futuremedicine.com/doi/full/10.2217/PME.14.74). To achieve this, a multiphased research approach was undertaken leading to the deployment of a consumer survey to a representative sample of 602 adults in the US, described below.
Description & Definition of Personalized Medicine
We formulated a description and definition of personalized medicine for consideration during survey development. After evaluating the feedback during the pretesting, final survey respondents were asked a series of questions regarding their familiarity with personalized medicine prior to being provided the following description and definition to complete additional questions on the survey.
Modern medications save millions of lives a year. Yet any one medication might not work for you, even if it works for other people. Your age, lifestyle and health all influence your response to medications. But so do your genes. Scientists are working to match specific gene variations with responses to particular medications. With that information, doctors can:
predict what diseases you may get in the future and attempt to either minimize the impact of that disease or avoid it altogether through the implementation of personalized, preventive medicine;
once diagnosed with a disease, tailor treatments, predicting whether a medication is likely to help or hurt you before you ever take it.
The first step was a targeted literature review focused on studies published in the past 5 years considering consumer or patient perspectives on personalized medicine and genetic testing. Databases searched include PubMed and Google Scholar. Search terms included 'personalized medicine +patient,' 'personalized medicine+consumer,' 'consumer+targeted treatment,' 'patient+targeted treatment,' 'patient preferences+oncology treatment,' and 'consumer and/or patient + individualized treatment.' Once identified, novel research studies or published papers reflecting relevant expert opinion were reviewed. Key issues related to consumer familiarity with and preferences towards personalized medicine were identified, and then validated with expert key informant interviews. Interviewees included three payer Medical Directors, a pathologist, two industry members developing oncology therapeutics and three oncologists. Key informants were selected based on personalized medicine content expertise and cascade sampling technique. Interviews were conducted by phone, lasted approximately 45 minutes and included a structured discussion of key topics expected to be covered in the survey instrument. The survey was then pretested with 15 consumers, randomly selected from the sample frame, to determine participant comprehension of survey questions and functionality. Pretest results were used to optimize the survey design. Questions were asked using Likert scales where possible, leveraging simple language to ensure a high level of participant comprehension. However, due to the respondents being a general population sample, variation in comprehension of scenarios presented is likely.
The survey was administered online to a representative sample of United States health consumers, ages 30 years and older. Participants were recruited by invitation through an internet-based survey panel, GfK KnowledgePanel. This panel has been used extensively in over 400 papers, articles and books, including several studies on genetic testing and is validated by the American Association for Public Opinion Research.[4,5,6,7] The panel uses address-based sampling with a published sample frame of residential addresses that covers approximately 97% of U.S. households. Participants without Internet access were captured by providing them with a netbook and Internet Service Provider. The sample of participants also included cell-phone only households. The sample was drawn from the 55,000+ member panel using a probability proportional to size (PPS) weighted sampling approach. KnowledgePanel participants are incented to complete the survey via a points system that is redeemable for rewards credits.
Confidentiality & Privacy Protections
The KnowledgePanel recruitment and empanelment process is designed to comply with CAN-SPAM and CASRO guidelines. Further, GfK policies conform to participant treatment protocols outlined by the federal Office Management and Budget, following guidelines from the Belmont Report. Survey responses are confidential; personally identifying information is never revealed to clients or other external parties without explicit respondent approval and a client-signed nondisclosure agreement. When surveys are assigned to KnowledgePanel panel members, they are notified in their password-protected email account that a survey is available for completion. Surveys are self-administered and accessible any time of day for a designated period. Participants can complete a password-protected survey only once. Members may withdraw from the panel at any time, and continued provision of the web-enabled device (e.g., laptop or netbook) and Internet service is not contingent on completion of any particular survey.
Participation in research is voluntary at the time that respondents are asked to join the panel, at the time they are asked to participate in any particular survey and at the time they answer any given question in a survey. KnowledgePanel participants are provided detailed privacy disclosure statements and releases prior to participating in the panel and subsequent surveys. More information about KnowledgePanel recruitment and privacy policies can be reviewed in Seep Supplementary Material: Appendix Bhttps://www.futuremedicine.com/doi/suppl/10.2217/pme.14.74/suppl_file/suppl_appendix_b.doc.
The survey data were analyzed using descriptive statistical for enumerating responses as proportions and percentages. Differences between subgroups were tested using t-tests and p values less that 0.05 were considered statistically significant (all differences noted below were statistically significant). When calculating percentages, participants who did not answer a particular question were excluded from the denominator for that question. Subgroups analyzed include differentiation by gender, sociodemographic characteristics and those reporting high or low preferences for personalized medicine.
Personalized Medicine. 2015;12(1):13-22. © 2015 Future Medicine Ltd.