Overall, among US consumers there was a lack of familiarity with personalized medicine. However, when provided definitions, respondents were optimistic about the prospect of personalized medicine providing safe and effective treatment options in oncology in the near future. Within that positivity there is significant variation across consumers in how they would embrace personalized medicine, their willingness to pay for it (both on the diagnostic and therapeutic side) and what they would do with test results that indicated they should forgo treatment.
Previous qualitative research by Bombard et al. has found that breast cancer patients value gene expression profiling (a form of personalized medicine testing for early stage breast cancer); however their understanding of the test was variable. Furthermore, factors relating to access to personalized medicine heightened the value of gene expression profiling to breast cancer patients. Similarly, we found that the understanding of personalized medicine was variable, however when explained further respondents were optimistic about the value of personalized medicine.
Our results are consistent with the available literature and recent consumer studies that speak to consumer familiarity and knowledge gaps, personalized medicine education challenges and preference variability. For example, a recent unpublished survey conducted by the Personalized Medicine Coalition found that a 'large majority of people have not heard of personalized medicine but react positively when it is described to them; most feel excited about the potential benefits of personalized medicine, including choosing a treatment that is most likely to work for them and the potential to prevent illness; and a large majority also recognize the value of these technologies and believe that they should be covered by insurance'. Another recent analysis of patients receiving genetic counseling associated with personalized medicine care found that participants had difficulty with basic genetic concepts and education to understand the complexities of genomic risk information was often needed. In another recently published study, authors found that 'a complex interplay of philosophical, professional and cultural issues can create impediments to genomic education of the public'. Other studies point out that levels of awareness related to genetics role in treatment selection were variable and that consumers are more willing to learn their risk for developing deadly diseases versus nondeadly ones.
Our study results add to the literature by exploring consumer preferences in greater detail among a representative sample where others use nonprobability-based samples like convenience, random dial, or voluntary sampling. Our study also adds by taking a specific focus on consumer perceptions related to genetic testing and oncology applications of personalized medicine. We also explore the differences in responses between both demographic subpopulations (i.e., education levels, gender) and between those who have had cancer and those who have not.
In concert with other studies in the literature, our study demonstrates a need for consumer education related to several aspects of PM's value proposition. For example, one critical need highlighted by the research is that consumers may not be willing to forgo treatment based solely on genetic testing. Compliance to testing and treatment algorithms, including forgoing treatments that are not expected to be effective, is required for personalized medicine to realize optimal value. If patients see genetic testing results as something to be ignored or challenged via second opinion when they suggest forgoing treatment, the paradigm loses significant value and reduces the potential for cost-effective care solutions. From a payer perspective, cost savings from personalized medicine depend on differentiated treatment pathways based on genetic profiling and associated response rates. As levels of awareness of and comfort with PM grow, it is expected that 'second opinion' redundancy would decrease and efficiencies would be realized.
Consumers' perspectives about personalized medicine and willingness-to-pay can provide useful insights for manufacturers as to the perceived value of different treatments in development. Today, patient cost sharing is routine and costs to the patient do play a significant role. The 2013 Employer Health Benefits Survey found that co-insurance rates of 16–38% of drug costs are typical within many health insurance plans, with higher rates associated with branded and/or higher tiered products. As patients are increasingly responsible for cost-sharing, their role as both patient and payer further supports the need to understand their perspectives on PM value. Consumer and payer preferences together will help align test and therapeutic product development programs with purchasing decision-makers. Additionally, the varying perspectives toward different attributes of personalized medicine captured in this study can inform development of value-based evaluations by industry, payers and clinicians.
Ultimately, to reach the true potential of a personalized medicine care paradigm, the perspectives of consumers must be understood and addressed within education and outreach initiatives. From familiarity and knowledge gaps of diagnostic and treatment options, to concerns about cost, there remain several unanswered questions for consumers related to personalized medicine.
While our research reviewed consumer perspectives from the US, consumer perspectives from other geographic contexts are likely to vary based on practice, cultural, and healthcare financing differences. For example, one recent study noted patient preferences for personalized approaches to breast cancer management, but systemic factors (payer and clinical gatekeepers within the Canadian health system) rather than treatment preferences prevented access. This example demonstrates the need to further assess patient preferences within markets rather than taking key findings and generalizing them across all settings. Future research focused on exploring the heterogeneity among consumer and patient perspectives within markets would be beneficial.
Furthermore, measures of patient preferences examining risk–benefit trade-offs in genetic testing need to be examined. Although previous research and our current research provide valuable qualitative information regarding the value of personalized medicine, further research needs to be done to quantitatively estimate the value of personalized medicine. Conjoint analysis is an accepted method in healthcare used to quantitatively measure stated patient preferences by forcing respondents to make benefit-risk trade-offs when making choices.[18,19]
Lastly, the differences between a healthy consumer, a current patient and a consumer who has experienced a life threatening disease must be considered. This is an area requiring further research to determine the nuances between these three groups, only briefly touched on within this research. Future studies should investigate how consumer trade-offs among the various attributes (i.e., costs, utility expectations) impact preferences. Measuring the risk–benefit trade-offs in this manner will affect estimates of willingness to pay, expectations of utility and other consumer preferences.
Strength & Limitations
In order to put study findings in context, the authors note several strengths and limitations. First, the large representative sample of consumers in the US provides an opportunity for generalizability of results not possible with smaller samples. Additionally, due to the design of the consumer panel, significant demographic data were known about each survey participant, including gender, sociodemographics, payer and relevant disease history. This enabled greater granularity in data analysis, without consumers having to complete an extremely long survey – which can lead to survey fatigue and drop-out. Additionally, knowing this information a priori decreases reporting bias of each of these variable points. There are also certain limitations of the study, including how financial incentive for participation may introduce incentive bias and the definition of personalized medicine used in this study may not account for all complexities associated with the discipline or regional variations in how the term is applied. Additionally, the survey instrument used was a nonvalidated tool, though tested with key stakeholders prior to use. Lastly, while the survey was cross-sectional in design and thus does not compare the evolution of thought or experience prospectively.
Personalized Medicine. 2015;12(1):13-22. © 2015 Future Medicine Ltd.