Hex image

Carrying the Patient’s Voice into Health Care Decision-Making: Five Recommendations for Preference Studies in Oncology Therapies

Carrying the Patients' Voice Into Healthcare Decision-making

Patient-centered drug development may be the goal of most manufacturers, but that goal can prove elusive if patient preference information is not considered along the product lifecycle. However, conducting patient preference studies is not always straightforward, with indication-specific challenges that must be addressed. In this blog, Sebastian Heidenreich, associate director and scientific lead of Evidera’s patient preference team, outlines five recommendations for patient preference studies in oncology.


Patient preference information can be considered for multiple decisions along a product’s lifecycle, including endpoint selection, target product profile refinement, decision maker engagement (e.g., regulators or payers) and post-launch support that aims to shape treatment pathways. To ensure this information is robust, reliable and impactful, considering application-specific challenges is vital.

Evaluating cancer treatments typically requires trade-offs among survival, adverse events and convenience, with potentially large differences in the relative importance among these treatment aspects (e.g., dominance of survival). Combined with potentially small sample sizes stemming from low incidence rates, and/or difficult access to patients, this makes preference elicitation in oncology challenging. It is with such challenges in mind that Evidera’s patient preference team recently conducted a systematic review of discrete choice experiments (DCEs) concerned with medical cancer treatments. The goal was to describe current practice, identify gaps and derive recommendations for an agenda in oncology preference research. Our five recommendations from that review are outlined below.


Despite an increasing amount of preference research in oncology, our insights are still limited. To obtain a wider perspective, more oncology indications need to be considered, given that 55% of oncology DCEs published since 1990 were conducted across only six indications: 1) breast; 2) skin; 3) prostate; 4) blood; 5) lung; and 6) bone. Yet the oncology Phase II pipeline reveals high activity in less researched indications like pancreatic, head and neck, brain, liver and bladder cancers, with no or limited supporting evidence on patient preferences.


There are notable differences in preferences among stakeholders, and all of these viewpoints should be explored to contextualize treatment decisions. Our systematic literature review revealed that clinician preferences were driven by overall survival more than preferences of any other stakeholder, including patients. The general population tended to place a greater weight on health-related quality of life and were less concerned about treatment risk. Caregiver preferences remain underexplored, including treatment preferences of parents of children with pediatric cancers. Broadening the valuation perspectives to multiple stakeholders provides a richer understanding of how patients’ treatment priorities align with the priorities of others involved in cancer care. This is especially important for access and reimbursement decisions that are currently not considering patients’ preferences on a routine basis.


As mentioned, preference research can be challenging in oncology, where sample size is often limited and the relative importance among attributes potentially can be unbalanced. For example, our systematic review found that 53% of the studies conducted with patients included a sample of 150 or fewer subjects. This raises important questions about the choice of an adequate instrument. While DCEs are often considered the gold standard of preference elicitation, they may not be fit for purpose in many oncology contexts. This is particularly the case if the sample size is small, the decision problem complex, and preferences are also potentially heterogeneous. Researchers and decision makers should be open to a wider choice of instruments, including thresholding and swing-weighting that can elicit preferences at the individual level.


Sample sizes in oncology clinical trials often make instrument pre-testing a challenge, with quantitative pilots rarely being feasible. Thus, we should aim to make the most out of qualitative pre-testing and explicitly assess challenges that we know can arise within an oncology context. Don’t just limit your pilot to determining the clarity of wordings and figures‒assess attribute dominance, attribute non-attendance and alternative neglect using structured note-taking systems and evaluation frameworks. More innovative researchers may consider going beyond classical think-aloud methods by employing eye-tracking approaches that have proven to be useful at the pilot stage. Additionally, be open to the idea of assessing multiple methods and design factors (e.g., attribute overlap) at the pre-testing stage.


Finally, researchers should report instrument development methods and results fully to improve the overall impact of their study. In our systematic review, only 66% of the studies explicitly reported on qualitative findings and, of those, only 31% reported qualitative findings in support of choice about these levels. This makes the appraisal of instruments challenging. There are many ways to select treatment attributes for inclusion in a preference elicitation instrument, so it is important to be honest about your process. If you used qualitative methods, demonstrate how qualitative findings led to the attributes included in the final instrument. To maximize the impact of study findings and related applications, researchers should adhere to best practice reports and guidelines.


As the U.S. Food and Drug Administration (FDA) and European regulatory groups encourage manufacturers to include patient preference data in submissions, pressure has grown on researchers to include this approach from discovery through commercialization. For oncology researchers, there may be added difficulties, but the opportunity to enhance patients’ lives in ways that are meaningful is significant.

Explore our patient-centered research offerings