Treatment decision making in papillary thyroid cancer: an online discrete choice experiment survey

The intervention aims to test the factors (trade-offs of risks including surgical side-effects, cancer progression and risk of mortality) that are most important to patients when deciding between treatment options (total surgery, partial surgery or active surveillance, which will be described as monitoring to participants) for papillary thyroid cancer, and whether this differs based on the terminology used to describe the condition. This understanding will enable further discussion and understanding as to how the varying terms to describe the same diagnosis change patient decision making and how powerful and influential the term ‘cancer’ can be. It will also provide insight into the acceptability of active surveillance as a management option for papillary thyroid cancer. Participants in this study will be men and women across the Australian community identified and recruited using a market research company specialising in online panel survey research (Survey Sampling International, SSI). This company maintains an extensive database of over 100,000 Australians, who have indicated their willingness to be involved in online survey research. To be eligible participants need to be above the age of 18, be fluent in English and have no previous history of a thyroid cancer diagnosis. Potential participants will be recruited to reflect the general population in terms of sex, age, cultural background, education, employment status, income. Participants will be excluded from the study if they are below the age of 18, unable to speak adequate English and have been diagnosed with thyroid cancer. The intended sample size for recruitment will be between 750-1000 participants in each arm (in total between 1500-2000 participants). Piloting the survey will produce results to allow for more a more definite sample size to be determined. Previous experience with discrete choice experiments (DCE), and the use of an efficient DCE design that will be generated using specific software to minimize sample size requirements, will ensure that this sample is sufficient to accurately and robustly estimate model parameters and subsequent benefit harm trade-offs. This quantitative survey study is intended to take approximately 3-5 months for recruitment and completion of the survey by participants. The study will be conducted via an online survey, accessed through SSI. Each survey will taking approximately 10-15 minutes for participants to complete. Data will be collected directly from participants in the form of an online survey. Once invited by email by SSI, participants will be provided a link to login in to the discrete choice survey. Participants will be provided with an online Participant Information Sheet. As this is an online survey, no signed consent will be sought, rather, if respondents proceed beyond the Participant Information Sheet and consent information and complete and submit the survey, this will be considered evidence of consent (as discussed in NHMRC National Statement on Ethical Conduct in Human Research (2007)). Participants will be randomised to either receive the cancer term (papillary thyroid cancer) or non-cancer term (papillary lesions). The discrete choice survey will consist of at least 6 choice questions comparing different decision making scenarios. Participants will be asked to choose the option they most prefer for each question. This will be presented as a two part question to give people an option not to choose either scenario. Following completion of the discrete choice questions, participants will be asked some additional sociodemographic questions such as age, education level, employment status and personal and family history of cancer. By observing participant choices for different scenarios using DCE, the analyst is able to construct sets of utility functions. These utility functions model the weights that individuals attach to each of the attributes, and as such, allow for an estimation of the relative importance of each attribute.