Research into patient preferences and the development of a concrete tool to measure them

During the last two decades, shared decision making (SDM) has been increasingly applied in multiple health care decision contexts. Within the model of SDM, a patient and physician share information regarding the benefits and risks of potential options and personal preferences. They discuss the patient’s preferences for each of the available options and discuss the patient’s desire for involvement in decision making. Finally, they make or defer a decision and arrange follow-up if applicable. SDM is the preferred model for medical decision making in case of preference-sensitive decisions. Within these decisions, two or more options are available, none of which is clearly superior. To facilitate the process of SDM, patient decision aids (PtDAs) can be applied. The aim of a PtDA is threefold: they make the decision explicit and provide evidence-based information on the potential options, they help patients clarify their values and preferences for the different options and they support patients to communicate their preferences to their health care provider. PtDAs exist in various formats, ranging from paper-based brochures to interactive digital applications, using audio and video elements. The International Patient Decision Aids Standards Collaboration has established a checklist for the development of PtDAs, based on 12 quality domains, covering content, development process and effectiveness of the PtDA. Internationally developed PtDAs have proven to increase patients’ knowledge and accuracy of risk perception, and decrease their decisional conflict. There are multiple PtDAs available in the disease areas of breast, prostate and lung cancer, osteoarthritis and osteoporosis, end of life decisions and cholesterol-related diseases. The variety in disease areas highlights that preference-sensitive decisions occur in many different health care decision contexts. The disease area of hormone-sensitive breast cancer is particularly suitable for the implementation of a PtDA. Patients receiving adjuvant endocrine therapy undergo a very long treatment that often causes a high impact on quality of life, resulting in suboptimal treatment adherence. The available treatments, tamoxifen or aromatase inhibitors, may cause adverse events such as increased risk of blood clots or joint and muscle pain, respectively. The efficacy of both treatments is comparable, although aromatase inhibitors are considered to be the most effective option. As treatment impact varies considerably between patients, a PtDA can help to tradeoff potential benefits and adverse events to determine the optimal therapy for individual patients. Although the practice of SDM and the use of PtDAs is internationally recognized, there are currently only very few PtDAs available in Belgium and there are no examples available for patients with breast cancer.

This PhD project aimed to improve SDM for Belgian patients with hormone-sensitive breast cancer by developing a PtDA for decisions regarding switching or continuing adjuvant endocrine therapy after 2-3 years. This PtDA should meet the needs of both patient and physician, by informing patients on the available decision options, eliciting their preferences for these options and supporting them to communicate their preferences to their treating physician. In order to meet this aim, four specific objectives were identified.

The first objective was to compare the process of decision making between patients and consumers and to identify innovative aspects of preference elicitation methods from the consumer research field. A literature review showed that the decision process between patients and consumers is highly comparable and identified five concepts from the consumer research field that might improve preference elicitation in healthcare. Preference elicitation methods that resemble real-life decision making as closely as possible, for example by providing time for self-reflection, are likely to generate the most accurate results.

The second objective was to assess the needs of both patients and physicians regarding a PtDA for breast cancer decision making. Four focus groups with 21 patients and five individual interviews with breast cancer specialists indicated that patients currently experience little involvement in their treatment decision making. Patients furthermore indicated a high need for information regarding treatment options, especially quantitative information on treatment benefits and risks. The breast cancer specialists acknowledged that SDM has become more important during the last decade but stated that it remains unclear how to implement this in clinical practice. Decisions regarding adjuvant endocrine therapy were deemed ideal for a PtDA intervention by both stakeholders as these treatments may have a high impact on quality of life and deliver limited benefits. Beneficial PtDA features were identified, such as the possibility to write down questions or to rate the impact of adverse events on daily life.

The third objective was to develop an interactive, online PtDA for patients with hormone-sensitive breast cancer switching adjuvant endocrine treatment. The combination of an in-depth literature review and stakeholder interviews were used to determine the content and design of the PtDA. Five attribute categories were identified using this approach: efficacy, adverse events, use, impact on quality of life and mechanism of action. Both patients and physicians rated potential attributes to determine the final selection, which included breast cancer mortality, risk of recurrence, treatment duration, joint and muscle pain, osteoporosis and increased thrombosis risk. The developed prototype PtDAs consists of three consecutive modules; an information module aiming to educate patients, a scenario-based module that will help patients to clarify potential impact on their everyday life and an adaptive conjoint analysis exercise to elicit patients’ preferences for various treatment characteristics.

The final objective was to test the developed PtDA in a two-stage process. First, alpha testing in a research setting yielded an average usability score of 78.75 out of 100 using the System Usability Scale. Furthermore, content and lay-out were scored 8.9 and 8.5 out of 10 respectively by 11 patients; and quality, completeness and lay-out were scored 8.4, 8.4 and 8.2 out of 10 respectively by five health care professionals. Qualitative feedback was gathered by applying cognitive interviewing while using the PtDA and a short interview afterwards. Next, after implementing the feedback received during alpha-testing, beta testing in a clinical setting was performed using a pilot trial. Nine patients tested the PtDA in the week before their planned follow-up consultation at the university hospital of Leuven. The effect of the PtDA was assessed by determining the impact on the quality of the decision process and the decision itself using the following constructs: knowledge, values-choice agreement, feeling informed, feeling clear about values, discussing goals with health care providers, and being involved. Patient knowledge increased from 5.33 before using the PtDA to 7.78 afterwards. Decisional conflict was low after the consultation, with a score of 18.06. One month after the consultation, decisional conflict had increased significantly, with a score of 41,67. This might indicate the need for a broader support framework for patients by providing information and support them to discuss their preferences over a longer time period. Making the PtDA available after the consultation and planning follow-up consultation when needed could be potential solutions. The extent to which SDM was applied during the consultation was assessed using the SDM-9 item questionnaire. The average score of 20 out of a possibly 45 indicated that no real SDM took place, highlighting that patient directed interventions only, are not sufficient to implement SDM in clinical practice. Usability was again assessed using the System Usability Scale, resulting in a mean score of 71.25. Finally, patient preferences elicited for different treatment characteristics revealed high variability between patients, with a total of five different attributes being selected as ‘the most important one’ by seven patients. These results indicate the need to discuss individual treatment preferences during consultations.

Based on the information gained through the different chapters, recommendations are formulated for the future development of PtDAs in Belgium and the implementation of SDM in clinical practice.

The first set of recommendations is related to the development process of PtDAs. As the current use and development of PtDAs is very limited in Belgium, more research in other disease areas is needed to truly enable patient centered care. Stakeholder input has proven to be indispensable during the development process. Therefore, stakeholder opinions regarding PtDA design and content should be assessed using a combination of qualitative and quantitative approaches. Different formats or visual displays should be tested to meet the specific needs of the target group. Interactive, online applications offer a range of advantages regarding implementation and use. Video and audio materials can facilitate the learning process for users. Interactive features such as content control or the use of narratives may improve decision making. Furthermore, explicit preference elicitation methods that simulate real-life decision making may improve value clarification. The generated preferences weights can open up the discussion during a consultation. More research is required to facilitate optimal implementation in Belgium. The implementation strategy for a PtDA should already be considered during the development process. A centralized platform, either for the whole of Belgium or for Flanders, may provide easy access to PtDAs. However, not all patients are familiar with an online environment, nor does everyone have internet access. The needs of the target group should be assessed to identify the best implementation approach to allow for patient-centered care.

The second set of recommendations is developed to improve SDM in Belgium. The main recommendation here is to create awareness for SDM and to support implementation in clinical practice. A combination of clinical guidelines and practical measures such as logistic and financial support are needed. More research is needed to identify and address the barriers of Belgian health care professionals for SDM implementation. An example can be financial support to reimburse physicians for the time spent on the conduct of true SDM, for example in analogy to the compensation pharmacists can receive for performing counseling on the use of diabetes or asthma medicines. Moreover, financial support will be required to stimulate research on the development of PtDAs and the implementation of SDM. Furthermore, initiatives from various stakeholders and organizations should be harmonized in one collective approach. We should capitalize on the knowhow built up during international research, especially from countries with a comparable health care system such as the Netherlands. Care pathways in both primary and secondary care might need to be revised, to allow patients the necessary time to inform themselves and participate in decision making. More quality measures are needed to monitor the extent of SDM in clinical practice and to improve where needed. Another important recommendation is to apply trainings in SDM for health care professionals. If we truly want to implement SDM in routine clinical practice, a change of culture and mindset is needed. Only focusing on patient directed interventions will not suffice to accomplish this. Finally, public awareness for the right to engage in SDM should be raised. Patients should know where to find relevant information regarding their medical condition and available options at all times. More patient directed interventions such as PtDAs should be made available, either by developing new interventions or by translating and adapting PtDAs from the international scene.

By implementing these recommendations, we could truly start our journey towards patient-centered care.