Stefan Elbers

166 Chapter 6 in the assessment and treatment of chronic pain and is particularly suited for tailoring specific strategies to individual needs and preferences (Keogh et al., 2010). With machine learning approaches, it is even possible to automate the process of personalizing the strategies based on user-generated data (Rabbi et al., 2018). Another advantage of mHealth is the opportunity of letting both strategies interact. For example, if a patient used an Insight Card to highlight an effective strategy to overcome barriers to physical activity, this card could also be used as a future solution to anticipated problems within the planning procedure of VBG. Although the idea of a digital intervention had already been suggested by stakeholders in earlier development stages, we did not make any decisions on its final form prior to this feasibility study. Because our study findings are in accordance with these earlier suggestions, we believe there is potential in transferring this prototype workbook into an mHealth app. Limitations Limitations of this study include a small sample size and relatively high dropout rate. Because the study was designed, conducted, and analysed by the same three researchers, who were also involved in the development of the workbook, confirmation bias and socially desirable responses may have resulted. Furthermore, due to organizational reasons at location Hoensbroek directly after the inclusion period, we were only able to collect evaluations from HCPs from location Maastricht, where regular reflections on patients’ values are beyond the scope of their treatment program. With these limitations in mind, it is important to reflect on the validity of the conclusions of this study. Concerning the adequacy of the sample, Malterud and colleagues (Malterud et al., 2016) have introduced the concept of information power, which is determined by five factors: narrow or broad study aim, sample specificity, established underlying theory, quality of dialogue, and type of analysis strategy.We believe that the specific focus in our study objective on 3 key factors for feasibility positively contributed to the information power of this dataset. Furthermore, all participants that we interviewed received instructions to use the workbook, participated in an IMPT program, and had—at minimum—made an effort to use the workbook in this setting, which not only resulted in high specificity but also to a high quality of dialogue. In addition, we included several established procedures to enhance the credibility of our findings and minimize bias, including member checks, triangulation of researchers and data sources, and including questions regarding negative experiences with the workbook to search for disconfirming evidence (Carter et al., 2014; Creswell &Miller, 2000; Nowell et al., 2017). However, we conducted a cross-case analysis and the low sample size resulted in limited variation on personal characteristics and a low likelihood that potential problems in use did occur within the sample (Malterud et al., 2016; Viechtbauer et al., 2015), which limits the generalizability of our findings.