Stefan Elbers

167 Preventing relapse after successful treatment Although multiple reasons could have contributed to these limitations, an important factor may have been our real-world approach toward the use of the prototype intervention within the inclusionperiod.We expected that the active participationof the treatment teams during previous development phases would contribute to high patient inclusion rates. However, it is likely that the limited guidance on when or how to explain the workbook and absence of fixed procedures regarding patient recruitment increased the required effort for HCPs to integrate this study into their treatment routine. Although this means that the extent to which this workbook can work in IMPT programs is inconclusive, we did obtain important insights for further developing the prototype intervention. From an intervention design perspective, the feedback from actual use within the intended environment is crucial to further refine the strategies and adapt them to that specific context (O’Cathain et al., 2015). Regarding the low response rate on the in-depth interviews, some patients indicated that they already provided a full evaluation of the workbook in the telephone interview. Other patients mentioned the traveling distance as main reason. In addition, we believe that a moment of direct contact with the researchers prior to the telephone interview could have helped to better explain the importance of the interview and establish a good rapport in advance. For these reasons, this study should be regarded as the first iteration in the overall process of transforming a prototype into an effective intervention for clinical practice. Czajkowski et al (2015) emphasize the need for initial prototyping before conducting more stringent tests in order to first align the behavioral strategies to the clinical context in which they will be implemented. In addition, experimental medicine highlights the need for a stepwise approach toward intervention development. This framework consists of multiple subsequent steps that should be undertaken to examine the relationships between the intervention and its effect on physical functioning and the modifiable behavioral factors that mediate this relationship (Sheeran et al., 2017). Consequently, further development and testing are required and should indicate whether these strategies lead to a change in specific health behaviors such as goal setting and problem-solving and to what extent this change causes clinically relevant long-term improvements for patients with chronic pain. In addition, these limitations provide valuable information in preparation of future trials, including more emphasis on training HCPs in how to use the strategies, more integration of study procedures within clinical practice, and improved patient fidelity procedures to decrease dropout. Conclusion This first test of the relapse prevention workbook in a real-world setting of IMPT programs resulted in important insights regarding form, content, and use, as well as its interaction with the treatment program and study design. Although these initial results indicate a favorable evaluation of behavior regulation strategies within the workbook, this study