**3. Conclusion**

This chapter contributes new knowledge about how to design the clinical components of a mobile health app. It does not replace existing guidelines and frameworks for mobile app development in general and mhealth apps specifically. Rather, it complements existing approaches and adds new information that may be helpful for app developers when designing the disease-specific features and functions of an mHealth app. Specifically, it adds the following to our current knowledge about designing mhealth apps: 1) It makes a distinction between disease-specific behaviors and needs and non-disease-specific barriers and needs. This enables app designers to dig down deeper into specific health-related experiences and encourages them to identify information in the literature which might illuminate app development. It provides a list of common barriers to mhealth app use and potential solutions to those barriers. 2) It makes a distinction between disease-related behaviors and app related behaviors. These are two very different types of behaviors and should have different design approaches, including use of different behavior change theories to underpin them.

This chapter also brings together several existing concepts and design patterns into a more coherent mhealth app design approach. We identify that human factors methodology is necessary but not sufficient for app development and recommend the expansion of user-centred design to human-centred design that considers a wider group of stakeholders, especially the HCP and the opinion leaders who influence them. We complement the use of personas with the Jobs to be Done framework and explain the key limitation of using personas in the design of apps compared to their use in marketing. We elaborate on the use of patient segmentation using disease-specific criteria, rather than generic or health-system level segmentation as a way of meeting the needs of different patient types and different app user types.

The limitations of this framework include the fact that the outcomes of using the framework have not been tested in the real-world and that there are two issues that the framework does not address. These include no research-based or proven guidance on 1) how to make a mhealth app engaging or 2) how to make an app visually appealing. In addition, we did not include any guidance on behavioral economic theory and choice architecture, also known as 'nudge theory' [91].

Overall, we believe this chapter will help app developers and publishers to build better apps that consider the disease-specific needs of patients and use behavior theories more effectively by conceptualizing disease-specific behaviors and appspecific behaviors as qualitatively different and therefore making apps more useful to patients and their caregivers.
