*1.7.2 Ability to change health behaviors requires an understanding of the range of ways that patients experience their disease*

For a mhealth app to be effective, it needs to not only provide evidence-based advice, but also needs to capture and track relevant disease-specific patient behaviors. In diabetes, this is likely to be nutrition, exercise, stress, medication adherence, regular monitoring of biomarkers and sleep. In myasthenia gravis, it will include medication adherence, exposure to exacerbating factors, stress and potentially muscle strength.

Regardless of the relevant disease-specific behaviors, different patients with the same disease have different preferences and beliefs which also impact their choice of treatments, their health behaviors and approach to interacting with the healthcare system [51]. These differences in patient experience need to be considered during the app design and development process. These differences can be captured as part of the persona development process.

#### **1.8 App-related behavior change is not health behavior change**

How users relate to apps is very different from the way they relate to their disease. Health-related behaviors are highly dependent on the nature of the disease, while app related behaviors are dependent on characteristics of the app and the clinical setting; i.e., ease of use, user experience, getting the job done and a sense of accountability to the HCP and the healthcare team.

A significant majority of mhealth apps do not use a theoretical basis for driving behavior change [52]. Geuens et al. [53] describe a tool which links behavior change theories to mhealth app features to make it easer for developers to incorporate behavior change theories into their applications [54]. The features in the Trustworthiness and Liking section are all related to better app use. An excellent and comprehensive list of methods to improve app usability mentions many good techniques to improve app usability and the user experience, including 1) minimizing cognitive load, 2) minimizing response burden, 3) breaking down tasks into smaller sub-tasks, 4) making it easy to navigate and know where you are in the app, etc. [55].

#### **1.9 Not all patients are alike. Patient segmentation is necessary for personalizing care**

Not all patient segmentation approaches are relevant to mhealth app design. For example, the approach recommended by Brommels [56] which breaks down

#### *Designing Disease-Specific mHealth Apps for Clinical Value DOI: http://dx.doi.org/10.5772/intechopen.99945*

patients into 7 categories (healthy, incidental, chronic disease, multi-morbid, needing elective treatment, trauma, requiring hospitalization) is more suited to health delivery systems and how to segment patients for health delivery planning. The approach used by Deloitte who categorize health consumers into 4 categories (trailblazers, prospectors, bystanders and homesteaders) is more suited to understanding patients as consumers and purchasers of healthcare as opposed to as individuals who are purchasing a product to solve a problem [57].

What app developers need is a method to segment patients to enable more personalized solutions and communications for patients [51]. A Finnish organization has developed such a segmentation approach for Diabetes [58] and Sandy et al. developed one for hypertension [59]. Ideally, app developers should find a well-researched patient segmentation approach for the disease for which they are building an app.
