This study was conducted with dietitians from the United Kingdom, Australia, and New Zealand. The dietetic associations for each respective country assisted in recruitment by distributing a link to the Web-based survey via their weekly member electronic newsletters, social media (Facebook) post, or emails directly to each of its members. Eligible participants had to be a Registered Dietitian (United Kingdom, New Zealand) or Accredited Practising Dietitian (Australia). The recruitment process has been described in more detail elsewhere [19]. Approval for this study was granted by the University Human Research Ethics Committee (approval number: 2015/701).

An interpretive paradigm was adopted by this study, which sought to understand the individual experience and the meaning they attribute to their actions [25]. As such, open-ended questions were selected as the research method to elicit new and more diversified information, especially on topics or experiences where there is limited information [26]. Open-ended questions also provide opportunities for respondents to share more rich and detailed opinions than that which could be achieved with close-ended questions alone [26,27]. Therefore, as part of a larger cross-sectional survey aiming to investigate dietitians’ use of smartphone health apps and other mHealth technologies in practice, the 3 open-ended questions shown in Textbox 1 were included to allow more specific exploration of app design features and additional resources or tools which could enable health apps to better support dietetic practice. Detailed methods about the piloting and development of the survey have been described elsewhere [19,28].

Inductive thematic analysis, guided by the framework described by Braun and Clarke [29], was applied to the survey questions. This involved 6 phases: (1) familiarization with the data, (2) initial code generation, (3) searching for themes, (4) reviewing themes, (5) defining and naming of themes, and (6) writing up of results. Coding of the responses was performed using the NVivo qualitative data analysis software version 11 (QSR International Pty Ltd). One researcher (JC) conducted the analysis, coding all the responses to allow for data immersion and to obtain an overall sense of the entire dataset. Generated codes and subsequent themes were checked through a process of ongoing discussion with a second researcher (MAF) who was familiar with the data, before finalization.

This theme captures dietitians’ uncertainty over the credibility of apps, making it difficult for them to recommend apps to their patients. Greater reviewing and endorsement of credible evidence-based apps by dietetic associations and collaboration between app developers and dietitians could improve the confidence of the profession in using and recommending apps in dietetic practice.

This theme explores the usability of health apps and the app design features which could enhance their ease of use across a range of users, both for the dietitian and their patients.

This theme describes the app design considerations which should be addressed with regard to dietary assessment and behavior change and the dietitian-specific tools which should be implemented to support dietitian-oriented tasks.

This theme identifies two main strategies for improving patient-oriented support, namely, through the option for tailoring apps to individual needs and by providing patient-specific tools for self-management of health.

This theme highlights how sharing of health-app data could improve patient-provider communication and care through enabling greater integration of these mHealth technologies into dietitian work systems.

To our knowledge, this study is one of the first to identify design considerations, including features and tools, for apps supportive of dietetic practice. Dietitians prioritized several design aspects, including the credibility and usability, including ease of use and the design of apps. Apps targeted toward dietitian- and patient-oriented support and that could integrate into dietitian work systems were also regarded favorably. These findings provide guidance to app developers about the fundamental characteristics to address while designing dietitian- and patient-oriented apps.

Dietitians are guided by codes of professional conduct to provide evidence-based practice, which extends also to the promotion of products, including health apps [30-32]. However, recommending apps in a professionally responsible way has been challenging since regulation only exists for apps considered to be medical devices [33-35], and health and wellness apps are left largely ungoverned. Furthermore, there are growing concerns over the credibility and evidence base of a range of mHealth apps that may be recommended in dietetic practice, such as weight management [36-41], diabetes [42-44], and physical activity [45,46] apps. Some studies have attributed the poor credibility of these apps to low health care professional involvement in app development [38,41]. Health care expert involvement in medical urology app development has been found to positively influence app downloads, suggesting that collaboration with health care specialists gives users greater assurance of the safety and credibility of an app [47]. As such, coinciding with our previous recommendations formulated on the basis of the COM-B model [19], involvement of dietitians and dietetic associations in the development as well as reviewing and subsequent endorsement of credible and reputable apps are necessary to enhance the confidence of the profession and their patients in recommending and using apps, respectively. Dietetic associations, such as Dietitians of Canada [48] and the US Academy of Nutrition and Dietetics [49] have headed the development and reviewing of credible apps for patient use and to support the dietetic profession.

According to the Technology Acceptance Model, perceived usefulness and perceived ease-of-use of a technology predicts users’ attitudes and behavioral intentions toward accepting the technology, thus affecting subsequent technology use [50]. Usability-related characteristics, such as ease-of-use, were not only valued among our responding dietitians but also were a positive contributor to users’ ratings within app stores [51] and were valued by mobile phone users in relation to wellness [23] and popular diet and weight loss apps [22,52]. An app requiring low effort to use is an imperative design consideration since both commercial and researcher-designed health apps typically experience a rapid decline in app use over time and low long-term retention [53-55]. Usability testing of a popular dietary app, MyFitnessPal, revealed users’ dissatisfactions over inconveniences in food logging and complex structure which resulted in a loss of interest in using the app [56]. Tools such as barcode scanners and image-based food logging could minimize user burden and allow patients to maintain compliance with tracking, while still providing valid measures of intake [57-59]. Furthermore, to improve adoption of apps, attractive user interfaces and simple to navigate designs should be included. In Web patient portal use, Web aesthetic simplicity (ie, cohesiveness, structure, and easiness to understand) was a significant antecedent variable to patients’ acceptance and use of patient portals [60]. App developers also need to engage in more user-testing during the development of apps and incorporate the service user feedback in an iterative design process to produce more dietitian- and patient- oriented apps.

When considering that diet and nutrition apps can automatically calculate energy, macronutrients, and micronutrient values from foods entered, apps present themselves as desirable tools to streamline and support dietitian-led dietary assessment. Yet, app developers should re-evaluate the quantitative approaches to dietary assessment currently implemented within apps, given that few dietitians consider apps as reducing the time for dietary assessment [19] and apps currently appear to lack of effectiveness in improving diet quality [61]. Instead, assessing the overall diet quality and interpreting and translating these dietary patterns into practical and meaningful food-based dietary advice would be more useful to both dietitians and their patients. Echoing conclusions drawn in the literature [36-38,45,62-64], our respondents recommended incorporation of a broader range of behavior change techniques, beyond self-monitoring. Notably, dietitians wanted apps to motivate patients and prompt them to practice health behaviors which could remind and encourage ongoing progress toward goals and encourage behavior change. Inclusion of automated motivational text messages or app push notifications have been found to improve physical activity [65], and when administered as part of a multicomponent mHealth lifestyle intervention, prevented weight gain and improved dietary behaviors [66].

As the health care system shifts away from the delivery of passive care to engaging patients as partners in their own health care, health apps present real-time opportunities to support and empower patients in making positive health behavior choices outside dietetic consultations. However, the absence of tailored goals and feedback is a major shortcoming identified in diet and nutrition apps for weight management [37,38,41]. The ability to input individualized goals within an app, such as those negotiated with dietitians would enable the tracking of more specific health behaviors relevant to the patient. There is only one known app—eaTracker developed by the Dietitians of Canada [67]—that supports the personalization of goals beyond generic pre-set targets of energy intake and weight loss. Additionally, providing personalized nutrition advice via mHealth technologies has been found to significantly improve selected dietary outcomes [68,69] and is an important consideration for developing effective apps. For example, remote and real-time delivery of daily tailored feedback messages significantly reduced energy and saturated fat intake, with changes maintained at 24 months [69]. Use of avatars might also be a method for personalizing the user experience in apps. They have been found to be a highly acceptable medium for modeling weight loss behaviors [70] and may engage and motivate users to change behaviors, such as promoting delayed gratification and dietary regulation [71] through embodying the patient’s ideal self.

Individuals have previously expressed that sharing health-app records with their health professional would be useful to their care [72]. Reports, however, indicate limited sharing of these records with health professionals or dietitians [72,73], possibly attributed to individuals’ perceptions that health professionals had little interest in their health-app records [72]. Contrary to patient beliefs, our responding dietitians wanted access to their patients’ health-app records, particularly to support the dietary assessment process. However, with few commercial mHealth apps having the capability to export user data [8,38], reviewing of patient progress with health apps has often been an infrequent and informal procedure for dietitians consisting of verbal discussion rather than direct viewing of the health-app data [19]. Enabling patients to synchronize, share, or export health-app data into their dietitian’s existing work systems could enhance the two-way communication between dietitians and their patients. The increased connectivity and access to records may create opportunities for dietitians to address patient lapses and deliver more dynamic behavioral strategies to support patient compliance with dietary recommendations. App developers also need to ensure that these functions integrate seamlessly into current practice workflow to avoid imposing addition burdens on time and effort for the dietitian in adopting new systems.

Although some respondents explicitly specified having no additional feedback regarding app design, it is not clear whether respondents who did not complete the open-ended questions had no further comments because they were in fact satisfied with the current state of apps, or whether they did not know what answers to provide and so left a blank response. The cross-sectional nature of this survey also poses the possibility of sampling biases, whereby greater willingness to respond to the open-ended questions may have come from more interested individuals and existing app users. However, the demographic profile of these respondents to the open-ended questions is comparable with that of the larger international survey, which was determined to be representative of the wider dietetic profession [19]. Furthermore, although adequate representation of the perspectives of nonusers is necessary, yet without experience in using existing health apps, the scope of suggestions provided by non–app users may be limited. If app developers perceive that the recommendations put forward regarding app design features already exist, they may be less inclined to develop apps further to support dietitian and patient needs. On the contrary, existing app users are likely to have richer and more feasible recommendations to guide and improve app development and design. It should also be noted that although understanding dietitian and user preferences may allow for more suitable apps to be designed for dietetic practice, this does not necessarily guarantee treatment effectiveness. Therefore, interventions studies are required to confirm which specific design features will provide the most support to dietetic practice and elicit significant effects on patient outcomes.

This study provides guidance to app developers of the features and characteristics of smartphone health apps valued by dietitians and highlights improvements for the design of health apps. In particular, dietitians asserted that apps should be credible and easy to use in order for them to more effectively support dietetic practice and dietitian’s recommendations of these apps. Greater collaboration between app developers and dietitians or their professional associations were also viewed as critical for achieving dietitian and patient-centered app design and integration into dietitian work systems. However, further investigation is required to determine the app features that offer the most support to dietitians in improving patient health outcomes.