Keywords

1 Introduction

Diabetes Self-Management.

Diabetes mellitus is a group of metabolic diseases characterized by chronic hyperglycemia [1]. In 2014, there were 29.1 million Americans with diabetes, including 8.1 million people who were undiagnosed [2]. Diabetes self-management education is a process of educating patient the knowledge of diabetes to improve their self-management behaviors [3]. Several studies suggested through appropriate diabetes self-management, diabetic patients can improve the long-term health outcomes [2, 4,5,6,7].

Diabetes Self-Management Applications (Apps).

Mobile health is defined as “mobile computing, medical sensor, and communications technologies” that can improve chronic disease care outside hospitals [8]. In recent years, there has been a rapid development of health apps for smartphones and tablets. Based on the report of IMS Institute for Healthcare Informatics, there were almost 165,000 health apps in 2015 [9]. With this increase in health apps there has also been an increase in the number of apps designed specifically for diabetic patients. In 2013, researchers searched Google Play, App iTunes, and BlackBerry World app stores. They found 1,812 diabetes-related apps [10]. DSM apps are tools on smartphones or tablets designed to help diabetic patients to achieve behavioral changes [11]. DSM apps provide functions, including monitoring carbohydrate intake, exercise, and blood sugar level.

According to the International Standards Organization (ISO 9241-11), usability is defined as “the extent to which a product can be used by specified users to achieve specified goals with effectiveness, efficiency, and satisfaction in a specified context of use” [12]. Usability testing is defined as “a systematic way of observing actual users trying out a product and collecting information about the specific ways in which the product is easy or difficult for them” [13]. Zhang et al. indicated “Usability testing is a mandatory process to ensure that a mobile application is practical, effective, and easy to use, especially from a user’s perspective [14]. The results of less attention to usability may include “frustrated users and decreased efficiency coupled with increased cost [15].

However, limited research and review have been conducted on the functionality and usability evaluation of these apps. The purpose of this review was to systematically review the studies related to the functionality and usability evaluation of diabetes mobile apps.

2 Method

Data Sources.

In October 2016, we searched three electronic databases: PubMed, Scopus, and Cochrane. The search terms used were “mobile app”, “mobile application”, “diabetes”, and “evaluation”. We limited the articles to those that were written in English and published from January 1, 2006 to October 4, 2016.

Inclusion Criteria.

The inclusion criteria were any research related to evaluation on functionality or usability of diabetes applications.

Study Selection and Data Extraction.

We reviewed the titles and abstracts of identified articles. Based on the inclusion criteria, eligible articles were included for full-text review. We collected data from eligible articles, the object of study, diabetes types, study sample, sample size, numbers of apps used in the study, evaluation types (functionality or/and usability), evaluation methods, findings of evaluation, app types (native or web-based app), app platforms (e.g., iOS, Android), device, and evidence-based guidelines used for developing the app.

3 Results

Study Selection.

Out of 200 articles, we identified 20 articles as eligible for our systematical literature review (Fig. 1). There were 13 articles from PubMed and 7 articles from Scopus.

Fig. 1.
figure 1

Study selection process

Full size image

Description of Included Studies.

Table 1 shows the summary of functionality and usability studies on diabetes mobile applications. The publication years ranged from 2008 to 2016. There were seven articles focused on type 1 diabetes [16,17,18,19,20,21,22], two articles focused on type 2 diabetes [23, 24], two articles focused on both type 1 and type 2 diabetes [25, 26], and nine articles did not report specific diabetes types [11, 27,28,29,30,31,32,33,34]. Thirteen studies evaluated one in-house mobile app [16, 17, 19,20,21, 23,24,25,26,27,28,29,30] and seven studies evaluated two or more apps developed by others [11, 18, 22, 31,32,33,34]. The number of apps for evaluation ranged from one to 656. Sample size for evaluation ranged from five to 123 subjects. The study sample types included adolescents [16,17,18], adults [19, 20, 24,25,26,27,28, 30], elderly people [23].

Table 1. Summary of functionality and usability studies on diabetes mobile applications. DMT: Diabetes Mellitus Type, F: Functionality, U: Usability, N: Native app, W: Web app, NR: No reported, N/A: Not applicable, ADA: American Diabetes Association, JDS: Japan Diabetes Society.
Full size table

Types of Evaluation and Evaluation Methods.

With regard to types of evaluation, only one study reported solely on functionality [21], seven studies reported usability [17, 19, 23, 25, 28,29,30], and twelve studies reported both functionality and usability [11, 16, 18, 20, 22, 24, 26, 27, 31,32,33,34]. The methods used for evaluations included survey [16, 25, 30], interview [17, 18, 24], laboratory testing [17, 20, 21, 27], user testing [18, 19, 22, 23, 25, 26, 28, 32, 34], questionnaire [18,19,20, 23, 26,27,28], expert evaluation [11, 22, 29, 31,32,33], and heuristic evaluation [22, 28, 32, 34].

Apps Types and Platforms.

Out of thirteen studies that evaluated in-house mobile app, two apps were implemented as web-based mobile apps [29, 30]. The other eleven studies did not report app types [16, 17, 19,20,21, 23,24,25,26,27,28].

With regard to the platforms, eleven studies used Android platform [11, 16, 18, 19, 21, 22, 24, 25, 27, 29, 33], six studies used iOS platform [11, 17, 22, 31, 32, 34], two studies used Windows platform [20, 28], one study used Blackberry platform [22], and one study used the Pebble (smartwatch) platform [26]. Two studies did not report the platform types on which the apps developed [23, 30].

Diabetes Guidelines for Developing Apps.

Even though evidence-based guidelines guide effective app development [35], only two studies reported the evidence-based guidelines used when the apps were developed [24, 25]. The other eleven studies which developed apps did not report the guidelines for app development [16, 17, 19,20,21, 23, 26,27,28,29,30].

4 Discussion and Conclusion

This study showed only six studies reported on usability. Usability plays an important role helping users to complete a task successfully with minimal cognitive load [36, 37]. This study revealed that only two studies provided the information about the diabetes guidelines for app development. It is not certain whether the studies used diabetes guidelines but failed to report them, or they did not consider using guideline when developing DSM apps. This result is consistent with the findings from several functionality studies on diabetes apps. Our team had conducted functionality analysis of current diabetes apps to investigate the presence of evidence-based guidelines while developing apps [38]. There were 168 diabetes eligible apps from iOS and Google Play included in the study. The functionality of each app was coded according to the validated AADE7 Self-Care BehaviorsTM by the American Association of Diabetes Educators. The results showed very few apps followed the AADE7 Self-Care BehaviorsTM guideline. Similarly, Chomutare et al., analyzed the functions of 101 DSM apps from Apple iPhone, Google Android, BlackBerry, and Nokia Symbian [35]. The study found that features of diabetes apps on the online market did not follow evidence-based guidelines either [35].

This study showed that most studies used triangular methodologies for functionality and usability evaluation, which was encouraging because mixed methods can reveal more comprehensive usability problems that a single method may not detect [39]. We also found lack of consistency in reporting evaluation findings that some studies did not provide scientific details on study subjects, such as level of prior mobile app experience, health literacy, and education level, which would influence the results evaluation outcomes. For future app development and evaluation, systematic and consistent reporting guideline including the methodical and scientific details should be used to inform research community.

Limitations of the Study.

Our study has limitations. First, we did not consider the clinical outcomes of the diabetes apps because we were interested in evaluation methods. The clinical outcomes may provide additional information on the intervention effect that may be influenced by the functionality or usability. Second, we only included studies that were published in English in our review. Inclusion of the literature published in a language other than English should have expanded the pool of literature.