The Use of Mobile Applications Among Adolescents with Type 1 Diabetes: Results from Diabetes MILES Youth

Abstract

Background:

The use of mobile applications (“apps”) for diabetes management is a rapidly developing area and has relevance to adolescents who tend to be early technology adopters. Apps may be useful for supporting self-management or connecting young people with type 1 diabetes (T1D) with their peers. However, outside controlled trials testing the effectiveness of apps, little is known about app usage in this population. Our aim was to explore app usage among adolescents with T1D.

Methods:

Diabetes MILES Youth—Australia is a national, online cross-sectional survey focused on behavioral and psychosocial aspects relevant to adolescents with T1D. Associations between app usage and demographic, clinical, and psychosocial variables were analyzed using logistic regression.

Results:

In total, 425 adolescents with T1D responded to the app questions (mean age, 16 ± 2 years; 62% female; diabetes duration 7 ± 4 years). Overall, 21% (n = 87) indicated that they used an app for diabetes management. Of these, 89% (n = 77) reported carbohydrate counting as the most common purpose. Of those not using apps, 44% (n = 149) indicated that this was due either to no awareness of suitable apps or a belief that apps could not help. App usage was associated significantly with shorter T1D duration, higher socioeconomic status, and at least seven daily blood glucose checks.

Conclusions:

Only one in five respondents were using apps to support their diabetes management; most apps used were not diabetes specific. App users can be characterized as having a more recent T1D diagnosis, checking blood glucose more frequently, and being from a middle-to-high socioeconomic background.

Introduction

Type 1 diabetes (T1D) is a burdensome chronic condition that requires the individual to make daily complex treatment decisions to successfully manage their glycemic levels. Research demonstrates the importance of achieving optimal glycemic control during the adolescent years for reducing future risk of complications.^1,2 Only one in five adolescents (aged 13–20 years) meet their HbA_1c target, and adolescents with T1D have higher HbA_1c compared to any other age group with T1D.³ Furthermore, around 30%–40% of adolescents with T1D are “lost” to specialist diabetes care each year, most frequently during the transition from pediatric to adult services^4,5; many only re-engage with the health system after complications have developed.

Several meta-analyses have indicated the efficacy of smartphone or tablet applications (henceforth “apps”) in improving self-management among adults with diabetes.^6
–8 For example, the meta-analysis by Liu and Ogwu showed an average 0.39% reduction in HbA_1c from 12 studies that used mobile phone interventions.⁷ As such, the suggestion that interventions can support adolescents with T1D is of significant interest.⁹ Apps can support a broad range of self-management activities through features such as blood glucose diaries,¹⁰ medication reminders,¹¹ carbohydrate and insulin dose calculations,¹² incentivized blood glucose checking,¹³ and peer support.¹⁴ A common rationale in app intervention studies for adolescents with T1D relates to the increasing ownership of mobile smart devices among adolescents; between 2009 and 2013, mobile Internet usage tripled among Australians aged 14–17 years, with 69% owning a smartphone in 2013.¹⁵ However, to date, it appears that none of the apps specifically designed for and evaluated among adolescents with T1D⁹ are currently available for public use. Therefore, it is unclear which apps, if any, are being used by adolescents with T1D outside controlled intervention studies.

In 2015, it was estimated that over 165,000 general health-related apps (largely focused on diet and physical activity, e.g., MyFitnessPal and Fitbit) were available for download.¹⁶ A recent industry report indicates there are >1100 diabetes-specific apps available for use.¹⁷ This suggests that the opportunity exists for adolescents with T1D who own smart devices to use apps to support their diabetes management. Based on these statistics, we might expect diabetes-specific apps to feature prominently among adolescents with T1D who use smart device technology to support their self-management and daily coping. Due to the popularity of a number of diabetes-specific apps,¹⁷ we would expect these to be reported frequently among adolescents with T1D. However, this expectation is tentative, given that the above market research did not provide data specific to different diabetes types or age groups. In addition, this market research did not explore the use of other generic apps that may be used for additional diabetes support. For example, generic apps (e.g., Twitter or Facebook) may be used by diabetes organizations or informal groups of individuals to enable diabetes-specific social networking.^18,19

With the exception of a recent small-scale study of 50 adolescents at a single pediatric diabetes clinc,²⁰ no study has examined the use of apps among adolescents with T1D. A better understanding of app use among this group will add to our knowledge of what this group wants and needs to support their diabetes management. It will also highlight which apps are most popular and identify opportunities and gaps for future app development. The primary aim of the current cross-sectional national Australian study was to assess the usage of apps, both in terms of prevalence and purpose, in the support of T1D self-management among adolescents. A secondary aim was to determine factors associated with app use and reasons for nonuse. Furthermore, as several meta-analyses have found a beneficial effect of apps,^6
–8 we aimed to investigate associations between that diabetes-specific app usage and markers of better physical and mental health, including lower self-reported glycated hemoglobin (HbA_1c), more optimal self-reported self-management behaviors (including higher frequency of self-monitoring of blood glucose [SMBG]), reduced diabetes-specific distress, and reduced diabetes-specific eating problems.

Methods

The Diabetes MILES (Management and Impact for Long-term Empowerment and Success) Youth—Australia study was a national, online cross-sectional survey of Australian youth with T1D and parents/carers of youth with T1D. The study aimed to explore the psychological and behavioral characteristics of this group and to investigate self-reported diabetes management and health outcomes. Full details of the methods and sample characteristics are reported elsewhere.²¹ The current study focuses on the adolescent respondents only and on the subsample that completed questions about app usage.

Participants

Youth respondents were eligible to take part in the Diabetes MILES Youth—Australia study if they were aged between 10 and 19 years, self-reporting T1D, living in Australia, and able to complete an online survey in English without any assistance. Parents were also invited to take part in the survey, although they were not asked any app-specific questions. For a period of 8 weeks (August to October 2014), targeted and general recruitment strategies were used. Targeted recruitment involved sending postal invites to 5928 National Diabetes Services Scheme (NDSS) registrants (or their parents if the registrant was aged below 18) who had agreed to be contacted for research purposes. This number represented ∼59% of NDSS registrants in this age group. General recruitment involved advertising through flyers in diabetes clinics, social media postings, and notices in relevant online and hardcopy publications (e.g., Diabetes Australia magazines). Overall, there were 781 eligible youth respondents, representing ∼13% of the invitations sent out. Although this response rate was low, it should be judged relative to the difficulty of engaging this younger age group in online surveys and the lack of comparable data reported elsewhere.

Respondents aged 10–12 years were excluded due to an erroneous omission of a crucial app usage question for that age group (n = 220). Furthermore, any respondent that did not answer the app use and frequency questions was excluded (n = 53). An additional nine respondents were removed, as they did not indicate how regularly they used apps. This question was also used to recode respondents who answered “yes” to the app usage question, but then indicated that they used them “rarely” for the follow-up frequency question (n = 49). This recoding allows for a more accurate estimate of regular app usage. To allow for adjustment from the time of diagnosis, all respondents diagnosed within the previous 12 months were excluded from the current study (n = 34). Finally, any respondents who reported that they did not own a suitable device or had no access to Internet were also excluded (n = 40). This last exclusion criterion was based on the premise that including respondents who do not have the physical capability to use an app would be misleading as the aim is to find what variables are associated with app usage. This left a final sample of 425 eligible respondents.

Measures

The Diabetes MILES Youth—Australia study included a large number of measures, the full details of which have been described elsewhere.²² All reported data, including demographic and clinical, were obtained through self-report. The current study used a subset of the data set related to app usage (the app use items are shown in Table 1) and relevant outcome measures.

Table 1.

App Usage Questions and Responses

Question	Response % (n/N)
Do you use “apps” to help you with your diabetes?	Yes 21 (87/425)
If “No” then…
Why don't you use apps to help you with your diabetes? (pick all that apply)
I don't know how to use apps	1 (3/338)
I haven't found any good ones yet	38 (129/338)
I don't think apps could help me manage my diabetes	44 (149/338)
Other	17 (58/338)
What would you like in an app to help you with your diabetes? (pick all that apply)
Carb counting	44 (147/338)
Recording blood glucose levels	44 (147/338)
Insulin dose calculation	32 (107/338)
To track my physical activity	30 (100/338)
To track my weight	29 (99/338)
Social support	25 (83/338)
Monitor my moods/well-being	24 (80/338)
Hypoglycemia symptoms	23 (77/338)
Other	6 (20/338)
If “Yes” then…^a
How often do you use apps to help you with your diabetes management? (pick one) ^b
Several times a day	22 (19/87)
Once or twice a day	20 (17/87)
Several times a week	26 (23/87)
Once or twice a week	32 (28/87)
What do you use your app(s) for? (pick all that apply)
Carb counting	89 (77/87)
Recording blood glucose levels	30 (26/87)
To track my weight	10 (9/87)
Insulin dose calculation	9 (8/87)
To track my physical activity	8 (7/87)
Other	6 (5/87)
Monitor my moods/well-being	5 (4/87)
Social support	4 (3/87)
Hypoglycemia symptoms	2 (2/87)

If respondents answered “yes,” they were then prompted to provide the name(s) of the app(s) they used.

A fifth response option was available (“I rarely use them”), which was used as an exclusion criteria in the current analyses.

Socioeconomic status (SES) was determined from the respondents postcode and based on the Index of Relative Socio-Economic Advantage and Disadvantage (IRSAD), one of the Socio-Economic Indexes for Areas (SEIFA).²³ Deciles of the IRSAD were used, where lower deciles indicate more disadvantage and less advantage. For the purposes of the current study, these deciles were collapsed into three SES groups: low (1–3), middle (4–7), and high (8–10). Respondent remoteness, also determined from postcode, was based on the Australian Statistical Geography Standard (ASGS) Remoteness Areas classification 2011.²⁴ This scale consists of five categories: major city, inner regional, outer regional, remote, and very remote. Due to low numbers in the two remote groups, they were merged with the outer regional group, making three categories: “major city,” “inner regional,” and “outer regional/remote.”

The following clinical measures were included: diabetes duration (years), number of daily blood glucose checks, number of severe hypoglycemic events in the past year (defined as “conscious but unable to recognize symptoms, ask for help, or treat yourself because you were confused”), impaired awareness of hypoglycemia (Gold score ≥4),²⁵ and self-reported HbA_1c (in the past 3 months).

Diabetes-specific emotional distress was assessed using the 26-item Problem Areas in Diabetes-Teen (PAID-T) questionnaire. Items are scored on a 6-point Likert scale (1 = Not a Problem, 6 = Serious Problem; range = 26–156).²⁶ The sum of all items was used as the outcome measure, with higher total scores indicating greater distress. Finally, due to the increased presence among adolescents with T1D,^27,28 diabetes-specific eating problems were assessed using the 16-item Diabetes Eating Problem Survey (DEPS).²⁹ Items are scored on a 6-point likert scale (0 = Never, 5 = Always; range = 0–80). The sum of all items was used as the outcome measure, with higher total scores indicating more disordered eating behaviors.

Statistical analyses

Descriptive statistics were used to report the occurrence of app usage in this population and the reasons for and against usage. Demographic, clinical, and psychosocial characteristics were spilt by app usage and examined using independent samples t-tests and the nonparametric Mann–Whitney and chi-square tests where appropriate. Logistic regression analysis was used to evaluate the relationships between several demographic and clinical variables on the binary outcome of app usage (weekly app use/no app use) among adolescents with T1D who have access to smart devices and have been living with diabetes for at least 1 year. Potential correlates of app use were entered simultaneously into a logistic regression model if they were found significant in univariate analyses at the 20% level (Table 2). Multicollinearity among these variables was checked before running the regression analysis (variance inflation factor <3; tolerance >0.2), which was then followed by calculating standardized Pearson's residuals and Pregibon's delta beta influence and leverage statistics.³⁰ To avoid overfitting the regression model, the event per variable (EPV) was calculated. The EPV is taken as the ratio of outcome events over the number of predictor variables and is expected to be 10 or more.³¹ The Hosmer and Lemeshow³² test was used to assess model fit. All reported analyses were conducted using STATA V.14,³³ and statistical significance was set at P < 0.05 unless otherwise specified.

Table 2.

Characteristics of Survey Respondents by Current Diabetes App Usage Status (N = 425)

Variable	N missing	No app usage (N = 338)	Regular app usage (N = 87)	Test statistic	P
Sex, female	0	62.1 (210/338)	62.1 (54/87)	χ² (1) = 0.0001	0.992
Age, years	0	15.9 ± 1.9	15.2 ± 1.9	z = 3.17	<0.001
Socioeconomic status	10			χ² (2) = 6.44	0.040
High		42.9 (141/329)	50.0 (43/86)
Medium		37.4 (123/329)	41.9 (36/86)
Low		19.8 (65/329)	8.1 (7/86)
Remoteness	10			χ² (2) = 1.69	0.43
Major city		71.4 (235/329)	65.1 (56/86)
Inner regional		20.7 (68/329)	23.3 (20/86)
Outer regional/remote		7.9 (26/329)	11.6 (10/86)
Diabetes duration, years	0	7.5 ± 4.1	5.3 ± 4.2	z = 4.56	<0.001
Primary diabetes treatment	1			χ² (1) = 0.84	0.36
Insulin pump		53.1 (179/337)	58.6 (51/87)
Insulin injections		46.9 (158/337)	41.4 (36/87)
Blood glucose checks, per day	1			χ² (2) = 23.84	<0.001
3 or less		26.7 (90/337)	11.5 (10/87)
4 to 6^a		58.2 (196/337)	51.7 (45/87)
7 or more^a		15.1 (51/337)	36.8 (32/87)
Self-reported HbA_1c in past 3 months % (mmol/mol)	148	8.2 ± 1.5 (66.1 ± 16.1)	8.2 ± 2.1 (66.1 ± 22.9)	z = 0.589	0.556
Impaired awareness of hypoglycemia (Gold score ≥4)	1	16.6 (56/337)	11.5 (10/87)	χ² (1) = 1.38	0.24
Severe hypoglycemia,^b ≥1 in past 12 months	4	32.0 (107/334)	33.3 (29/87)	χ² (1) = 0.053	0.818
Diabetes-specific distress (PAID-T total score)	0	77.9 ± 31	69.8 ± 28	t = 2.32	0.011
Diabetes eating problems (DEPS total score)	1	20.2 ± 15.5	15.8 ± 12.7	z = 2.37	0.018

Values are mean ± SD or % (n/N).

Adjusted residual significant at P < 0.05, Bonferroni corrected.

Defined as “conscious but unable to recognize symptoms, ask for help, or treat yourself because you were confused.”

PAID-T, Problem Areas in Diabetes-Teen.

Results

Sample characteristics

Of the 425 eligible respondents, 62% (n = 264) were female and the mean ± SD age was 16 ± 2 years (min = 13, max = 19). Two percent (n = 8) of respondents identified as of Aboriginal or Torres Strait Islander origin. Eighty percent (n = 340) were in secondary school and 10% (n = 41) at university, with the remainder working, unemployed, or not reporting their employment status. Their average duration of T1D was 7 ± 4 years (min = 1, max = 18), with an average HbA_1c of 8.2% ± 1.6% (66.1 ± 6.0 mmol/mol).

App usage

Overall, 21% (n = 87) of respondents indicated that they used an app to support their diabetes management at least once a week. They reported a total of 21 different apps (respondents could report multiple apps). The most commonly used app was “CalorieKing,” an Australia-specific food database and weight management app, used by 66% (n = 57) of app users. The second most popular app was the “iBGStar” diabetes manager app, with 12% (n = 10) of app-using respondents using this. The “iBGStar” app can present real-time blood glucose information from the iBGStar blood glucose meter, providing additional functionality that may explain its apparent popularity. The remainder comprised miscellaneous apps of relatively low popularity (<10%). Of those using apps, 89% (n = 77) reported carbohydrate counting as the most common purpose, with 74% (n = 57) of these using “CalorieKing” for carbohydrate counting (Table 1).

Of the 80% (n = 338) not using apps, 44% (n = 149) indicated that this was due to their belief that apps could not help in the management of their diabetes, while 38% (n = 129) indicated they had not yet found any useful apps. These nonusers were asked what they would like in an app. The three most commonly endorsed responses were carbohydrate counting (44%; n = 147), recording blood glucose (44%; n = 147), and insulin dose calculation (32%; n = 107). One in five respondents (21%; n = 70) selected the above three functions together, suggesting a desire for multifunctional diabetes-specific apps.

Univariate and multivariate correlates of app usage

Univariate analysis (Table 2) showed mean age and diabetes duration to be significantly lower among regular app users than nonapp users (P < 0.001). The distribution of daily blood glucose checks differed significantly by app usage (P < 0.001), with a higher frequency of SMBG observed among respondents who used apps. Finally, both diabetes-specific distress and diabetes eating problems were significantly lower among app users (P < 0.02).

The seven variables that were significant at the trend level (P ≤ 0.2) in the univariate analysis (age, diabetes duration, SES, SMBG frequency, diabetes-specific distress, and disordered eating) were checked for collinearities. The largest variance inflation factor observed was 2.28 in the DEPS (tolerance = 0.48), which showed that collinearity was not an issue. Inspection of the regression diagnostics relieved two high-influence observations, both of which appeared to have legitimate values for variables. As there were no substantial changes with these observations removed, these cases were left in the analysis (Table 3).

Table 3.

Logistic Regression with App Usage as Dependent Variable (N = 414)

Variable	Odds ratio	95% confidence interval	P
Age, years	0.90	0.78–1.03	0.151
Diabetes duration, years	0.87	0.81–0.93	<0.001
Socioeconomic status
Low^a
Medium	3.20	1.29–7.92	0.011
High	2.78	1.14–6.76	0.024
Blood glucose checks, per day
3 or less^a
4 to 6	1.60	0.73–3.48	0.240
7 or more	4.72	1.98–11.26	<0.001
Diabetes-specific distress (PAID-T total score)	1.00	0.99–1.01	0.982
Diabetes eating problems (DEPS total score)	0.99	0.96–1.01	0.360

Eleven cases were deleted from analysis because of missing values; this did not reduce the event per variable below 10.

Reference group.

The model correctly classified 81% of the respondents, and the Hosmer–Lemeshow test was nonsignificant (χ ² = 5.08, df = 8, P = 0.749), indicating that the model fit the data well. The EPV was over 10 (87/8), indicating that the model could be estimated accurately. In the model, the factors associated with app usage were as follows: shorter diabetes duration (odds ratio [OR] 0.86, confidence interval [95% CI] 0.81–0.93 P < 0.001), high/middle SES compared to low (OR 2.78, 95% CI 1.14–6.76 P = 0.024; OR 3.20, 95% CI 1.30–7.92 P = 0.011), and at least seven blood glucose checks per day compared to three or fewer (OR 4.72, 95% CI 1.98–11.27 P < 0.001).

Discussion

In this first national survey of app usage among adolescents with T1D, one in five respondents reported using apps regularly to support their diabetes management, with the most popular app not being diabetes specific. The most common reason for app use was carbohydrate counting to support self-management of their T1D.

The most commonly reported app (used by 65% of users) was CalorieKing, a free generic app that can be used to support dietary decisions (weight loss or maintenance, but also for insulin dosing calculations). A separate diabetes-specific app made by the developers of CalorieKing (HEALTHeDiabetes) was not mentioned by our sample. This may be due to the fact that HEALTHeDIABETES is not a free download (cost AUD$9.99). A reasonable assumption is that the popularity of the CalorieKing app is, at least in part, due to its large Australian-specific food database, including over 19,000 foods with carbohydrate listings, and being free to download. However, CalorieKing is primarily a weight management tool and it presents weight loss messages within the app itself. Thus, its focus on weight and weight loss is concerning, given the increased frequency of disordered eating among adolescents with T1D.^27,28 However, it should be noted that DEPS scores were not significantly associated with app use in the multivariate analysis.

The main reasons given for not using apps were either a lack of knowledge about suitable apps or a belief that they would not help. This suggests that any future efforts to encourage app usage for diabetes self-management may need to raise awareness of the existence and potential benefits of app use. Among nonusers, 20% indicated an interest for multifunctional diabetes-specific apps. It is also noteworthy that approximately a quarter of nonusers expressed an interest in apps that could monitor exercise, weight, mood, hypos, and social support. Although this could also help to inform future app development, it should be noted that previous work in this area has highlighted such multifunctional apps as problematic in terms of their usability.³⁴

Both the univariate and multivariate analysis suggested a relationship between duration of T1D and app usage; a shorter duration was associated with increased use of apps. One possible cohort effect could be that respondents whose T1D was diagnosed before the rapid growth of mobile devices and apps (e.g., before Apple's App Store went live in 2008) would have had their initial diabetes education without reference to mobile app technology. In contrast, those whose T1D was diagnosed more recently, and the healthcare professionals they encountered, had the opportunity to incorporate mobile app technology into their initial self-management skill acquisition (e.g., using an app to find carbohydrate content for specific foods). This tentative interpretation is consistent with the number of respondents who indicated that they did not use an app due to either a lack of knowledge of suitable apps or belief that apps could not help. Future research could explore the role, if any, of app use in self-led and professional-led early diabetes self-management education.

The data also suggest a relationship between frequency of SMBG and app usage, with seven or more daily checks associated with over a fourfold increase in the odds of using an app. This finding is consistent with the large proportion of app users who list carbohydrate counting as the primary reason for using apps and would therefore be likely to be checking their blood glucose levels premeal. Again, the data suggest that app usage in this population is primarily function led, used to support the specific self-management skill of carbohydrate counting.

When considering the meta-analysis by Liang et al.,⁶ which showed mobile phone interventions led to significant improvement in glycemic control, the finding that app usage was not associated with lower self-reported HbA_1c was unexpected. Our inability to validate respondents’ self-report with clinical records of HbA_1c may be a factor here. However, research in this area is new, and the relationship between app usage and HbA_1c is uncertain.³⁵ Furthermore, despite a quarter of nonusers indicating an interest in using apps to support their mood/well-being, it is notable that very few app users (5%) were actually using apps to support their mood/well-being or for social support.

The data suggest that app usage is being driven by carbohydrate counting, and if this is the case, it is unsurprising that respondents in middle and high SES groups were more likely to be using an app to support their diabetes self-management.³⁶ Furthermore, it should be noted that this data set included an over-representation of respondents in the high SES group compared to the total invited sample,²¹ which may have increased its effect in the regression model.

Future research may look to compare the possible impact of dietary apps, which do and do not include weight loss messages, for use by adolescents with T1D. Given the popularity of CalorieKing in our sample, it seems reasonable to suggest that a diabetes-specific Australian food database app would be of value to an adolescent population. Furthermore, developing an app that complies with current privacy policies would also be important considering recent reports of diabetes app developers sharing sensitive health information with third parties.³⁷ However, it should be acknowledged that development of an alternative to CalorieKing would require significant resources. A more economical and sustainable approach might be the creation of an Australian open-source food database that would provide the necessary data for interested app developers (e.g., see the American DocGraph open food database project).³⁸

With regard to broad study limitations, it should be acknowledged that these data are subject to issues of self-report and selection biases.²² Young people from all states and territories took part. However, the predominance of female respondents and high SES level indicates that our findings may not be representative for Australian youth with T1D. Furthermore, the current study provides no information regarding how and why participants began using app(s), history of previous use, how long they have been using particular app(s), and who, if anyone, recommended the app(s) to them. These are important avenues for future research. Given the high frequency of CalorieKing usage compared to other food database apps, such as MyFitnessPal, it may be that specific app use is the result of common health professional recommendations or current adolescent trends for health-related app use more broadly. It would also be interesting to explore the extent to which app use is promoted by health professionals and the perceived role of apps in diabetes management by health professionals. A further limitation is the lack of data collected about self-management activities (such as carbohydrate counting) more generally. Unfortunately, it is unclear whether carbohydrate counting, as a self-management skill, was more common among app users; hence, the direction of the relationship is unclear. These cross-sectional findings also require further longitudinal exploration to identify the benefit, if any, that app usage has for adolescents with T1D in terms of self-management skills, biomedical outcomes, and psychosocial outcomes, such as diabetes distress and social support.

In conclusion, the Diabetes MILES Youth—Australia study has found that adolescents with T1D who use mobile apps to support self-management can be characterized as having a more recent T1D diagnosis, checking blood glucose frequently, and being from a middle-to-high socioeconomic background. It appears that a lack of awareness and doubt over their efficacy were the main reasons for not using apps in this way. This suggests that any future efforts to encourage app usage for diabetes self-management may need to raise awareness of the existence and potential benefits of app use. In addition to objectively assessing the benefits, future research needs to identify what drives app usage and how to support adolescents wanting to adopt this new technology to support their diabetes self-management.

Footnotes

Acknowledgments

The Diabetes MILES Youth—Australia 2014 Survey was funded by the National Diabetes Services Scheme (NDSS) Young People with Diabetes National Development Programme. The NDSS is an initiative of the Australian Government administered by Diabetes Australia. The authors thank the young people with diabetes and the parents who took part in the cognitive debriefing interviews and the Diabetes MILES Youth—Australia 2014 Survey. They acknowledge the generous advice of the NDSS Young People with Diabetes National Development Programme Expert Reference Group and the Diabetes MILES Youth—Australia Reference group for their consultation and contribution throughout the study and for their continued collaboration on the dissemination of findings.

Author Disclosure Statement

J.S. has received unrestricted educational grants, consultancy income, and sponsorship to host or attend educational meetings from Sanofi Diabetes (makers of iBG Star). J.L.B., in addition to serving on an advisory board, has received an unrestricted educational grant, consultancy income, and travel expenses from Sanofi Diabetes. C.H. has received an unrestricted educational grant from Sanofi Diabetes. No competing financial interests exist for the other authors.

References

Samuelsson

, Steineck

, Gubbjornsdottir

: A high mean‐HbA1c value 3–15 months after diagnosis of type 1 diabetes in childhood is related to metabolic control, macroalbuminuria, and retinopathy in early adulthood—a pilot study using two nation‐wide population based quality registries. Pediatric Diabetes, 2014; 15:229–235.

Svensson

, Eriksson

, Dahlquist

: Early glycemic control, age at onset, and development of microvascular complications in childhood-onset type 1 diabetes a population-based study in Northern Sweden. Diabetes Care, 2004; 27:955–962.

Wood

, Miller

, Maahs

, et al.: Most youth with type 1 diabetes in the T1D Exchange Clinic Registry do not meet American Diabetes Association or International Society for Pediatric and Adolescent Diabetes clinical guidelines. Diabetes Care, 2013; 36:2035–2037.

Kipps

, Bahu

, Ong

, et al.: Current methods of transfer of young people with type 1 diabetes to adult services. Diabet Med, 2002; 19:649–654.

Mater Hospital. Best Practice Guidelines for Health Professionals: For the Effective Transition of Young People with Diabetes from Paediatric to Adult Care. Brisbane: Queensland Health, 2008.

Liang

, Wang

, Yang

, et al.: Effect of mobile phone intervention for diabetes on glycaemic control: a meta‐analysis. Diabet Med, 2011; 28:455–463.

Liu

, Ogwu

: A meta-analysis of mobile health and risk reduction in patients with diabetes mellitus: challenge and opportunity. J Mob Technol Med, 2012; 1:17–24.

Whitehead

, Seaton

: The effectiveness of self-management mobile phone and tablet apps in long-term condition management: a systematic review. J Med Internet Res, 2016; 18:e97.

Barnaba

, Burr

: The role of mobile phones in adolescent T1DM: a review of the literature. J Pediatr Nurs, 2014; 29:387–389.

10.

Frøisland

, Årsand

, Skårderud

: Improving diabetes care for young people with type 1 diabetes through visual learning on mobile phones: mixed-methods study. J Med Internet Res, 2012; 14:e111.

11.

Hanauer

, Wentzell

, Laffel

: Computerized Automated Reminder Diabetes System (CARDS): e-mail and SMS cell phone text messaging reminders to support diabetes management. Diabetes Technol Ther, 2009; 11:99–106.

12.

Gilport Enterprises. RapidCalc Diabetes Manager. Retrieved from https://itunes.apple.com/au/app/rapidcalc-diabetes-manager/id400469609?mt=8 (accessed May 23, 2016 ).

13.

Cafazzo

, Casselman

, Hamming

, et al.: Design of an mHealth app for the self-management of adolescent type 1 diabetes: a pilot study. J Med Internet Res, 2012; 14:e70.

14.

HelpAround Inc. HelpAround Diabetes. Retrieved from https://itunes.apple.com/us/app/helparound-diabetes/id654991205?mt=8 (accessed May 23, 2016 ).

15.

Authority ACaM. Research snapshot—Aussie Teens online. Retrieved from www.acma.gov.au/theACMA/engage-blogs/engage-blogs/Research-snapshots/Aussie-teens-online (accessed November 11, 2015 ).

16.

Research2Guidance. mHealth App Developer Economics. 2015. The current status and trends of the mHealth app market November 2015.

17.

Research2Guidance. Diabetes App Market Report. 2014 January 2014.

18.

Advocacy

: DOC (Diabetes Online Community). www.inputdiabetes.org.uk/doc-diabetes-online-community (accessed June 6, 2016 ).

19.

Edwards

: Diabetes Counselling Online. www.facebook.com/diabetes.counselling (accessed May 30, 2016 ).

20.

St. George

, Delamater

, Pulgaron

, et al.: Access to and interest in using smartphone technology for the management of type 1 diabetes in ethnic minority adolescents and their parents. Diabetes Technol Ther, 2016; 18:104–109.

21.

Hagger

, Trawley

, Hendrieckx

, et al.: Diabetes MILES Youth–Australia: methods and sample characteristics of a national survey of the psychological aspects of living with type 1 diabetes in Australian youth and their parents. BMC Psychol, 2016; 4:42.

22.

Speight

, Hagger

, Trawley

, et al.: Diabetes MILES Youth Study 2014 Survey Report. Retrieved from the NDSS Resources for Young People with Type 1 Diabetes website: www.ndss.com.au/resources-for-young-people-with-type-1-diabetes (accessed May 24, 2016 ).

23.

Pink

: Information Paper: An Introduction to Socio-Economic Indexes for Areas (SEIFA): Canberra: Australian Bureau of Statistics (ABS), 2008.

24.

Pink

: Australian Statistical Geography Standard (ASGS): Volume 5—Remoteness Structure. Canberra: Australian Bureau of Statistics, 2011.

25.

Gold

, Macleod

, Frier

: Frequency of severe hypoglycemia in patients with type I diabetes with impaired awareness of hypoglycemia. Diabetes Care, 1994; 17:697–703.

26.

Weissberg‐Benchell

, Antisdel‐Lomaglio

: Diabetes‐specific emotional distress among adolescents: feasibility, reliability, and validity of the problem areas in diabetes‐teen version. Pediatr Diabetes, 2011; 12(4 pt 1):341–344.

27.

Jones

, Lawson

, Daneman

, et al.: Eating disorders in adolescent females with and without type 1 diabetes: cross sectional study. BMJ, 2000; 320:1563–1566.

28.

Young

, Eiser

, Johnson

, et al.: Eating problems in adolescents with Type 1 diabetes: a systematic review with meta‐analysis. Diabet Med, 2013; 30:189–198.

29.

Wisting

, Frøisland

, Skrivarhaug

, et al.: Psychometric properties, norms, and factor structure of the diabetes eating problem survey-revised in a large sample of children and adolescents with type 1 diabetes. Diabetes Care, 2013; 36:2198–2202.

30.

Pregibon

: Logistic regression diagnostics. Ann Stat, 1981; 9(4):705–724.

31.

Peduzzi

, Concato

, Kemper

, et al.: A simulation study of the number of events per variable in logistic regression analysis. J Clin Epidemiol, 1996; 49:1373–1379.

32.

Hosmer

Jr. , Lemeshow

: Applied Logistic Regression. New York: John Wiley & Sons, 2004.

33.

Stata Statistical Software: Release 14. [computer program]. Version: StataCorp LP., 2015.

34.

Arnhold

, Quade

, Kirch

: Mobile applications for diabetics: a systematic review and expert-based usability evaluation considering the special requirements of diabetes patients age 50 years or older. J Med Internet Res, 2014; 16:e104.

35.

Kumah-Crystal

, Hood

, Ho

Y-X

, et al.: Technology use for diabetes problem solving in adolescents with type 1 diabetes: relationship to glycemic control. Diabetes Technol Ther, 2015; 17:449–454.

36.

Valenzuela

, Seid

, Waitzfelder

, et al.: Prevalence of and disparities in barriers to care experienced by youth with type 1 diabetes. J Pediatr, 2014; 164:1369–1375.e1361.

37.

Blenner

, Köllmer

, Rouse

, et al.: Privacy policies of android diabetes apps and sharing of health information. JAMA, 2016; 315:1051–1052.

38.

Medstartr. DocGraph Open Food Database. Retrieved from www.medstartr.com/projects/232-docgraph-open-food-database (accessed May 26, 2016 ).

The Use of Mobile Applications Among Adolescents with Type 1 Diabetes: Results from Diabetes MILES Youth—Australia