Randomized controlled study shows mobile app improves glucose control in Type 1 Diabetes #Med2

[This is a preview of some of the exciting mHealth research being presented at the upcoming Medicine 2.0 Congress in September. This abstract and others are candidates for the iMedicalApps-Medicine 2.0 mHealth Research Award]
by: Morwenna Kirwan, MHMSc, Corneel Vandelanotte, PhD, Mitch J Duncan, PhD, Andrew Fenning, PhD 
iMedicalApps-Medicine 2.0 Award Finalist
Persistently poor glycemic control in adult type 1 diabetes patients can lead to significant damage to the cardiovascular, renal, neural and visual systems. In many patients HbA1c levels are unsatisfactory, with levels consistently above 8.0%.
Only a small number of prototypes of type 1 diabetes smartphone applications have been developed and tested in clinical settings even though there are a plethora of low-cost and free self-management diabetes applications currently available.
The aim of our study was to examine the effectiveness of a free smartphone application combined with text-message feedback to improve glycemic control and other diabetes related outcomes in adult patients with type 1 diabetes in a two group randomized controlled trial.

Research Design and Methods

Design

• The study utilized a two-arm randomized controlled trial method including a usual care and intervention arm. It was conducted with the assistance of a Certified Diabetes Educator.

Participants

• Participants were recruited nationally by means of sending an invitation letter to type 1 diabetes patients registered with Diabetes Australia in New South Wales (n=3809) and Queensland (n=3207), as well as an advertisement in a type 1 diabetes newsletter emailed to over 5,000 recipients and promotion in an online community forum.

Study inclusion criteria were:

1. aged 18 to 65 yrs
2. diagnosed with type 1 diabetes > 6 months
3. HbA1c >7.5%
4. treated with multiple daily injections or insulin pump
5. own an iPhone
6. not currently using a diabetes self-management app.

There was no face-to-face contact between patients and researchers during the study. Participants lived throughout Australia. Patients and their primary diabetes health care practitioner were required to provide informed consent.

Intervention

Patients in both groups were asked to continue with their usual care which included a visit to their primary diabetes health care practitioner every 3 months. Patients assigned to the intervention arm were given instructions to download the smartphone application Glucose Buddy (developed by SkyHealth LLC), a free diabetes self-management iPhone application that allows users to manually enter blood glucose levels, diet, insulin dosages, other medications and physical activities.
Glucose Buddy users can also view their data on a personalized graph and email their information. Glucose Buddy has been reported to be the most downloaded diabetes management software on iOS, with downloads in excess of 100,000.
There was no minimum amount of logging in required and intervention patients were able to utilize the accompanying website to keep track of diabetes parameters at their discretion. The information logged into Glucose Buddy was reviewed by a Certified Diabetes Educator via a web interface on a weekly basis.
All patients in the intervention arm were sent a minimum of one personalized text message communication per week for the first 6-months of the study. These text messages were personalized for each patient and were based on the information logged in the Glucose Buddy app – this advice was of the same standard that would be given in a face-to-face consultation with a Certified Diabetes Educator.
Those intervention patients that did not log into Glucose Buddy were also text-messaged to see how they were doing. It was optional for patients to respond to the text-messages and converse with the Certified Diabetes Educator. At the 6-month timeframe all text-message communication ceased.

Measures

All measures were collected at baseline and every 3-months over the 9-month study period for both groups (making 4 time points in total). The primary outcome measure was change in glycemic control assessed by HbA1c. The secondary outcome measures, being diabetes-related self-efficacy (perceived ability to engage in the activities related to diabetes self-management), self-care activities and quality of life, were collected via a web-based survey emailed to patients every 3 months.
Intervention participants’ level of engagement was also measured in terms of text-message communications with the Certified Diabetes Educator and utilization of the app.

RESULTS

Participant Characteristics

197 adults with type 1 diabetes indicated their interest online or via phone to the research team and were assessed for eligibility (Figure 1), with 125 excluded for not meeting inclusion criteria. Seventy-two individuals were randomized to the two groups. There were significantly (p =0.02) more females (75%) in the control group with no other baseline differences observed.
Mean age of patients was 35.20 ± 10.43 years and patients had been diagnosed with type 1 diabetes for a mean of 18.94 ± 9.66 years. The intervention group had a significantly higher (p=.02) baseline HbA1c (9.08 ± 1.18) than the control group (8.47 ± 0.86), and reported a healthier diet (3.56 ± 1.70 healthy days per week for the intervention group versus 2.60 ± 1.98 days for the control group, p=.03).
Dropout was 26% (11 males, 8 females) with logistic regression analysis revealing no significant difference in age, gender, diabetes duration and baseline HbA1c among those that completed the study and those that were lost to follow up.

Intervention Effects

The intervention group had a significant decrease in HbA1c (-1.10 ± 0.74, p Engagement by Intervention Patients.
Over the 6-month intervention period, the Certified Diabetes Educator sent in a total of 1,714 text-messages, which equates approximately to 2 text-messages per patient per week.
Patients sent in a total of 559 text-messages to the Certified Diabetes Educator over the 6-month period.
Using the Glucose Buddy application, patients logged 24,720 diabetes parameters in total: 54% of the logs related to blood glucose levels, 31% to insulin, 12% to diet, and 1% to exercise.

Conclusion

In adjunct with usual care, use of the Glucose Buddy application combined with weekly text-message feedback from a Certified Diabetes Educator led to a significant decrease in HbA1c in comparison to a control group receiving only usual care.
Study limitations include:

• This study was a randomized controlled trial with a small sample conducted over a short duration
• There were differences in glycemic control and gender between groups at baseline
• Despite these limitations we did find that integrating a smartphone application into secondary care was effective in improving glycemic control in patients with type 1 diabetes
• Our findings can be applied to adults with poorly controlled type 1 diabetes that own a smartphone, though larger studies over a longer duration need to be conducted to validate our findings.

Researchers

The Authors are researchers at the Institute for Health and Social Sciences Research, Central Queensland University, Rockhampton, Australia.

• This study was funded by Central Queensland University, Australia
• The authors thank Certified Diabetes Educator Veronica Mills and SkyHealth, the developers of Glucose Buddy application and website
• C Vandelanotte is supported by a National Health and Medical Research Council of Australia (#519778) and National Heart Foundation of Australia (#PH 07B 3303) postdoctoral research fellowship
• M Kirwan is supported by a Queensland Government, Department of Tourism, Regional Development and Industry, SmartFutures PhD Scholarship.