Diabetes mHealth Apps:
Can they be effective?
Ronak BRAHMBHATT a, Shadi NIAKAN b, Nishita SAHA d, Anukriti TEWARI c,
Ashfiya PIRANI d, Natasha KESHAVJEE a, Dora MUGAMBI a, Nasrin ALAVI a,
a InfoClin Inc, Toronto, ON
b University of Toronto, Toronto, ON
c York University, Toronto, ON
d University of Waterloo, Waterloo, ON
e Ryerson University, Toronto, ON
Abstract. mHealth apps are not being used. Over 45,000 mhealth apps are
languishing in mobile app stores. We evaluated over 200 diabetes mobile apps
found in the Apple and Google app stores using a framework that we recently
published. None of the apps met all 15 criteria identified by our framework. The
largest number of apps fell into the category of Type 1 diabetes blood sugar and
medication trackers. Other types of apps included educational apps such as recipe
apps, guideline dissemination apps, simple diabetes education apps, etc. There is a
need for more Type 2 apps and for all types of apps that are better integrated into
EMRs for more holistic care that can be prescribed by clinicians and monitored and
supported by the health care team.
Keywords. mhealth apps, diabetes, Type 1, Type 2, effectiveness, mhealth app
architecture, mobile apps, EMR integration.
Diabetes mellitus is a chronic disease whose prevalence is increasing rapidly
world-wide. The US spends US$245 billion and Canada spends CAN$12.5
billion annually on the direct and in-direct costs of diabetes [1, 2]. There is
strong support for the efficacy of self-management for diabetes, since the
treatment for the disease has a significant component of lifestyle change and
long-term adherence to medications .
One way of supporting patients in self-care and self-management is through
the use of technology, including mhealth apps [4, 5]. However, uptake of
mhealth apps is particularly poor . In a recently accepted paper, we developed
a reference architecture for the design and development of mhealth apps for
chronic disease management that addresses many of the barriers identified in the
literature . Based on this reference architecture, we developed screening
criteria for what should be in a mhealth app for optimal management of patients
Brahmbhatt R, Niakan S, Saha N, Tewari A, Pirani A, Keshavjee N, Mugambi D,
Alavi N, Keshavjee K. Diabetes mHealth Apps: Designing for Greater Uptake. Stud
Health Technol Inform. 2017;234:49-53. PubMed PMID: 28186014.
with diabetes. We were interested in learning whether diabetes mhealth apps
available in the apps stores meet the criteria that would allow them to function
well and solve previously identified barriers to mhealth app use.
We searched through the Apple iTunes and Google Play apps stores. Our search focused
on diabetes management and prevention apps using search terms such as diabetes,
diabetes management and diabetes tracker; 201 apps were identified. Using our mhealth
reference architecture , we devised 15 major functions that a diabetes management
app should perform (Table 1) because they enable care that is supported by guidelines
and by patient engagement best practices. Each major function also has about 5 to 10
descriptors; e.g., when “Patient Information” is a major app function, information such
as “date of birth” or “sex” is a descriptor. Every app was screened against these 15 major
Due to a limited budget, we did not download apps. Two individuals evaluated each
app based on a review of the product description, a careful review of the screen shots
provided and based on the reviews by current users. If an app had at least one attribute
of a major function, it was considered to have that function.
We had 9 reviewers, several of whom are clinicians (RB, KK, NA), students (AT,
AP, SN, NS) or research associates (DM, NK). Prior to screening initialization, we
carried out a calibration exercise with 5 apps. Reviewers had 84% agreement rate, which
was close to our target of 85%. Following the calibration exercise, we identified areas of
major discrepancy and standardized our approach. All reviewers were trained in the
standardized approach. Any new information was noted and a training program was
created which was used to train for the final screening. We also tailored our screening
sheet after the discrepancy resolution session. For final screening, each app was allocated
to two independent reviewers.
All data was collected in a Google Sheets. A list of apps and related major and minor
functions was collected in Airtable and shared with the reviewers to maintain blinding
for screening in Google Sheets. After screening was complete, all data was collated into
a single spreadsheet and descriptive analysis was performed using Google Sheets.
Of the 201 apps reviewed, none met all the functional criteria identified as being
necessary for the management of patients with diabetes (Table 1). A majority of apps
had some form of educational or recommendation component. Many were recipe apps
or provided basic information about diabetes. Most apps (approximately 90) were
replacements for paper journals or diaries, allowing patients with Type 1 diabetes to keep
track of their blood sugar readings and their insulin intake. A number of those apps also
provided basic education or made some minor recommendations, such as on exercise or
diet, which contributed to making education and recommendations the most prevalent
feature. A reasonable number of functions are available in about 25% of apps, including
patient information, notifications, lab results, physiological measurements and
messaging with their provider.
Surprisingly, very few apps (3) helped patients keep track of their risks factors, such
as smoking and alcohol consumption, or their vaccinations (0), which are recommended
for patients with diabetes to keep up with. Only 6 used risk scores to help patients connect
their daily care to future events so they could see the impact of their current actions on
potential events in the future.
Number of Apps
% of Apps
Health System Utilization
Local Device Integration
Messaging with Provider
Table 1. Features present in diabetes apps
Number of Apps
% of Apps
Table 2. Total scores and the number of apps that achieve them
Only a small number of apps were integrated with a local device such as a
glucometer or blood pressure cuff (28). These integrations are increasingly built into
smartphones by manufacturers and tapping into them is not particularly difficult .
Table 2 displays total scores and the number of apps that attained that score. An app
could obtain a maximum score of 15; none achieved it. A number of apps received a
score of 0. These were apps that were related to the concept of diabetes, but were not
meant for patients; e.g., diabetes conference apps, guideline apps for health professionals
and apps for educating nursing students about diabetes.
Only a small number of apps attained a score of 8. Of those, all had the following
features built in: patient demographics, medications, notifications, lab results and
physiological measurements. Other features which were more variable included,
messaging with physician, educational features, data export feature and integration with
a local device
It is not surprising that apps were not able to get past a score of 8. To achieve a
higher score requires interoperability with EMRs and the health system to obtain
information about risk factors, vaccinations and other important clinical endpoints. It
would be interesting to see interoperability kits built that would allow mobile health apps
to integrate with EMRs.
Overall, apps in the diabetes space appear to serve mostly educational or informational
purposes and blood glucose and insulin tracking for patients with Type 1 diabetes. The
blood glucose trackers seem to be well-received and many are well-liked in the app store.
Although these tools are valuable and play an important role in bridging the paper-
electronic gap, there is a great need for high quality tools that can be prescribed by
physicians and whose use can be monitored by a health care team. The tools need to
capture additional data that can be used to manage the whole patient and their disease,
not just a small part of the patient’s care which they are expected to manage themselves.
If we are to engage patients in self-management, we need to provide them with more
sophisticated tools within the context of the patient-physician relationship that ensures
that patients have the guidance they need to succeed in their own care on all disease
dimensions, not just one or two. The apps need to also provide health care teams the
tools they need to follow-up on patients, ensure lab tests are done, treatment is working
and that patients are adhering to diet, lifestyle and medication recommendations.
Given the high level of heterogeneity in the apps we found, we believe there is a
need for standard mhealth app certification criteria developed and regulated by a credible
organization. Any app that enters the market for use as a prescribed disease management
app should have at minimum the baseline functions outlined in our major functions list.
Certification can provide a basic level of trust to both physicians and patients, similar to
the way people trust drugs regulated by Health Canada. We also believe that a
standardized interoperability ‘kit’ for mhealth apps developed by EMR vendors would
go a long way to make important data for the care of diabetes available to patients and
make it easier for clinicians to analyze and trend.
Limitations of this study include: 1) due to budgetary constraints, we did not
download apps from the stores; 2) some vendors had poorer descriptions of their product
than others; 3) a very small number of apps were in languages that are not understood by
the people conducting the review; 4) we were not able to quantitate which apps are used
and which ones are not; 5) we did not include any patients in defining the criteria nor in
reviewing the apps.
In the next iteration of our project, we aim to include patients, download the apps
that got the highest scores, review them in greater depth, attempt to translate apps in
other languages and attempt to quantitate actual use of apps.
 The Cost of Diabetes. (2016). American Diabetes Association. Retrieved 11 September 2016, from
 An economic tsunami: the cost of diabetes in Canada Diabetes.ca. Retrieved 11 September 2016, from
 Lavelle, D., Zeitoun, J., Stern, M., Butkiewicz, E., Wegner, E., & Reinisch, C. (2016). Diabetes Self-
Management Education in the Home. Cureus, 8(7).
 Brzan, P. P., Rotman, E., Pajnkihar, M., & Klanjsek, P. (2016). Mobile Applications for Control and Self
Management of Diabetes: A Systematic Review. Journal of Medical Systems, 40(9), 210.
 Goyal, S., Morita, P., Lewis, G. F., Yu, C., Seto, E., & Cafazzo, J. A. (2016). The Systematic Design of a
Behavioural Mobile Health Application for the Self-Management of Type 2 Diabetes. Canadian journal
of diabetes, 40(1), 95-104.
 Research2guidance. (2015). App Developer Economics 2015. Retrieved December 16, 2015, from
 Chindalo P, Keshavjee K, Karim A, Brahmbhatt R, Saha N. (In press). Health Apps by Design: A
Reference Architecture for Mobile Engagement. International Journal of Handheld Computing Research.
 Diabeto iOS app is now available to download! https://blog.diabe.to/diabeto-ios-app-is-now-available-to-
download-72d93cc73d66#.fjg3hqeqx 2016. Web. 18 Sept. 2016