Attendees of an online webinar debate about smartphone apps to help manage diabetes were initially very sure these apps are worth it, but the proponent for the con side managed to temper this enthusiasm, leading to a drop in enthusiasm at the end of the debate.
Initially, 91% of the 57 professionals who care for patients with diabetes agreed with the premise, “Are Diabetes Apps Worth It?” However, after the debate this dropped to 83%.
The webinar by the Healthcare Delivery and Quality Improvement group of the American Diabetes Association was held on January 16.
“We do want to see [diabetes] apps play a role — just in the right way,” summarized Josip Car, MD, PhD, arguing for the con side. Car is director of the World Health Organization Collaborating Centre for Digital Health and Health Education at Nanyang Technological University, Singapore,
He convincingly argued that studies to show that diabetes apps are effective have been “disappointing,” and that there are important issues regarding data privacy.
He did concede, however, that “there is no denying the usefulness of apps,” acknowledging that “they work for certain patients.”
Arguing for the pro side, endocrinologist David Ahn, MD, program director, Mary & Dick Allen Diabetes Center, Newport Beach, California, agreed that the use of apps might not translate into better patient outcomes.
And because there are hundreds of apps related to diabetes management, which are not regulated, it is a bit like the Wild West, as previously reported by Medscape Medical News.
On the other hand, apps can provide high-quality data that can be exchanged from patient to clinician, they enable telemedicine, and they can give patients access to a social support network, Ahn noted.
Both debaters admitted that although they were presenting opposing views they agreed that clinicians would benefit from a type of independent clearinghouse that would identify the current best diabetes apps.
In future, the “best apps would be paired to diabetes devices, supported by companies,” Ahn speculated.
Diabetes Apps Are Worth It
Ahn began by describing the different types of smartphone apps available for patients with diabetes.
The most “clearly useful” ones are tied to continuous glucose monitors (CGMs), he said, and showed examples of data from the InPen app (Companion Medical), which is approved by the US Food and Drug Administration (FDA), as well as data from the Dexcom Clarity app.
Meanwhile, standalone apps to help patients manage diabetes can assist with meal planning, track physical activity, monitor glucose, and help with insulin dose calculation, he noted.
And importantly, diabetes apps can also improve quality of life by providing patients with a social support network that gives them near real-time “wisdom of the crowd” responses to questions or concerns.
However, there are few clinical studies — mostly of bolus calculator apps — and efficacy outcomes have been inconsistent, Ahn conceded, possibly anticipating what his debating opponent might say.
A study published in 2016, he noted, reports that “Many mobile medical applications (apps) are available to consumers with diabetes, but only 14 currently have clinical outcomes data published in the peer-reviewed literature or have been cleared by the US FDA or have received a Conformité Européenne (CE) mark in Europe.”
“Why do apps lack good evidence?” Ahn asked rhetorically, and then suggested this may be because clinical trials are too costly for software developers, it is difficult to design a blinded study, and software/hardware changes so rapidly that it is difficult to track during any ongoing study.
Nevertheless, at present, apps facilitate and enable sharing of high-quality data that make telemedicine possible, and apps can give patients a better quality of life because they can have a social support network of their peers, he reiterated in his summary.
In future, Ahn concluded, apps could potentially be used as a digital therapeutic, for example, as a smart insulin bolus calculator.
Diabetes Apps Are Not Worth It; Insulin Dose Calculators Especially Bad
Car began by noting that, in 2017, there were more than 300,000 healthcare apps, and more than 200 new apps added daily.
However, “from excitement, reality kicks in,” he explained.
Most diabetes apps currently fall outside FDA regulations, and the quality and safety of these apps need to be investigated, something his research team has been doing.
A study from their group, published in 2019, which looked at 371 apps to help manage type 2 diabetes, found more than four in 10 did not alert users to hypoglycemia (41.2%) or hyperglycemia (41.6%), and more than 85% did not provide real-time support on blood glucose self-management.
In a second study, published back in 2015, they reported that of 46 insulin dose calculator apps, 91% lacked numeric input validation, 59% allowed calculation when one or more values were missing, and 67% provided an inappropriate output dose recommendation.
“The quality of insulin dose calculators in diabetes apps has not improved over time,” Car emphasized.
Moreover, in a third study, published in 2019, they found that “diabetes apps lack essential features to support self-management.”
That is, of 143 apps, 42% did not have medication reminder features; only 5.6% provided information about medication; 60% did not allow entry of medication-taking instructions; 83% did not have a feature to review medication adherence; and only two apps prompted users about complementary medicine.
Importantly, apps with insulin calculators have strikingly “inappropriate disclaimers,” Car stressed.
For example, the disclaimer of one app reads, “The entire risk as to the performance and quality of the app is with you.” Another reads, “Keep in mind that this is just a suggestion.”
Another important area of concern is that diabetes apps often do not take appropriate measures to keep data private, he noted.
The question is, “How can we develop app models that have both the rigor required for app safety and the nimbleness needed for sustainability?” Car wondered.
Rebuttals and Discussion
In his rebuttal, Ahn said that without bolus calculator apps, insulin on board is nearly impossible to determine; there is overuse of round numbers and mental math can be discriminatory.
In his rebuttal, Car said, “There’s good evidence for diabetes apps — it’s just that the evidence shows that apps are not yet amenable for proper use and self-care.”
In reply to a question about direct-to-consumer apps, Car said he teaches medical students that in future, when they prescribe metformin, they should prescribe a diabetes app at the same time — but “we’re not there yet,” he cautioned.
“In the interim, Veterans Affairs could do some curation of apps,” he suggested.