A new systematic review claims that most type 2 diabetes management smartphone apps have only minimal benefits, and are typically not worth the time, money, or effort. Even the most flattering studies tend to show that these applications offer only small improvements over usual care and that those benefits tend to evaporate over time.

The claim comes from Peterson Health Technology Institute (PHTI), a non-profit founded to evaluate the efficacy and cost-effectiveness of healthcare tech. The report (PDF) offers pages of analysis arguing that these management programs just aren’t very helpful — with a small number of exceptions.

“Technology should make healthcare more efficient,” says Caroline Pearson, executive director of PHTI. “But when we looked around, we felt like the way that this tech is being adopted is not meeting that goal.”

Skepticism About Diabetes Management Apps

There is a rapidly growing world of smartphone applications available for type 2 diabetes. The PHTI report focuses on a specific type: remote monitoring and lifestyle modification apps that require the use of a blood glucose meter, or fingersticks. The eight apps that PHTI researchers analyzed in detail: DarioHealth, Glooko, Omada, Perry Health, Teladoc (Livongo), Verily (Onduo), Vida, and Virta.

While these apps naturally vary, they have similar goals: to help people with type 2 diabetes achieve healthier blood glucose levels through the use of data, digital reminders, education, and behavioral coaching. A smartphone notification, for example, might remind you to test your blood sugar and may help interpret the data. The system will also prod patients to make healthier food and exercise sources. Some platforms grant online access to a clinician or diabetes educator (or, sometimes, an AI chatbot). Some integrate with other technology, such as smart scales or nutrition trackers. The end result is a steady stream of automated custom guidance to help people with diabetes make healthier decisions.

Typically, these platforms are purchased by organizations such as insurers or large employers rather than individuals, and then made available to members or employees with diabetes, often at little or no cost. Large organizations have a clear incentive to support interventions that improve the health of people with diabetes — it reduces the likelihood of more expensive (and disruptive) treatments such as kidney dialysis. But Pearson reports that, not long after the invention of the diabetes management app, the organizations purchasing access are increasingly skeptical of their efficacy.

“We spoke to hundreds of people throughout the industry to understand where the challenges were. One thing we heard over and over from healthcare purchasers, including health plans, employers, and providers, was that they were feeling inundated by sales pitches with claims for performance, but they really were not convinced that these were backed with evidence.”

The Results

The PHTI team assessed all of the published medical literature on these programs, whittling over a thousand papers down to the most relevant 69, and considered the results. They found that diabetes management apps improve outcomes, but not necessarily to a significant degree. According to a summary in a press release, the report found:

Small reductions in A1C (0.23 to 0.60 percentage points) compared to usual care. For reference, adding a new glucose-lowering drug such as a sulfonylurea lowers A1C by an average of more than 1.0 percentage points.
This effect was not durable. While most studies were only a few months long, those who followed patients for a year typically showed that the benefits in comparison to usual care disappeared.
No significant benefits for other metabolic conditions such as obesity, hypertension, or high cholesterol.
No evidence that glucose management apps are especially used by people who might have the most to benefit, such as those with high A1C levels or limited access to diabetes care.

The finding is especially striking when you consider that a great deal of this published data was organized and funded by the companies themselves: “We’re largely looking at the very same studies that the companies are touting, but we’re doing it through a specific frame,” says Pearson. “Because these are solutions that are adding to usual care delivery, we have evaluated their performance in comparison to usual care.” While diabetes is a progressive disease, oftentimes the control or placebo group in a clinical diabetes trial enjoys enhanced blood sugar control, perhaps because the trial setting serves as inspiration to take glucose management a little bit more seriously.

The bottom line for the large companies paying for these programs? They’re probably not worth the money. “Most of them are not delivering clinically meaningful benefits, and after you account for the price of the product, they’re generally increasing healthcare costs,” says Pearson.

Two Exceptions

The PHTI report notes that there are two exceptions to this otherwise rather pessimistic conclusion.

The first exception is “patients with high A1C (>9.0 percent) who are initiating insulin for the first time.” One small study, in particular, showed that diabetes management apps could have a powerful positive effect on this subpopulation of people who are not meeting glucose management targets and are about to start using insulin, a complex medication with a steep learning curve.

“There are moments in a person with diabetes’ care where they are establishing new habits,” says Pearson. “Those may be critical care moments where better self-management is possible with the use of digital tools.” But we are not currently deploying those tools in a focused way to best address those transitional moments.”

The second exception is the Virta Health diabetes management platform. Though Virta may use a similar smartphone interface to some of its competitors, it has one very big difference: It asks its users to commit fully to a radical diet change. Virta is one of the diabetes world’s biggest proponents of a very low-carb diet. The diet appears to work. Virta users were far more likely to enjoy significant glycemic improvement and related benefits, such as a reduction in the number of diabetes medications taken, and the improvements seemed to be long-lasting.

“In Virta’s success cases, the health effects are really impressive,” Pearson. “The question is, what portion of all people with diabetes will be able to stick to Virta’s program, or want to? We know it won’t be a solution for many people.”

The Bottom Line

Many people with diabetes enjoy and benefit from management apps. PHTI’s study casting doubt on these programs’ efficacy is limited only to certain kinds of apps: those usually purchased by larger health systems and employers rather than individual patients, and those that rely on old-school fingerstick blood sugar measurement. PHTI did not evaluate apps that use continuous glucose monitor (CGM) technology, nor those that focus on nutrition, exercise, sleep, stress, or other type 2 diabetes factors.

With all that said, maybe it’s smart to be skeptical that a smartphone application that communicates primarily through push notifications will be the impetus for change. The evidence suggests that these programs work best when they help patients navigate major treatment turning points, such as the use of a complex new medicine like insulin or the adoption of a dramatic dietary change.

“At the end of the day, these solutions rely on motivated behavior change, and that’s incredibly difficult,” Pearson says.

“No one should think that these programs are likely to improve their health outcome dramatically. If patients like using them, and they’re available freely, that’s fine, but they’re not a panacea.”