Report: Data on diabetes apps' clinical impact scarce, inconsistent

By Dave Muoio
Share

Of the hundreds of diabetes self-management apps currently available, very few have been evaluated with health outcome-focused studies, according to (AHRQ). Among this subset, even fewer have been shown to actually help diabetes patients significantly improve their HbA1c levels.

Prior research has shown that 16 percent of the 318,000- health apps available worldwide in 2017 were related to diabetes, AHRQ investigators wrote. The functions and features of these apps are highly variable, and there is little reliable information as to their clinical validity.

“In early 2017, the AHRQ Scientific Resource Center (SRC) identified several systematic reviews that could potentially address the [effectiveness of apps for diabetes self-management]. However, these reviews and others typically apply one of two strategies: they look exclusively at the published literature and include apps that are not available to consumers, or they review features of commercially available apps and do not consider whether the apps have evidence of clinical efficacy,” the AHRQ investigators wrote. “We determined that a marriage of these two strategies could address both research and consumer needs.”

To do so, the investigators searched Ovid/Medline and the Cochrane Database of Systematic Reviews for any recent and relevant technology assessments or systematic reviews pertaining to diabetes apps. The team also conducted online searches for any other recent primary studies that may be applicable. Data from those that qualified for inclusion were reviewed and extracted by the team, with the intention of identifying the specific characteristics of each offering, what kinds of patient outcomes had been associated with use of each app, and any harms or costs that have been associated with these specific technologies.

The investigators identified 15 qualifying pieces of literature, which included analyses of only 11 unique diabetes apps (two of which had separate paid and free versions of their service).

While the team deemed the population samples to be generally representative to most Type 1 or Type 2 diabetes patients, they noted the relatively short duration of the identified studies (range, two to 12 months) when compared to the lifelong duration of diabetes. Among these, eight apps demonstrated improvement in at least one clinical outcome compared to controls, and just five apps had data supporting a clinically meaningful HbA1c reduction of 0.5 percent or greater (Diabeo Telesage and Glucose Buddy for Type 1, BlueStar, WellTang, and Gather Health for Type 2). Only two of the examined apps were deemed to have an “acceptable” usability score by the investigators (Glucose Buddy and BlueStar), with the researchers noting that the majority of the apps for which clinical data were available were outshone by the usability and aesthetic presentation of the Google and Apple storefront’s top diabetes app results.

Of similar concern to the team was the quality of the data, and the general lack of consistency with how they were being reported across all of the identified studies. Driving this was the limited information on randomization, allocation, masking, and drop-out analysis in several studies; varying measures of app success; and study designs that could confuse any positive effects of an app with other added interactions. Between these issues and the low overall number of apps with appropriate clinical investigations, the investigators wrote that they lacked the data to empirically assess and compare the relationship between costs, features, efficacy across apps.

“Although there is limited evidence that commercially available mobile apps improve diabetes-related outcomes, patients are downloading and using them anyway. Strong evidence can help people make informed choices, but when evidence is limited, patients who use these apps are essentially experimenting on themselves,” the investigators wrote. “Considering this, clinicians should consider asking their patients if they use apps in their self-management, and determine if the information provided by these apps adheres to current guidance for diabetes self-management. Patients should be aware that there is little evidence supporting the effectiveness of these apps, and should be wary of claims that these apps will improve their outcomes if not supported by evidence.”

In many ways, these results echo the findings of another overview of systematic reviews that was published yesterday in Nature Partner Journals: Digital Medicine. Here, researchers from the Centre for Research in Evidence-Based Practice at Bond University in Queensland, Australia reported they could find only 23 randomized controlled trials of available mobile health apps, and that fewer than half showed a positive health effect resulting from the studied app.