As the title of this blog suggests, I tend to be skeptical about things. One of the things I am most skeptical about these days is the whole realm of digital health/mobile apps, etc. People may like them, but will they actually improve health outcomes. A new study in Journal of the American Medical Association Network Open suggests that a mobile app for diabetes care management did have positive effects. For a patient with diabetes, there are always medications to take and tracking blood sugar levels is important. For that reason, a number of mobile apps to encourage adherence to treatment regimens have been developed. In the current study, the Kaiser health system in northern California offered members free access to an app that tied them to information in their health records. Patients with diabetes who had at least one prescription for a diabetes medication other than insulin were included in the study and the primary outcome was adherence to those medications. Note that the measure of adherence was prescription-filling behavior, not actual use of the drugs, which is harder to monitor but might be inferred from prescription-filling activity. Blood sugar levels, as recorded in the EHR, were also analyzed. Patients were categorized as never accessing the portal, doing so by computer only, by mobile app only or by computer and mobile app. Once people signed up, the presumption was that they continued using the portal. The researchers attempted to account for certain characteristics and factors in the analysis, but you have to wonder if they adequately adjusted for the likelihood that portal users were more engaged in their health and health care.
Over 90,000 patients were involved in the analysis. During the study period, roughly from 2015 to 2017, the percentage of people who never accessed the portal declined from around 35% to 24.5%. Most people used both the computer and mobile app to access the portal. Patients who began using the portal showed a modest increase in medication adherence. Patients who had higher blood sugar levels showed greater increases in adherence behavior. These patients also showed a statistically significant reduction in those blood sugar levels. Patients who added mobile access to pre-existing computer use of the portal also showed greater reductions in blood sugar levels. While potentially encouraging, I am not sure this basically observational study actually suggests that the use of the portal, by computer or by mobile app, is the cause of the improvement in adherence or blood sugar levels.