If You Don’t Take Your Stroke Meds, This App Will Tell on You

A smartphone app that verifies whether patients are appropriately taking oral anticoagulation after an ischemic stroke improves adherence to prescribed therapies, according to results from a small, randomized study.

Daily monitoring with the artificial intelligence (AI)-based app was associated with 100% adherence based on plasma drug concentration levels compared with a rate of only 50% in control patients, researchers led by Daniel Labovitz, MD (Montefiore Medical Center, New York, NY), report in a study published recently online in Stroke.

An advantage also was seen in the intervention group when adherence was based on pill counts (97.2% vs 90.6%).

“Patients, some with little experience using a smartphone, successfully used the technology and demonstrated a 50% improvement in adherence based on plasma drug concentration levels. For patients receiving direct oral anticoagulants, absolute improvement increased to 67%,” the authors write. “Real-time monitoring has the potential to increase adherence and change behavior, particularly in patients on direct oral anticoagulant therapy.”

The direct oral anticoagulants—also called non-vitamin K antagonist oral anticoagulants (NOACs)—carry some advantages over warfarin, including a wider therapeutic window, fewer drug and food interactions, and fixed dosing that precludes the need for routine blood tests. But the move away from regular monitoring and the shorter half-lives of the newer agents have raised concerns about medication adherence.

Prior interventions to boost compliance with treatment have included counseling, educational programs, text messages, and electronic monitoring, but those have all been limited by the inability to verify that patients are actually taking their medications, Labovitz et al say. Directly observed therapy has been used to optimize adherence to treatment for other conditions, like tuberculosis and HIV.

Harnessing New Technologies

In this study, the investigators evaluated that concept using AI instead of human observers. The app (AiCure; New York, NY), which can be used on any smartphone, includes HIPAA-compliant facial recognition to identify patients, drug recognition, and ingestion confirmation. The software also provides medication reminders and dosing instructions to patients, as well as automated text messages to clinic staff when doses are missed, late, or incorrect.

The researchers randomized 28 patients with recently diagnosed ischemic stroke who were taking warfarin or one of the following NOACs: dabigatran (Pradaxa; Boehringer Ingelheim), rivaroxaban (Xarelto; Janssen Pharmaceuticals), or apixaban (Eliquis; Bristol-Myers Squibb). Those in the intervention group received smartphones with the app for daily monitoring and those in the control group did not.

According to visual confirmation of drug administration using the app, 90.5% of patients in the intervention group were adherent through 12 weeks. Pill counts and plasma sampling confirmed better adherence compared with the controls.

A substudy of patients taking one of the NOACs, who generally had worse adherence compared with their warfarin-treated counterparts, provided similar results but with a larger absolute advantage for those using the app.

“Suboptimal adherence across all [NOACs] confirms previous findings, although this study suggests that adherence might be lower than previously recognized,” the authors write. Routine laboratory testing with warfarin, in contrast, showed value in ensuring adherence, they add. 

“Consistent use and general likability of the AI app over 12 weeks underscores the possibility of harnessing new technologies to optimize adherence in patients on [NOACs], for whom routine laboratory tests are not good indicators of adherence,” they conclude.