Understanding and Predicting Medication Adherence in Clinical Trials

Traditional adherence methods such as using pill counts are unreliable, said Dooti Roy, PhD, U.S. Clinical Data Sciences Chapter Head at Boehringer Ingelheim (BI). Too often, researchers are left to “hope and pray missing pills have gone inside the patient and not somewhere else like the trash can,” she added.

To bring more certainty to adherence, Dr. Roy and her team at BI engaged with AiCure on several Central Nervous System pilot studies. The goals were to reduce missed doses and participant dropouts, and increase data quality. Learn more in “Leveraging Predictive Dosing to Reduce Clinical Trial Risk.”

AiCure Chief Medical Officer Rich Christie, MD, PhD, and Ken Getz, executive director of Tufts Center for the Study of Drug Development joined the webinar to discuss project specifics and research around predictive adherence.

The industry gold standard for approaching non-adherence has been direct patient observation, which is costly to deploy; however, the use of artificial intelligence (AI) reduces that cost and complexity. For over 10 years, AiCure has been providing remote assessments of adherence using direct video and audio collection. The AiCure App provides reminders and steps trial participants follow to complete and confirm dosing. Learn about the app here.

Unlike self-reported and clinic-reported methods which are executed by participants and sites respectively, AiCure confirms dosing via mobile computer vision. The app monitors the exact time and date of doses, identifies intentional non-adherence (using different pills or feigning ingestion), and enables supportive intervention.

Accessing this data in near real time enables better recruitment strategies, accurate predictions of non-adherence, better balance across treatment arms, and timely interventions by sites.
Begin your journey today to improve adherence with AiCure. Contact us to learn how.