Using Artificial Intelligence on Mobile Devices to Measure and Maximize Medication Adherence in CNS Trials
In this poster you will learn:
- Twenty to thirty percent of CNS trials fail because participants are not dosing correctly. The issue of adherence has received increasing attention as researchers find ways to reduce the high failure rate of clinical trials, particularly in CNS.
- Nonadherence, defined as failure to take medications and/or to follow treatment recommendations, is an increasing problem in randomized clinical trials to support product development. Nonadherence affects the quality of a trial and compromises a sponsor’s ability to detect a treatment effect or decide whether a failed study is indeed due to an ineffective treatment.
- Select studies in schizophrenia have focused on the use of mobile technologies, particularly around patient self-assessment and the value of psychoeducational tools. Early results are promising even in symptomatic patients.
- The majority of clinical trials use a combination of pill counts and self-reported data to measure drug adherence, despite the drawbacks of relying on these types of indirect measures. It is assumed that doses are taken, but the exact timing of these is often incomplete and imprecise. Pill counts have frequently been shown to underestimate poor adherence and nonadherence.
- Because of its ability to ensure treatment adherence, directly observed therapy (DOT) has been used for decades to both measure and maximize adherence for treatment of tuberculosis infection and antiretroviral therapy and to ensure ingestion in inpatient settings or in early-phase clinical trials when subjects are dosed in the clinic. However, for trials conducted in outpatient populations, the cost and logistical complexity of administering DOT forces clinical trials to switch to less intensive monitoring despite the continued and largely unmeasured risk of nonadherence.
- The artificial intelligence (AI) platform (AiCure, New York, NY) uses AI to visually confirm medication ingestion using software that can be downloaded as an application onto any mobile device. Software algorithms identify the subject, the drug, and confirm ingestion. Encrypted data are sent to cloud based dashboards for real-time monitoring and intervention, with suspicious activity, duplicate enrollment, or incorrect usage triggering alerts.
- To evaluate the use of an AI platform on mobile devices in monitoring and increasing medication adherence, pooled data from 6 CNS trials were evaluated.