Prediction of Medication Adherence in Clinical Trials Using Machine Learning

In this poster you will learn:

Patient non-adherence to study medication in a clinical trial can lead to negative results for potentially promising treatments.
Selection of patients who are most likely to demonstrate high adherence throughout a trial can facilitate the accurate measurement of the therapeutic effects of a compound.
In the current investigation, we apply machine learning-based forecasting models to assess their ability to predict future patient adherence based on patterns observed in patients’ behavior.
The ability to identify patients at risk of low or non-adherence will allow clinical site staff to engage in proactive interventions that can prevent medication non-adherence prior to the occurrence of repeated low or non-adherence.
Similarly, such algorithms can allow studies to avoid enrolling low or non-adherent patients prior to randomization when utilizing a lead-in period.