Digital Phenotyping and Post Approval Monitoring of Drugs

Alessio Signorini’s presentation focused on how digital phenotyping and artificial intelligence can fundamentally expand what clinicians are able to observe and measure about patients outside the clinic. He highlighted a core limitation of traditional medicine: we primarily see patients through episodic, visible data points (visits, labs, imaging), while the vast majority of health-related signals—daily behavior, sleep, activity, heart rate, voice, and environment—remain “invisible” to the healthcare system .

The talk showed how wearable devices, smartphones, and connected apps now generate continuous, passive data streams that can be transformed into clinically meaningful digital biomarkers. Examples included monitoring functional decline, detecting influenza-like illness, assessing treatment effects on sleep and activity, identifying atrial fibrillation or migraine events, and measuring cognitive decline through voice analysis. Importantly, these signals often change earlier and more sensitively than traditional endpoints, offering opportunities for earlier intervention and better longitudinal monitoring.

From a physician’s perspective, the key message is not replacing clinical judgment, but augmenting it. Digital biomarkers can contextualize what we see in the clinic, help distinguish true change from noise, and provide objective trends over time rather than single measurements. Signorini emphasized that this work is grounded in peer-reviewed research, regulatory-compliant platforms, and partnerships with healthcare institutions and Pharma. The broader implication is a shift toward more preventive, personalized, and continuous care—where clinicians can better understand the “other 99%” of patients’ lives and make more informed decisions about treatment, monitoring, and outcomes.