The data dilemma

Unlocking Health Insights: AI, Data, and the Future of Patient Care

The panel, moderated by Mr. Amit Katwala, Features Editor at Wired, explored “The Data Dilemma” in healthcare, focusing on AI’s role in transforming data into actionable insights. Mr. Matt Pavelle, Co-founder and Co-CEO of Doctronic, and Dr. Emilė Radytė, Co-founder and CEO of Samphire Neuroscience, shared their perspectives on data utilization and privacy.

Mr. Pavelle introduced Doctronic as an AI doctor that has processed 18 million patient consults. The platform requires accurate health data to provide reliable information, extracting it conversationally or through imported health records. Crucially, Doctronic never trains on user data, asserting that patient data is sacred and owned by the individual, with AI primarily assisting human doctors.

Dr. Radytė described Samphire Neuroscience’s work developing medical-grade neurotechnology for women’s health, targeting hormone imbalances during various life stages. Her company manages sensitive cycle-related, brain, and symptom data. This data helps predict conditions and tailor therapeutic outcomes, reflecting a deeply personal interaction with patient information.

Regarding privacy, Mr. Pavelle explained Doctronic’s anonymous default, allowing users to interact without providing personal identifiers unless they choose to save their chats. This approach ensures privacy, especially where discussing health topics might be sensitive. Doctronic does not track anonymous users, reinforcing data ownership and control.

Dr. Radytė noted Samphire’s commitment to anonymity, preventing direct linking of data to individuals. Patients can opt-in to share data, driven by a desire to advance research in underserved women’s health. Many users willingly contribute anonymized data, recognizing its potential to improve collective understanding and treatment.

The discussion highlighted the utility of data. Mr. Pavelle stated that Doctronic’s AI achieves 99.2% treatment alignment for common primary care issues, relying on doctor-written guidelines rather than user data for training. He observed that while diagnoses might vary, treatment plans often converge, making the AI’s advice highly effective.

Dr. Radytė emphasized subjective symptom data as a significant opportunity, particularly in women’s health. She noted that diverse patient descriptions like “brain fog” or “emotional exhaustion” often map to similar underlying profiles. AI can analyze these patterns to create standardized classifications, such as “Type 1 PMS,” which could lead to more precise treatment guidelines.

The panel also addressed the interplay of subjective and objective data from wearables. Dr. Radytė argued that subjective experience, reflecting brain activity, is paramount, even when physiological markers are absent, as seen in depression research. While objective data is valuable, subjective insights are often the primary indicator of a patient’s condition.

Mr. Pavelle suggested that more data could enable proactive healthcare, with AI identifying troubling trends from health records and wearables to recommend timely interventions like cholesterol tests. However, Dr. Radytė cautioned against the potential for medical anxiety caused by excessive monitoring and uncorrected statistical findings, advocating for careful presentation of information to patients.

Looking to AI’s future role, Dr. Radytė proposed “cohortization” – using AI to identify groups with shared symptoms to apply proven therapeutic patterns, rather than solely individual personalization. This approach facilitates early intervention and leverages collective data for broader benefit in medical science.

The conversation concluded on the indispensable role of human doctors. Mr. Pavelle stressed that AI would alleviate physician burnout by automating routine tasks like charting and prescription renewals. This frees up doctors to focus on complex cases and meaningful patient interactions, with AI functioning as a “chief resident,” providing comprehensive information and suggestions.

Dr. Radytė shared that users of Samphire’s Nettle device reported fewer doctor visits and less medication use, indicating improved self-management. This suggests that empowering patients with effective tools can lead to more purposeful doctor engagements. Ultimately, human doctors remain essential for decision-making and accountability, roles that AI cannot yet fully assume. Education for clinicians on AI’s supportive capabilities is vital for its successful integration.