Big Data Meets Personalized Care

"We want to do for patient information what Google did for general information." Riley Bove, MD

The buzz around precision medicine has been building for years. But despite its promise and PR, the infrastructure for incorporating precision medicine approaches into clinical care remains largely inadequate, especially in fields like neuropsychiatry, where patient and population data can be immense in quantity and varied in quality and category. The potential payoff, though, makes it a challenge worth pursuing.

That’s why Drs. Riley Bove and Kate Rankin, Department of Neurology; and Stephan Sanders, Department of Psychiatry, joined forces to develop the “neuropsychiatry clinic of the future,” a virtual space that puts every patient’s data in one place, allowing for quick comparisons of individual data against broader data sets and more precise, tailored care.

The neuropsychiatry clinic of the future: “We want to do for patient information what Google did for general information: put all clinical data in a central repository that is available to clinicians and researchers. Clinicians would see the information they’d need to inform their care recommendations and fuel their conversations with patients, and researchers would be able to mine the data to identify potential biomarkers for neurological and psychiatric diseases, develop predictive models about how patients might fare, and much more.” Stephan Sanders

The chance to be truly unbiased: A tool like this could remove some normal, but still problematic, clinical bias. That kind of bias – often based on a clinician’s firsthand patient experience – impacts how clinicians perceive their patients’ health, the way they anticipate how their patients will do in the future, and even the diagnoses they make. If a physician has been treating patients with a particular disease for decades, he or she might have an outdated view of how treatable that disease is. With this new tool, clinicians will be able to compare their new patients with research cohorts that have been beautifully characterized and treated according to the most modern standards of care. It eliminates therapeutic nihilism by creating expectations based on the most current data, rather than sometimes outdated experiences.” Riley Bove

The unique challenges of aggregating data in neuropsychiatry: “In neurology and psychiatry, much of our understanding of a disease comes from the clinical features that we observe and test. Interpersonal behavior, social behavior, cognition, and motor skills all help us understand what’s happening with a brain. There are so many different elements to creating a full picture of a patient that we need to consider. Getting data in front of a clinician in a way that makes sense, can be reproduced, and can be compared across patient populations is a unique challenge.” Kate Rankin

When diagnosis remains imprecise: “The sad truth is that even specialists often misdiagnose patients with neurodegenerative diseases. They say, ‘Oh, you have Alzheimer’s,’ when the patient actually has progressive supranuclear palsy, or vice versa. That’s a problem, because the trajectory of the patient is different, and the treatments are different. It can take months, years – forever – to get the right diagnosis. If we can get the right data in front of doctors quickly, we can help them make the right diagnoses from the outset.” Kate Rankin

A better way to help patients understand what’s next: “When patients are diagnosed with a neurological or psychiatric disease, they want to know what is going to happen to them, and they want that information in the simplest way. We plan to create categories based on disease severity and other characteristics so clinicians can tell their patients that if they fall into this category, they will get this treatment, then this might happen – essentially creating a map for patients and providers to navigate together.” Stephan Sanders


UCSF Weill Awards

Joseph DeRisi, PhD; Samuel Pleasure, MD, PhD; and Michael Wilson, MD