Amae Health, a behavioral health provider focused on serious mental illness, is teaming up with Google Health Enterprise to integrate wearables data into its precision psychiatry model.
Amae will provide Fitbit devices to patients while they’re in treatment. If they consent to data sharing, Amae clinicians will receive regular summaries of their sleep patterns, exercise and heart rate variability data. Amae’s long-term goal is to develop a predictive model that pulls on wearables and other data to measure disease state and progression and spot early warning signs of clinical decline.
Amae treats conditions like schizophrenia and bipolar disorder, with programs across partial hospitalization, intensive outpatient and maintenance therapy settings. It features a multidisciplinary care team, including therapists, primary care docs, dietitians, health coaches, peer supporters and social workers.
Psychiatry has historically lacked objective tools to measure disease progression and treatment response, Amae executives say. This is in part due to the lack of scalable solutions capturing longitudinal data. Wearables make that possible.
“Biomarkers have a long and frustrating history in psychiatry, even going back to the 70s,” Scott Fears, M.D., Ph.D., Amae’s chief medical officer and professor of psychiatry at the University of California, Los Angeles, told Fierce Healthcare.
Just as there are genetic factors associated with psychiatric illness, Fears said, so there are biomarkers that may indicate a mental health condition. Yet both fields face the challenge of untangling the many factors that may all contribute some risk in tiny amounts. “They’re small biomarkers, so we really need to aggregate these things over lots of biomarkers,” Fears said.
It’s why Amae is looking to collect medical records, lab data, medication data, EEG data, voice data, wearable data and patient-reported outcome questionnaires to develop a mental health composite score. This will be supported by machine learning and artificial intelligence.
In an ideal future, the company said it would work like this: a patient is assessed at intake and their biomarkers indicate a likely direction of a diagnosis and intervention. Within a week, their biomarkers are analyzed to see if the treatment is working or needs to be adjusted. Then, when they’re stable, biomarkers could indicate wellness. If someone is in decline, an algorithm could pick up on that, alerting an Amae provider to proactively check in with that patient.
As part of its ongoing work, Amae is committed to research and publishing its findings. “That whole peer-review process really will keep us on track,” Fears said, inviting academics to poke holes in Amae’s research.
“As part of our ethos and company aspirations, we want to be doing research and collecting data. At the same time, we aren’t like the typical academic institution [or pharma] that has a very, very focused question,” Fears noted. Amae is not only focused on a single disorder or intervention. It is analyzing trends at the individual and population levels across conditions and factors.
Amae has seven clinics and clinical partnerships with Cedars-Sinai, Mass General Brigham and New York-Presbyterian. These partnerships allow Amae to operate as the outpatient SMI arm of a hospital, integrating with its EHRs and supporting transitions from EDs and inpatient units. Because Amae offers different levels of care, it aims to be a long-term provider managing patients across phases of recovery. Its goal is to reduce repeated hospitalizations and failed handoffs between disparate programs.
Because of its value proposition—keeping patients engaged in care and reducing downstream costs—Amae has some value-based arrangements and is looking to expand with more. Today, Amae works with commercial payers and some Medicaid plans. It is currently operating in New York, North Carolina and California.