Mercy hopes Ayasdi’s big data platform can save $100M and change doctors’ habits
August 26, 2014 6:00 am by Dan Verel
Silicon Valley’s Ayasdi has teamed up with Mercy and its more than 40 hospitals on a new big data effort that seeks to drastically reduce the time needed for improvements in clinical pathways on a number of procedures.
That, in turn, could help the health system reduce costs by improving efficiency, clinical knowledge, outcomes and eventually readmission rates, according to Gurjeet Singh, CEO of Ayasdi, which launched recently but was born out of Stanford University’s Mathematics Department.
While myriad big data efforts are underway throughout the healthcare industry, the Ayasdi Care platform is unique in that the data gleaned from EHRs, along with financial information, can be specifically tailored to individual health systems’ patient data and demographics, Singh said. It can also be put together much faster than it would ordinarily take to implement clinical pathways or best practices, which typically take years.
“It’s all very hypothesis-oriented,” Singh said. “A human being has to come up with the hypothesis, and with the data going the way it is, that’s not a very efficient way to do things. What we have done is developed a platform that sifts through the data and points people in the right direction.”
St. Louis-based Mercy, with its dozens of hospitals and hundreds of clinics and outpatient facilities, expects it can both significantly improve clinical outcomes while saving hundreds of millions of dollars in the near term.
“Ayasdi Care provides an intuitive interface that allows our physicians and staff to understand the ‘why’ behind specific care pathways which makes it easier for them to adopt and adhere to our own best practices,” said Seth Barbanell, MD, VP of Mercy’s Clinical Performance Acceleration.
Singh said Mercy anticipates it could save as much as $100 million over two years.
The speed in processing the data is a key part of achieving those types of savings, Ayasdi CMO Patrick Rogers said.
“It would take them months to come up with a recommended pathway, and we can cut that from months down to weeks or even days,” he said. “That’s very powerful.”
It’s perhaps analogous to how genomic data is comparatively easy to sequence, but requires significant resources to analyze the data and gain insights. If the process can be sped up while maintaining quality data, programs can then be developed that take into account the patterns gleaned from the data.
Also important is the rate of adoption of any new pathways among physicians. Typically, healthcare providers rely on peer-reviewed studies and expert physician knowledge. But that approach relies on consensus, across a limited number of physicians, resulting in greater variation in treatments, which lowers physician adoption, the company said.
“They are expecting a much higher rate of adoption because it’s their own data,” he said.
“The ability to combine the complexity of medical data with recent innovations in machine learning will quicken and deepen the pace of medical innovation,” said Skip Snow, Senior Healthcare Analyst, Forrester Research. “New standards of care will be discovered. Software assisted diagnosis will become commonplace. Discovering new uses for medicines will regularly occur.”
Co-founder and CEO Singh said Ayasdi Care is ideal for large health systems that generate significant amounts of data, and that the company is hoping to replicate its model with Mercy across the industry.
Ayasdi, with about 100 employees, isn’t limited to healthcare, either, having applied its analytics to the pharma, finance and oil industries. It also counts General Electric among its clients. Last year it raised $30 million and has grown by about 4x year-over-year, according to Singh.
