S.F. startup builds Medicaid database
Jini Kim’s relationship with Medicaid is business and personal.
Her San Francisco startup, Nuna, while working with the federal government, has built an online database of the nation’s 74 million Medicaid patients and their treatment.
Medicaid, which provides health care to low-income people, is administered state by state. In California, it is known as Medi-Cal. Extracting, cleaning and collating the information from so many disparate and dated computer systems was an extraordinary achievement, health and technology specialists say. This new collection of data could inform the coming debate on Medicaid spending.
Andrew Slavitt, acting director of the Centers for Medicare and Medicaid Services, described the cloud database as “near historic.” Largely because Medicaid information resides in so many state-level computing systems, he explained, “We’ve never had a systemwide view across the program.”
This week, for the first time, Nuna’s executives are talking about the company’s funders, business strategy and work for Medicaid at the J.P. Morgan Healthcare Conference in San Francisco.
Health data on its own — billing, diagnostic and treatment information, typically recorded in arcane, shorthand codes — is not very useful. But if it can be aggregated and analyzed economically and quickly, that data is seen as a vital ingredient in transforming health care.
The health care marketplace in the traditional fee-for-service model values volume. More doctor visits, hospital stays, operations and pills mean more revenue and profit for health care providers. But the push in recent years has been toward what is known as value-based health care. In the value model, medical groups are paid for outcomes: patients treated more efficiently and people who are healthier.
But that transition becomes possible only with accurate, reliable data as the raw material for measuring outcomes and discovering what works and what does not. That is the reason for the excitement about and investment in health information technology, and the goal of the Medicaid work with Nuna. The data will be stripped of identifying information before it is released to researchers.
Kim, 35, a former Google product manager, has firsthand experience with Medicaid. Her brother, Kimong, who is a year younger, is severely autistic. When he was 8, Kimong started having monthly grand mal seizures, which are characterized by loss of consciousness and violent muscle contractions. The ambulance, hospital, doctor and therapy bills piled up. As a 9-year-old, Kim helped her Korean immigrant parents complete the Medicaid application forms.
“Our family would have gone bankrupt without Medicaid,” she said. “It saved us.”
The Medicaid system covers millions of working families, older people, children and people with disabilities. In fact, 40 percent of Medicaid spending goes to the people with disabilities. Half of long-term care in the United States, mainly for older people, is through Medicaid. And nearly half the children born in the U.S. are in the Medicaid system.
The trouble with traditional health data warehouses, specialists say, is that they resemble digital vaults. It is difficult and time-consuming to get information in or out, and only people with specialized skills can use them.
But the new cloudbased technology, using Internet-era software, is flexible and interactive. It opens the door to the monitoring of emerging disease clusters, billing patterns and program effects. For example, did the percentage of low birth-weight babies decline after a Medicaid program was put in place? If so, how much?
“This kind of data can help move health care policy from a partisan ideological debate to one informed by knowing who the people affected are and what will likely happen to Medicaid recipients,” said Drew Altman, president of the Henry J. Kaiser Family Foundation, a nonprofit health policy research organization in Menlo Park.
Though young, Kim has spent years learning hard-earned lessons about health care. At Google, she was a product manager on Google Health, a failed effort to attract millions of people to use its free, online personal health records.
“Health care is hard, and humility is important,” Kim said. “You can’t just put technology on something and assume it’s going to work. You really have to understand the ecosystem in health care.”
In late 2013, Kim got a call from Washington and became one of the small cadre of Silicon Valley technology specialists called on to fix HealthCare.gov, the application website for health insurance under the Affordable Care Act. That was six months of 18-hour days, seven days a week through Thanksgiving, Christmas, New Year’s and beyond, until the end of the spring 2014 enrollment period.
Nuna, founded in 2010, had only one full-time employee, Kim, until 2014, when it got an early round of venture financing. And no one was paid until then. David Chen, the company’s co-founder and chief data officer, holds a doctorate in bioinformatics from Stanford University, but he also was a data scientist at Netflix for three years while Nuna was getting off the ground.
Today, Nuna has 110 employees. It has raised $90 million in venture capital led by Kleiner Perkins Caufield & Byers and prominent individual investors, including John Doerr, Kleiner’s chairman, and Joi Ito, director of the MIT Media Lab.