Many quality improvers know that data should help you understand your performance over time. But what if your data could show you the future? In this blog post, Andrey Ostrovsky, MD, a practicing physician and social entrepreneur, and Lynne Chase, Massachusetts Program Director of The New England Quality Innovation Network-Quality Improvement Organization, explain how analytics are helping health systems improve outcomes for patients in crises — and transition from fee-for-service care into value-based models. Dr. Ostrovsky is CEO of Care at Hand, a health care survey start-up, and will be presenting as part of the IHI Innovation Relay at the National Forum.
Andrey Ostrovsky (left) and Lynne Chase (right)
More and more, health systems are shouldering the responsibility for their patients’ health through risk-bearing arrangements like accountable care organizations (ACOs) and bundled payments.
By 2018, more than 50 percent of Medicare spending will go toward reimbursing health care in such value-based models, according to the US Secretary of Health and Human Services. Providers are feeling more pressure than ever before to improve outcomes and patient experience at lower cost. One solution they’re turning to is innovation in data and on the care team.
Predicting the Future with Aggregate Data
How do you know which of your patients is most likely to be hospitalized? Right now, data for quality improvement is largely limited to claims and electronic health records (EHRs). Medical providers in the clinic collect this data, and it has a big limitation — it’s episodic, leaving a blind spot between claims.
Frontline workers collecting data in the patient’s home can create a much different picture of a patient’s health. New technology is helping to fill the blind spot missed by claims data. The graphic below shows the difference this can make for providers trying to identify high-risk patients before it’s too late.
New Roles, More Data, Better Care
New technology plays a big role in the shift from episodic to continuous big data. But the other key piece is what we call workforce innovation — re-imagining the care team to deliver higher quality care at a lower cost.
Doctors’ salaries are often three times higher than other care team members, as can you can see in the graph below. As a result, physician-led care coordination models are not the best way to drive down cost of care, especially when better care requires more frequent touch points with patients.
The drawbacks of physician-led care aren’t only about cost. Doctors know that unmet social needs directly lead to worse health, but they don’t feel confident in their ability to address those social needs, according to a Robert Wood Johnson Foundation study called “Health Care’s Blind Side
.” Other team members may be better equipped to meet these social needs.Data Success Stories
Innovative care teams are exploring new approaches that use big data and high-touch, low-cost frontline staff to improve care for patients. Here are some examples:
In Washington state, recent wild fires led to massive evacuations and a state of public health emergency. A few hundred miles downwind of the fires, a community-based care transitions program in Spokane, Washington, used predictive analytics to map patients with increasing risk of becoming hospitalized.
The community provider, Aging and Long-term Care of Eastern Washington, was able to identify 55 high-risk patients, and providers responded with proactive care for 23 of those patients. The provider tested several approaches to optimize the risk stratification based on diagnoses, medication use, and durable medical equipment use. The provider also tested several approaches in its outreach efforts, including remote and in-person interventions.
On the other side of the country, West Baltimore is recovering from its own state of emergency in the wake of the death of Freddie Gray, a black man killed in police custody. The recent riots highlighted the challenging environment of some of Maryland’s most marginalized people. West Baltimore's population suffers from many disparities, including a life expectancy of 67 years compared to mid-70s for the surrounding zip codes.
Despite the health disparities, a decades-old community-based organization, The Coordinating Center, was able to substantially reduce 30-day readmissions, according to data from the Centers for Medicare & Medicaid. The key to their success? A high-touch, low-cost intervention using non-medical community coaches to meet patients were they were, in the community.
As part of its quality improvement strategy, The Coordinating Center implemented predictive analytics to help their non-medical staff detect health decline much earlier than before. Since the introduction of the technology in January, The Coordinating Center was able to reduce readmissions another 5 percent, to a readmission rate of about 20 percent (orange bars below).
These examples highlight providers that are not only improving the quality of care for vulnerable populations, but also discovering delivery models that will enable them to survive the transition to an outcomes-based health care system.
Innovations from community-based providers are rarely disseminated in academic publications. It would be powerful to bring these little-known improvement initiatives into the academic discourse with review, critique, and dissemination among leading journals
, which could speed up the spread and adoption of these value-based innovations.
You may also be interested in:Video: What Are the Phases of IHI Innovation Projects?