The Kaiser Permanente Evidence-based Practice Center synthesizes scientific evidence on health care topics and produces comprehensive reports that are used to guide decisions about health care.
With support from the Garfield Memorial Fund, Kaiser Permanente researchers are using big data to help customize and improve care for depression
By Jessica Ridpath
Thanks to modern technology, finding the best way to travel from point A to point B has never been easier. Between our phones and our cars, many of us have personalized, real-time driving directions, traffic updates, and maps of alternative routes at our fingertips.
But with depression treatment, finding the best way forward still involves a lot of guesswork. With dozens of options for antidepressant medications and psychotherapy available, depression care can take many different paths. How can clinicians know which one is likely to work best for individual patients?
For answers, Kaiser Permanente researchers are tapping into the power of big data. With a $380,000 grant from the organization’s Sidney Garfield Memorial Fund, they aim to create a one-of-a-kind data resource to support a more individualized approach to depression treatment called “feedback-informed care.”
A modern navigation system for depression treatment
Traditional approaches to depression care are informed by general guidelines based on what works on average. In contrast, feedback-informed care matches treatment recommendations to individual circumstances.
“Feedback-informed care is sort of like our modern-day navigation systems,” explains project co-leader Gregory Simon, MD, MPH, a Permanente psychiatrist and senior investigator at Kaiser Permanente Washington Health Research Institute. “Instead of an old-fashioned paper map, clinicians and patients get customized, turn-by-turn directions.”
And like modern mapping tools, what makes feedback-informed care possible is the timely collection, coordination, and analysis of vast amounts of data.
“At the national level, Kaiser Permanente has the best data ever available to look at what works and what doesn’t work in depression care,” says Dr. Simon. “Previously, the largest studies done on treatment outcomes included 1,000-1,500 people. Across Kaiser Permanente, we have data on depression treatment for 500,000 people or more. That’s multiple orders of magnitude bigger than what has ever been possible before.”
Using this unparalleled data resource to improve feedback-informed care for depression is a top priority for Kaiser Permanente’s national behavioral health leadership. Heeding their call, Dr. Simon worked with Arne Beck, PhD, director of quality improvement and strategic research at the Kaiser Permanente Colorado Institute for Health Research, to develop a plan to turn that vision into reality.
“Over the last decade, Kaiser Permanente regions across the country have made huge strides in implementing feedback-informed care,” says Dr. Beck. “This grant from the Garfield Fund gives us a chance to coordinate and scale up these efforts.”
Turning complex math into easy-to-use reports
In the project’s first phase, the research team will improve upon Kaiser Permanente’s existing data resources in two of its regions: Washington and Colorado. The improvements will allow for more data on depression treatments and outcomes to be pooled together more quickly. The next step is to translate those data into reports that clinicians and patients can use to guide individual treatment choices for depression.
“The idea is to give clinicians and patients standardized information about which treatments are most likely to work well in the patient’s particular situation,” says Dr. Simon. “Secondly, we want to give them a way to compare how well an individual patient is doing based on all the data we have about depression outcomes for people with similar characteristics. If a patient is not doing as well as expected, this alerts the clinician that it might be time to adjust the treatment plan.”
To ensure that these reports are user friendly, the research team will develop prototypes in collaboration with behavioral health leaders, clinicians, and patients.
“We’re focused just as much on implementation as we are on analytic methods,” explains Dr. Beck. “The math behind this work is complicated. But our reporting tools need to be easy for patients and clinicians to use in the real world.”
Taking science off the shelf
The project’s second phase involves taking the data resources, methods, and reporting tools developed in the first phase and spreading them across Kaiser Permanente’s six other regions. If successful, the result will be the largest data resource on depression care ever available—with information on the treatment of more than half a million people. Future work could expand the approach to a range of other conditions, including anxiety, suicidal ideation, and alcohol use disorders.
“The exciting thing about this project is that it’s really taking science off the shelf,” says Dr. Beck. “It’s giving us the resources we need to effectively implement and study feedback-informed care in the clinic. By developing these new tools and creating useful reports for administrators, clinicians, and patients—and then scaling this up to the national level—we have the potential to improve care for hundreds of thousands of people.”
“I used to joke that analytic methods for predicting a person’s video rental choices were more sophisticated than the methods we have for predicting which depression treatments will work best for different people,” adds Dr. Simon. “With this project, we’ll bring data science to personalizing mental health care.”
Jessica Ridpath is a senior research communications consultant with the Kaiser Permanente Washington Health Research Institute, where this article originally appeared.