By Deborah Borfitz
July 16, 2019 | Planned enhancements to the primary care model of Stanford Health Care will blend the workflow of its Center for Personalized Wellness with the "democratization of healthcare" features of the All of Us Research Program of the National Institutes of Health. The focus is on detecting disease earlier, strengthening patient-provider relationships and deploying the latest health technology.
In a recently completed pilot, this "Humanwide" approach succeeded in not only identifying previously undiagnosed conditions and future health risks, but also reducing provider burnout in one of the most at-risk fields of medicine, according to Megan Mahoney, Stanford Medicine's chief of general primary care. Initial learnings from the one-year study were outlined in a paper recently published in Annals of Family Medicine (doi: 10.1370/afm.2342).
The most successful features of Humanwide will be rolled out systemwide, she says, including a digital health component linking data from mobile monitoring devices to patients' electronic health records. Other likely additions are genetic testing of patients at significant risk of inherited disease and pharmacogenomic screening to inform treatment decisions for certain patients—services that insurers increasingly are covering. Stanford Health Care Alliance may well add those benefits to its value-based payment plan, Mahoney says.
Humanwide was built on the foundation of Primary Care 2.0, the practice redesign effort of Stanford Medicine that brings together inter-professional healthcare teams to collaboratively care for patients. Each team is comprised of a nurse practitioner or physician assistant working alongside a primary care physician, supported by highly trained care coordinators who do health coaching and in-room scribing, Mahoney explains. That core group is backed by a nutritionist, clinical pharmacist, behavioral health specialist, nurses and physical therapists.
The staffing component is critical, and not impossible to replicate elsewhere provided care teams can respond to data in a timely fashion, says Mahoney. Primary care as currently rendered—typically a physician spending 15 minutes with patients once a year—is compromising patient health and contributing to professional disillusionment.
Physicians quite literally do not have enough time to manage the chronic and acute illnesses of their patients and deliver all the recommended preventive services, she says. Published studies on the subject indicate it would take the average physician at least 15 hours a day to achieve that goal.
Primary Care 2.0 "leapfrogs" out of that situation with physician extenders and virtual visits, says Mahoney. At the first such clinic in Santa Clara, California, where the pilot was based, 10,000 patients are served by three core teams. "We try to keep patients outside the walls of the clinic, and in their homes and at work, by bringing services to them."
The team-based approach is also less onerous for primary care physicians than "playing the role of social worker, clinical pharmacist nurse and nutritionist all wrapped in one," says Mahoney, while dealing with mounting paperwork and communication requirements associated with the job.
High Tech and High Touch
Fifty patients were recruited into Humanwide through their primary care provider and interest in participating was high along the spectrum of medical complexity, says Mahoney. Well patients, chronically ill patients, and medically complex patients were nearly evenly represented.
Pilot participants regularly interacted with care teams at the clinic and worked with a certified health coach to identify wellness goals and create a plan for achieving them, says Mahoney. Their health status was continuously monitored based on initial genetic assessments and pharmacogenomic screening and incoming data from home-based and wearable devices. The mobile monitoring devices included a glucometer, pedometer, scale, and blood pressure cuff.
The idea, she says, was to capture data on a variety of factors known to influence health: activity, behaviors, physical characteristics, genetics, biological conditions, care utilization, and environmental exposures. Team members met regularly to review the multiple streams of data for each patient and took preventive action when health risks were discovered.
Using technology in healthcare both creates efficiencies and helps ensure medically-complex patients benefit from concierge-level services, says Mahoney. In fact, Humanwide borrowed from the "best and most effective components" of the well-respected Stanford Coordinated Care primary care model that caters to high-need patients collectively accounting for over half of all healthcare spending, she adds.
Humanwide also paralleled the All of Us Research Program in that it considered individual differences, only "in a clinical setting so patients could directly benefit from… the promise of precision medicine," Mahoney says. "We intentionally recruited a diverse patient sample along the lines of race, gender, age, and medical complexity. We weren't really sure who would benefit the most when we started integrating all these new streams of data."
High Satisfaction
Demonstrated outcomes of the Humanwide study included the detection and treatment of previously overlooked health concerns. Among 33 women screened for breast cancer risk, five were identified as having a very high risk and in need of ongoing, enhanced surveillance, Mahoney offers as an example.
More than a dozen changes were made to medication prescriptions or dosages as a result of the pharmacogenomics screening, she continues, one of which relieved pain for a patient experiencing persistent leg cramps from statins. And early diabetes or hypertension was identified in several patients thanks to continuous readings from home-based devices, leading to medication and lifestyle changes.
Patients and providers both expressed satisfaction with the new care approach, Mahoney adds. Enrollees liked the strong connection with their care team and the opportunity to apply their personal data to their health. Clinicians reported that they felt more engaged in their work when sharing the goal of caring for a patient with a like-minded team.
Since the pilot was implemented in an area where patients are generally well insured, well employed, and confront relatively few social determinants of health that might impact their care, Mahoney was initially concerned that Humanwide interventions might exacerbate disparities.
Quite the opposite, Mahoney says. Over 90% of participants reported their care was "very accessible," including the oldest patient—an 86-year-old African American woman. Similar digital health interventions have been successfully implemented in the Medicaid population, she adds, making that aspect of the program generalizable across populations.
A similar primary care model was developed by Bellin Health System and deployed across dozens of practices in Wisconsin, says Mahoney. As with Primary Care 2.0, Stanford Coordinated Care inspired its design.
Next Steps
Stanford Health Care is now looking to add Humanwide's digital health component to the Primary Care 2.0 model, as well as the genetic and pharmacogenomic testing features for select populations, says Mahoney. "Currently, we're designing scripting for the medical assistants so that they can provide the health coaching that is linked to incoming data from the digital health devices."
Blue-tooth-enabled home devices used during the pilot remain with participants and the healthcare teams at the Santa Clara clinic continue to respond to the incoming data, Mahoney says. Study participants also have access to a dynamic database that keeps them updated on any new clinical findings related to their genetic test results.
Predictive analytics will be required to interpret complex patterns in the big data amassing from incoming data that can be presented to care teams as actionable findings, she says. Industry partnerships are being explored to develop solutions. Stanford already has publicized a partnership with Lumeris, a consulting firm with expertise in population health initiatives with a predictive analytics component.