Skip to content Skip to navigation

Getting Beyond Encounter-Based Care to Include Social Determinants

June 2, 2015
| Reprints
Panel describes challenges to Policy Committee Workgroup

On June 2, the federal Health IT Policy Committee’s Advanced Health Models and Meaningful Use Workgroup heard from several innovative organizations about both the challenges and benefits of integrating data across settings, care teams, and human service organizations. The workgroup hopes to use the input to develop recommendations about the IT infrastructure that would support advanced models of care.

Speaking in a panel session were representatives of several organizations that have worked to develop patient care plans that reach out to include groups such as social service organizations, food banks, corrections, and home health providers.

“If you take care of somebody with diabetes and they live under a bridge, you have to pay attention to housing and food issues, or you will have no success at medical care,” said Nancy Garrett, chief analytics officer at Hennepin County Medical Center in Minnesota.

Hennepin Health is an 11,000-member, defined provider network with a shared EHR and a risk-sharing funding model. It integrates medical and social services to address social determinants of health under a consensus-based governance model.

Garrett said Hennepin uses risk stratification to identify the highest risk patients and surround them with support. The organization is bringing data from a variety of sources, including EHRs, claims, dental care and social service case information, into a data warehouse. “We try to look across the continuum,” she said. “We have had challenges with building the data warehouse, because it was built on a shoestring. You have to make sure you have enough analyst bandwidth,” she said.  “I would recommend thinking of how many analysts you need and then double it, because insights are key to population health management.”

Garrett said both state and national laws make two-way sharing of social services data with medical providers difficult. “It took our privacy officers over a year of meetings to make happen, so there is a lot of waste there,” she said. Hennepin also has struggled with patient identification matching, Garrett said. “This legislative session our state legislature passed laws to make that easier.”  (Other providers echoed her concerns about the variety of federal and state regulations around consent and sharing data with social service agencies.)

She added that payers need to make appropriate risk adjustment of payment rates for social determinants of health. Finally, she noted that EHRs are designed to be transactional – one patient at a time – and for medical settings. “We need policies that incent vendors to develop systems that work across social service, public health and medical care,” she said.

L. Allen Dobson Jr., M.D., president and CEO of Community Care of North Carolina, said one lesson his organization has learned is to try to hand off insights rather than just documents. “Eighty-six-page Continuity of Care Documents do not change care delivery or health outcomes,” he said. “We try to hand off actionable insights that inform the team member receiving the information.” CCNC, a statewide network of medical homes with 1,800 primary care practices, transfers actionable insights beyond just the medical community. “It requires data liquidity from all electronic systems of record to get data into the hands of people who can help such as pharmacists and care managers,” Dobson said. As an example, CCNC’s PharmaceHome program combines medication use information from multiple sources (claims or fill transactions, hospital discharge summary, PCP EMR, brown bag review done by care manager) into a single community record. That enables a much more robust and efficient medication reconciliation process, he said, integrating community pharmacy, medical home, and community-based care managers into a common workflow. 

Dobson said CCNC does not yet have a single shared care plan for patients. “We are struggling with how to do that,” he said. “We have claims data and are integrating that with clinical data, but how do you take social determinants and characterize them and give them back to the EMR or social service agencies in a way that is useful to them? That is future work that needs to be done.”

After a full day of panel sessions June 2, the workgroup plans another full day of meetings June 3 to discuss translating the panel discussions into recommendations for the full Health IT Policy Committee.