Almost no topic is garnering as much attention in healthcare as social determinants of health (SDH). As public health researchers continue to point out, your zip code is a better predictor of your health than your genetic code.
Along with identifying which social determinants have the most impact on health outcomes, researchers are also working on how to get the SDH data into EHRs, and the exercise is not just academic. CMS’ Accountable Health Communities initiative will soon require 32 sites across the country to screen 75,00 patients around their social needs.
Earlier this year I wrote about a project sponsored by the National Association of Community Health Centers called PRAPARE (Protocol to Respond to and Assess Patient Assets, Risks, and Experiences), which was designed to create and implement a national standardized patient risk assessment protocol to assess and address patients’ social determinants of health. Questions developed in the protocol were incorporated into a tool in the EHR so that data could be collected in the clinical workflow. Researchers also structured the data so it could be used for reporting and analysis.
Yesterday I saw an interesting presentation by Erika Cottrell, PhD, an investigator at Oregon-based nonprofit OCHIN, which has piloted an effort to build on PRAPARE by working on an implementation of SDH tools it developed in Epic for three of its community health center partners.
Founded in 2000, the Oregon-based nonprofit OCHIN offers a centrally managed instance of Epic to more than 100 member organizations and 500 member clinics focused on the safety net in 18 states. For this group’s patients, social determinants are especially important.
In a two-year, grant-funded project, OCHIN sought to learn how to systematically track SDH in community health center EHRs.
“We worked with our community health center stakeholders to develop the tools,” Cottrell said. “They gave input on which domains to include, and how to best design tools to collect, summarize and review data in existing work flows. We also worked with stakeholders to identify and make referrals to community resources.”
Although the study focused on just three clinics, other members are eager to work with the SDH tools, so the researchers made them available OCHIN-wide.
Among the 14 SDH domains selected were questions about alcohol use, stress, sexual orientation, housing, food insecurity, tobacco use and exposure, race and ethnicity, as well as exposure to intimate partner violence. A few of these are already routinely collected in the EHR, and that data could be routed to SDH screens automatically to avoid duplicative data entry.
Cottrell said the data collection tools were designed to be flexible so that anyone on the care team could enter it. “We didn’t tell them to do it a certain way,” she explained. “We want to be adaptable to a number of workflows.” Most had patients fill out a paper questionnaire to be entered into the EHR later by staff members, although she noted that there is a MyChart patient portal version that could sent out prior to a visit and the data would automatically be transferred into the EHR, but she said that wasn’t used much during the pilot. (The lag time between the patient filling out a paper form and data entry into the EHR was sited as a barrier by study participants.)
OCHIN built a summary tool that allows staff members to see all the SDH information in one place and see trends over time. In this view, responses are highlighted if a need is identified or follow-up is required. “This is a nice way to get a snapshot of patient needs,” she said.
The project also sought to built referral tools for the pilot clinics. Clinicians want to find a way to do something with the SDH information. “There is so much national attention being paid to screening in primary care, which is great, but they are asking what do we do with the information?” Cottrell said.
There are groups creating online directories of social service agencies. “We thought it would be easy to connect the EHR to one of those organizations and make automatic referrals, but it is not an easy process,” she said. “We are still thinking about it, but it was beyond the scope of this pilot.” As a work-around, they created order sets with referrals to community services that included details on how to access those services.
The pilot also created data rosters of those who had already been screened and those who had not, as well as identifying which patients had said they did not want help with these issues.
The clinics had to work through what to do with positive responses without getting overwhelmed. If the patient did want help with SDH needs, did the clinic have something it could offer? The clinicians reported being concerned that if you ask about these issues, do you have an ethical obligation to do something about it?
Cottrell noted that this study was not trying to link SDH data with any changes in clinical outcomes, but rather just to study the feasibility of using the EHR tools. From June 2016 to July 2017 the researchers did on-site observations and interviews focused on clinical processes, and they also extracted EHR data on SDH data collection and referrals. Across the three clinics, 1,130 patients were screened, 211 had an SDH referral documented in the EHR. Social isolation was one of the top areas identified, along with financial resource strain, stress and food insecurity.
Among the key lessons of the pilot, she said, are that clinics must decide which patients to screen and how often. Also, the tool must be flexible enough to be used by diverse team roles. Clinics need a way to track referrals and document them. Also, the SDH data may be valuable in advocacy for policy change as well as for clinical purposes.
In its next step, OCHIN is working with the Kaiser Foundation Research Institute to expand this work to many more community health centers with a 5-year study called ASCEND (Approaches to CHC Implementation of SDH Data Collection and Action). Rachel Gold, Ph.D., M.P.H., an investigator at the Kaiser Permanente Northwest Center for Health Research with a joint appointment at OCHIN, is the principal investigator.
Cottrell said that OCHIN was aware that there are many efforts nationwide to create SDH screening tools. For instance, Epic is developing its own tools, she said. “We want to be as aligned as possible with other efforts,” she added. “There is so much more attention being paid to this, but there is some confusion that there is no national standard yet. We are working hard to streamline efforts and make sure our tool is as aligned as possible.”
