In recent years the federal government has used financial incentives and penalties to promote use of electronic health records (EHR), but in these regulations the issue of EHR data interoperability — that is, data standardization — has been left largely unresolved, dimming prospects for data exchange, population health surveillance and systematic improvement across health networks.
This interoperability impasse has been top of mind for the Committee on Recommended Social and Behavioral Domains and Measures for Electronic Health Records. Co-chaired by William Stead, M.D., of Vanderbilt, and Nancy Adler, Ph.D., of the University of California, San Francisco, the committee was impaneled in 2013 by the Institute of Medicine of the National Academy of Sciences.
On Thursday the committee published its recommendations.
“We’re talking about a handful of standard measures that can be implemented in the EHR the same way everywhere. By standardizing the measurement, we build interoperability in from the beginning, and from my perch that’s the huge breakthrough,” said Stead, McKesson Foundation Professor of Biomedical Informatics, associate vice chancellor for Health Affairs and chief strategy officer at Vanderbilt University Medical Center.
According to the committee’s initial report (issued in April), the best available evidence suggests that health behaviors and social conditions account for more than half of all premature deaths in the United States.
“The United States has poorer health status than most developed countries and finishes near the bottom when it comes to dealing with social and behavioral determinants of health,” Stead said.
The basic premises adopted by the committee led initially to “more than 70 social and behavioral domains and subdomains, where evidence of the association with health was clear, and if you had the data it was reasonable that you could use it for individual patients or population management or research,” Stead said.
Settling on the 17 most promising domains, “What we said is, we’re only going to deal with things that have a scientifically proven standard measure that’s freely available.”
Also considered were feasibility of record keeping (time, cost), and whether patients would be uncomfortable releasing information.
The committee arrived at a short list of domains and measures for routine inclusion in the EHR: race/ethnicity, alcohol use, tobacco use/exposure, residential address (used to infer estimated income), physical activity, educational attainment, social connections/isolation, depression, stress, financial resource strain and exposure to violence.
The committee concluded that the information could be captured with as few as 25 self-reported patient questions.
With Web-based patient portals (such as My Health at Vanderbilt), the information could be gathered from home before the patient encounter.
In an era of financial incentives tied to hospital readmission rates, social and behavioral determinants of health presumably have new influence on hospital financial performance.
The report addresses potential barriers to implementation, and sets out the implications of each domain for patients, clinicians, health systems and researchers.
The report is out in time to influence the final version of forthcoming EHR regulations from CMS (Center for Medicare and Medicaid Services).
“This is about personalized care, more holistic care. We need to help the public get comfortable with the fact that their clinician ought to know and be aware of these factors.
“The clinical team may not be the right people to intervene, but they need to connect the patient with the right intervention. We’re going to have to provide education for the care team on how to work with these questions,” Stead said.
The committee also recommended further research with regard to measures that, while important, happen currently to be less feasible to record and less clinically actionable.
