Using social data to help value-based care begins with ICD-10 codes

Twitter icon
Facebook icon
LinkedIn icon
e-mail icon
Google icon
 - data mining

The transition to value-based care may require making it simpler for information on social determinants of health to be pulled from electronic health records.

In an article published in the public health-focused Nov. 2016 issue of Health Affairs, Laura Gottlieb, MD, a family and community medicine professor at the University of California, San Francisco and her coauthors examined how aggregating social data may inform value-based payments, boost quality and drive strategies to improve population health from both inside and outside traditional care settings.

Gottlieb and her coauthors wrote there’s already ample evidence for tailoring care to individual patients based on social data, like identifying at-risk patients who may need cardiovascular screenings or deciding when phone consultations may be more effective than office visits. Readily available social data could also inform risk adjustments for performance-based payments from CMS.

“For example, if readmissions are more likely for individuals with inadequate housing, then data on patients’ housing status could enable risk adjustments of payment penalties based on the number of patients ‘diagnosed’ with that condition,” Gottlieb and her coauthors wrote.

Most population-based payment models also fail to factor in social determinants. There are a few exceptions, like Medicare Advantage using Medicaid eligibility as a measure of poverty, but Gottlieb and her coauthors said such a simply proxy is insufficient.

Utilizing social data could be challenging, the article said, thanks to existing ICD-10 codes.

The system could be the most effective way of incorporating social determinants into payment systems, if and when those factors become billable diagnoses and have existing codes attached to them. Some already match social domains covered in other screening tools, like the Z56 category for unemployment. In other cases, however, important social factors had no clear equivalent in ICD-10 codes.

“Responding to these code gaps could also inform ICD-11 development,” Gottlieb and her coauthors wrote. “One way to create these codes could include assigning specific uses for the category ‘other’ that is associated with many of the z-code domains of social determinants of health, such as ‘Other problems related to social environment.’

Some factors could fit under multiple codes, like challenges for a patient paying for food or housing, adding another obstacle to using social determinants. In the article’s conclusion, Gottlieb and her coauthors said healthcare professionals must begin integrating social data into care by first building a consensus about which ICD-10 codes to use.

“Identifying a clear process for collecting and aggregating data on social determinants of health is an important next step toward transforming health care, refining value-based payment, and ultimately influencing both health- and non-health-sector strategies to improve population health,” Gottlieb and her coauthors wrote.