How should we capture social determinants of health as coded data? CDC invites public comment on ICD-10 codes for SDOH

April 8, 2019 / By Rhonda Butler

I’ve got a two-part (maybe three-part) homework assignment for you:

  1. Read a pair of proposals for new ICD-10 codes, presented at the March C&M meeting. They are each two pages long. The first one, “Food Insecurity,” starts on page 20, and the second one, “Social Determinants of Health,” starts on page 44.
  2. Watch the YouTube video of the presentation (You can witness an industry first—the CMO of United Health Care asking for new ICD-10 codes) and the spirited discussion that follows, on effectively capturing SDOH data. The presentation by the United Health Care CMO starts at 1 hour 44 minutes in the video, lasts about ten minutes, and the discussion that follows is another 30 minutes.
  3. If you are inclined to comment on the ICD-10 proposals or the general subject of codes for SDOH, send your written comments to NCHS/CDC at nchsicd10CM@cdc.gov.

The deadline for accepting comments is May 10. The discussion ranged beyond the specific ICD-10 code proposals, so feel free to comment on what you heard on the video as well.

Reading the proposals and watching the video takes about an hour. If you are anything like me, the writing will take…as long as it takes, and always longer than you would wish. So why do I think this is important enough to ask you to spend the time? Judge for yourself. Read this snippet from United Health Care’s written request for new ICD-10 codes and see if you agree with me that the subject is worth checking out. I’ve bolded the text where I thought, Whoa, I would have loved to be there when the people involved were choosing these words (carefully).

“UHC believes expanding the ICD-10-CM code set would allow for population health improvement, along with the opportunity for National Committee for Quality Assurance (NCQA) to expand HEDIS measurements around social barrier identification and assistance in the future. Today, capturing SDOH barriers appear to vary widely throughout the industry, rendering it a fragmented, inconsistent way of both capturing and using this information.

Systematized Nomenclature of Medicine-Clinical Terms (SNOMED-CT) and Logical Observation Identifiers Names and Codes (LOINC) are other coding standards that have also identified some SDOH codes, however for varying reasons, they would be a substandard solution.

UHC believes utilizing the ICD-10-CM codes is a logical choice, as it is the standard language between care providers and payers. The existing SDOH range Z55-Z65 in the ICD-10-CM has been labeled as ‘Persons with potential health hazards related to socioeconomics and psychosocial circumstances,’ validating that an expansion of these codes would be warranted within the ICD-10-CM classification.”

To whet your appetite for more, here is a tasting menu of proposed new ICD-10 codes intended to capture specific socioeconomic risks to an individual’s health:

Z59.41     Lack of adequate food

Z91.110    Patient’s noncompliance with dietary regimen due to financial hardship

Z56.84     Unemployed but not seeking work

Z59.61     Unable to pay for prescriptions

Z59.64     Unable to pay for transportation for medical appointments or prescriptions

Z60.82     Inadequate social interaction – limited to once or twice a week

I do hope you will check this stuff out for yourself. If you want to join the discussion, email your comments to NCHS/CDC at nchsicd10CM@cdc.gov by May 10. Donna Pickett of NCHS/CDC made it clear at the meeting that the SDOH proposals are preliminary—if any new ICD-10 codes are created along these lines, they would be added for FY 2021 (at the earliest). With preliminary C&M proposals, typically a revised version of the proposal that takes account of comments received is presented at the next C&M meeting, so there is likely more to come in September.

Rhonda Butler is a clinical research manager with 3M Health Information Systems.


Striving to achieve health equity by addressing social and clinical risk.