Housing, food, transportation, and safety shape whether someone can stay in treatment. That is an operational fact for a behavioral health program. It is not the same thing as a single percentage of “health outcomes” assigned to social factors.
What the county rankings actually weight
The University of Wisconsin Population Health Institute revised its model in 2025. The methods page uses the weights to build county summary scores, then sorts counties into health groups. The groupings are a way to compare counties. The page says the groups themselves do not necessarily represent statistically significant differences from county to county [1].
For the 2025 Population Health and Well-being summary, length of life is weighted 50% and quality of life is weighted 50% [1]. For the Community Conditions summary, health infrastructure is weighted 25%, physical environment is weighted 25%, and social and economic factors are weighted 50% [1]. The evolution page describes these as nominal weights for summary measures, replacing the weighting scheme of the 2014 model, and it says the new model does not display weights the way the 2014 model did [2].
Those percentages rank counties. They are not a measured share of one patient’s outcome, and they are not a license to write that clinical care accounts for a leftover fraction. The 2014 model is not the current one [1, 2].
What a payer can require, and what it cannot be assumed to require
A screening duty lives in a specific program: a contract, a quality measure’s specifications, or a state’s Medicaid policy. This review did not find a CMS rule that makes social-needs screening a general condition of participation for every Medicaid managed care plan. Do not write that sentence into a policy manual.
If a contract requires a named instrument, follow that contract and keep the instrument’s current form from its publisher. Item counts change. A template that hard-codes an old “core measure” count will drift from the tool. The Accountable Health Communities health-related social needs instrument was built for a CMS Innovation Center model. Availability of a tool is not the same thing as a standing requirement that every CMS-funded program use it. Confirm the model’s status and the measure’s status on the day you rely on them, and date the policy “as of” that month.
Accreditors publish their own standards. A general claim that “standards increasingly mention social needs” is not a citation. If a survey will score social-needs assessment, quote the standard you are surveyed under.
Z codes
ICD-10-CM includes Z codes for social circumstances. Reporting a Z code documents a circumstance in the coding system. It is not, by itself, a risk-adjustment payment. Before a billing policy says Z-code capture changes reimbursement, read the current CMS ICD-10-CM guidance and the payer’s risk-adjustment instructions. If those documents do not say the code changes payment, do not put that claim in the staff script.
A workflow that does not stop at the form
Screening without a next step leaves the need in the chart. A practical sequence is:
- Ask with the instrument the payer or the program actually adopted, and store the answers as structured fields.
- Flag a positive screen where the care team will see it, not only inside a scanned form.
- Record the referral: where it went, and whether the person got there.
- Write the follow-up in the plan of care, including when the need is unchanged.
None of that requires a product to invent codes. It requires the instrument, the payer’s rule, and a place to see the referral status. If software maps a screen to a Z code, the mapping is only as good as the coding guidance it follows, and a person should be able to see the code before it is submitted.
Sources
- County Health Rankings methods, including 2025 summary weights, University of Wisconsin Population Health Institute. Accessed September 28, 2026.
- The evolution of the County Health Rankings model, University of Wisconsin Population Health Institute. Accessed September 28, 2026.