Most hospital language-access policies were written before the 2016 Section 1557 update and haven’t been touched since. The checklist below is what an OCR investigator actually asks for when they show up.
The four documents you need on file
- Language-access plan: a written policy naming the languages you serve, how staff identify LEP patients, and the workflow for requesting an interpreter. One page is fine. The plan must be reviewed annually.
- Vendor contracts with interpreter qualifications: your interpreter roster (in-house or contracted) must document each interpreter’s credentials. “Bilingual nurse” is not a credential.
- Notice of right to interpretation: posted in the top 15 languages of your service area, in the patient registration area, on your website, and on any form a patient signs. The HHS-translated taglines are free.
- Interpretation event logs: every interpreted encounter, dated and timed, with language and modality. This is the document an OCR investigator opens first. Most hospitals can’t produce it.
Where most policies fail
The plan exists. The contracts exist. The notice is posted. The logs are missing, or they exist in three different systems and nobody can produce a clean export.
This is the single most common Title VI finding: organizations that had interpretation but couldn’t prove it had been interpreted by a qualified person on a specific date. The remedy is uniformly painful: voluntary resolution agreements, monitor visits, staff training mandates.
What “good” looks like
A clean program produces, for any patient encounter on any date, a single line of evidence: interpreter name + credential + modality + duration + patient identifier. If you can produce that line in under five minutes, you are above the median.
If you can’t, the issue isn’t policy. It’s that interpretation lives in a spreadsheet, an email thread, or a vendor portal that doesn’t talk to your EHR. The next step isn’t a new policy. It’s a system that produces the log as a byproduct of the work.