Your equity report shows nothing for Afghan patients because the data cannot see them — not because the disparity isn't there. Ariana Nexus stratifies race, ethnicity, and language data down to Pashto, Dari, and the other Afghan languages, so the population becomes measurable to the Section 1557 and NCQA standard.
Both the 1997 federal standard and the 2024 SPD-15 revision code Afghan patients as “Asian.” The aggregate satisfies the report. It also erases the signal.
Afghan patients are coded “Asian,” recorded as “Other,” or split so finely across 24 languages that no group registers — so the disparity never enters the analysis, while NCQA, HEDIS, and Section 1557 all assume it does. Ariana Nexus governs the data so the population becomes visible. What you can see, you can close.
U.S. residents who speak English less than “very well.”
Video telehealth use, by English proficiency.
The federal category that contains Afghan patients.
Population-level findings from the cited studies and federal publications — not Ariana Nexus client data. Verification pass completed June 2026.
Three institutional capabilities, run as one system that makes an invisible population measurable.
Afghan subject-matter experts define the categories — ethnicities, languages, dialects, sub-populations — and read a finding for what it is: an empty cell, a real gap, or an access barrier. Pashto and Dari are separated rather than merged, and Hazaragi is recorded as a dialect of Dari, not a language of its own.
The REL data model aligned to SPD-15 and the 24 Afghan languages, with Pashto and Dari carried as distinct fields rather than collapsed into one; record linkage, the stratification engine, and de-identified analytics.
Methodology and data-governance sign-off — categories valid, stratification sound, dialect and gender fields validated before they enter the model, and findings defensible to NCQA and an auditor.
Three capabilities. One population your system can finally see.
The Population Visibility Audit™ finds where the population disappears; the Five-Gate Validation Protocol™ governs every deliverable that follows.
Race, ethnicity, and language fields incomplete, defaulted, or inconsistently collected — the cohort does not exist in the record.
REaL data is captured, but Afghan patients dissolve into “Asian” or “Other” — the report is compliant, the disparity invisible.
Most systems sit hereDetailed origin and preferred language captured at granular codes, small cells governed by written suppression rules.
Quality, safety, access, and utilization measures stratified by cohort, with denominators an auditor could check.
Disparity found, intervention launched, effect re-measured — governance owns the delta.
Where the work leadsThe Population Visibility Audit™, delivered — where Afghan patients are mis-coded, collapsed, or absent, mapped with the fix.
A race-, ethnicity-, and language-data model aligned to OMB SPD-15 and the 24 Afghan languages, with Pashto and Dari held separate.
HEDIS measures stratified by race, ethnicity, and language, with a disparity dashboard benchmarked against the Diaspora Health Equity Index.
A Section 1557 evidence file — the data to demonstrate non-discrimination, if asked.
A board and quality-committee brief, and an NCQA-survey-ready package.
Population-level analytics only. Small-cell suppression by default. No individual-level data exposed. Every methodology and finding documented to an audit-grade standard.
Ariana Nexus is a Washington, D.C.–area firm providing Afghan language services and cultural intelligence — interpretation, translation, cultural training, compliance support, and AI data — across 24 Afghan languages.
Because the categories collapse them. Both the 1997 federal standard and the 2024 SPD-15 revision code Afghan patients as “Asian”, and language fields often carry a single “Other” value. The cohort is in your data; it has no cell of its own. That is a measurement failure, not an absence of disparity.
By making language a real field rather than a checkbox. Pashto and Dari are held separate, the remaining Afghan languages are coded distinctly, and Hazaragi is recorded as a dialect of Dari rather than a language of its own. Where cells get small they are suppressed, not merged into something meaningless.
It is the floor below which no cohort cell is ever reported. Small cells can re-identify patients, so they are suppressed rather than published. It matters because the tempting alternative to a suppressed cell is a merged one — and merging is exactly what made the Afghan cohort invisible in the first place.
Section 1557 requires non-discrimination, and NCQA and HEDIS expect measures stratified by race, ethnicity, and language. In practice you cannot demonstrate the first without the second: an equity file with no Afghan cohort in it does not show the absence of disparity, only the absence of data.
Yes. No patient-level data leaves your environment at any stage. The data model, the stratification logic, and the governance run inside your boundary; what comes out are governed aggregate findings and the methodology behind them.
All 24, with Pashto and Dari as the highest-volume fields and the rest coded distinctly rather than pooled. Dialect is captured where it changes meaning — Kandahari, eastern and central Pashto; Kabuli, Herati, Badakhshani and Hazaragi, a dialect of Dari. Uzbeki and Turkmeni are separate languages, not variants.