Most health systems cannot say how many Afghan patients they serve, which language each one speaks, or how their care turns out. Ariana Nexus builds the race, ethnicity and language data that makes Afghan patients visible, in Pashto, Dari and 22 more Afghan languages, and the outcomes analytics that shows where their care is falling short.
A consulting engagement delivered entirely by our own people, from the registration desk to the board report.
Reviewed by Tamana Ghaznawi, M.P.H., Senior Partner
Figure 1
The Afghan diaspora in the United States
Afghan immigrants living in the United States, 2010 and 2022
Federal standards disagree on Afghan patients: SPD 15 lists Afghan under Asian, while the CDC code set carried by EHR interfaces codes “Afghanistani” as White.
Language fields fail quietly. Across 13 primary care clinics, one in five patients recorded as English-speaking chose a Spanish-language survey.
Pashto and Dari each need their own entry, and dialect, interpreter need and interpreter gender each need their own field.
Health plans can stratify 22 HEDIS measures by race and ethnicity in 2026, and readmission penalties now count Medicare Advantage patients.
1 in 5
patients recorded as English-speaking in the EHR chose a Spanish-language survey, across 13 primary care clinics.
Klinger et al., Journal of General Internal Medicine, 2015
22
HEDIS measures can be stratified by race and ethnicity in measurement year 2026.
National Committee for Quality Assurance
Explained
It is the work of recording each Afghan patient’s self-identified origin, preferred language, dialect and interpreter needs correctly, mapping those fields to federal and industry standards, and reporting quality, safety and cost measures for Afghan patients as their own group. The result is a patient population you can count, reach and improve care for, instead of one averaged into a larger category.
Table 1
The same patient can be counted as Asian in one system and White in another, depending on which federal standard the software follows.
| Source | Category assigned | How Afghan appears |
|---|---|---|
| OMB SPD 15, the 2024 federal standard | Asian | “Afghan” is a listed example of another Asian group |
| CDC Race and Ethnicity Code Set, carried by EHR interfaces | White | “Afghanistani,” under Middle Eastern or North African |
| SEER cancer registry coding manual, 2026 | White (code 01) | “Afghanistani” |
| The patient | Their own answer | “Afghan,” plus any other group they select |
We record what the patient says, Afghan, and map it to every standard your reports need, so the same patient counts the same way everywhere.
Sources: Federal Register, Mar. 29, 2024 (FR Doc. 2024-06469); CDC National Center for Health Statistics, Race Code List, Appendix E; National Cancer Institute, SEER Program Coding and Staging Manual 2026, Appendix D.
Afghan patients are in almost every large U.S. health system, yet most records hide them. Few registration screens offer Afghan as an answer, Dari is often filed as Farsi, Pashto may be missing from the language list, and no field shows whether a Pashto speaker from Kandahar was matched with an interpreter who speaks her dialect.
Figure 2
3 measures
3 measures
3 measures
3 measures
3 measures
2 measures
3 measures
3 measures
Table 2
Seven fields turn an unseen population into one you can report on. Each is asked of every patient and answered by the patient, never inferred.
| Field | What the patient can answer | Standard | Why it matters |
|---|---|---|---|
| Self-identified origin | Afghan, plus any other group the patient selects | SPD 15 detailed category; CDC code crosswalk | Counts Afghan patients the same way in every report |
| Preferred spoken language | Pashto, Dari or one of 22 more Afghan languages | BCP 47 language code (ps, prs) | Separates Pashto from Dari and both from Farsi |
| Preferred written language | Pashto, Dari, English, or spoken only | BCP 47 language code | Many patients read a different language than they speak |
| Dialect or variety | Kandahari or central Pashto; Kabuli, Herati or Hazaragi Dari | Ariana Nexus value set | Drives interpreter matching and comprehension |
| Interpreter needed | Yes, no or declined | US Core Interpreter Needed (LOINC 54588-9) | Shows who needs support, not only who speaks another language |
| Interpreter gender preference | Female, male or no preference | Ariana Nexus value set | Recorded as the patient’s choice, not assumed |
| Country of birth | Afghanistan or any other country | ISO 3166 country code | Used only to confirm a cohort, never to label a patient |
Figure 3
A field that is missing at registration cannot be recovered in the report. This map shows which fields each report depends on, so the fix starts at the right desk.
| Field | HEDIS stratification | Readmissions by language | Community health needs assessment | Interpreter demand forecast | Research cohort |
|---|---|---|---|---|---|
| Self-identified origin | Required | Required | Required | Not used | Required |
| Preferred spoken language | Supports | Required | Required | Required | Required |
| Preferred written language | Not used | Supports | Supports | Not used | Supports |
| Dialect or variety | Not used | Supports | Supports | Required | Supports |
| Interpreter needed | Supports | Required | Supports | Required | Supports |
| Interpreter gender preference | Not used | Supports | Not used | Required | Not used |
| Country of birth | Not used | Not used | Supports | Not used | Supports |
Reports not listed under a field do not use it.
Ariana Nexus field map. Required: the report cannot be produced for Afghan patients without the field. Supports: the field sharpens the report.
DeliverableReadiness scorecard
DeliverableValue sets and crosswalks
DeliverableScripts and training module
DeliverableConfirmed patient cohort
DeliverableStratified outcomes report
DeliverableDemand forecast
DeliverableLive dashboard
DeliverableFindings brief
Figure 4
The readiness assessment places each organization on this scale and shows the work between one stage and the next.
Stage 0
Stage 1
Stage 2
Stage 3
Stage 4
Weeks 1 to 3
You receiveReadiness scorecard and fix list
Weeks 3 to 6
You receiveAfghan patient data standard
Weeks 5 to 10
You receiveTrained staff and a confirmed cohort
Weeks 8 to 14
You receiveOutcomes report and dashboard
Week 12 onward
You receiveAction plan and quarterly review
About three weeks, fixed fee
12 to 16 weeks
Quarterly
Scoped per study
A bilingual speaker can translate a word. Our people are matched to your patients on five points, and the same five points become fields in your data, so you can see which patients were matched and which were not.
Afghan Dari and Iranian Persian share a script, not a clinical vocabulary. Pashto is a separate language altogether.
| English | Iranian Persian | Dari | Pashto |
|---|---|---|---|
| Hospital | بیمارستان | شفاخانه | روغتون |
| Doctor | دکتر | داکتر | ډاکټر |
| Nurse | پرستار | نرس | نرس |
| Pharmacy | داروخانه | دواخانه | درملتون |
| Medicine | دارو | دوا | درمل |
Common usage shown; native-language reviewers confirm the terms for each engagement.
We ask every patient these questions so that everyone receives the right care.
Pashto
موږ دا پوښتنې له هر ناروغ څخه کوو، ترڅو هر چا ته سمه پاملرنه وشي.
Dari
ما این سوالها را از هر مریض میپرسیم تا همه مراقبت درست دریافت کنند.
Native-language reviewers confirm every script before registration staff use it.
Pashto and Dari are the most common. We record every language to the dialect, and we list Hazaragi as a variety of Dari that patients and interpreters treat as its own match.
Most analytics firms can build a dashboard. Few can hear the difference between Kandahari and central Pashto, or read an interpreter log and see which dialect went unmatched. We do both, in one team.
The national medical interpreter certifications test bilingual skill in only a handful of languages, and no Afghan language is among them. We train and assess our own interpreters and reviewers, and we train Afghan diaspora interpreters and students beyond our bench.
Our standardA college degree, assessment in the exact dialect, structured medical interpreter training, NCIHC ethics, annual HIPAA training, supervised entry and periodic review.
Alumni and scholars of leading universities who come from the Afghan community and work in public health, psychology, data engineering and AI.
Managing Partner
Grew up in Herat
Senior Partner
Lived in Kabul
Senior Partner
Lived in Kabul
Engagement Manager
Engagement Manager
Lived in Kabul
Analyst
Every engagement is staffed with a senior partner, an engagement manager, a data engineer and native-language reviewers matched to your patients’ languages.
Table 3
| Standard | What it sets | How we apply it |
|---|---|---|
| OMB SPD 15 (2024) | One combined race and ethnicity question, a Middle Eastern or North African category and detailed categories by default. Federal information collections must comply by September 28, 2029. | Afghan offered as a detailed answer and stored exactly as the patient gives it. |
| CDC Race and Ethnicity Code Set | Detailed codes carried by EHR interfaces; lists “Afghanistani” under White, Middle Eastern or North African. | A documented crosswalk, so Afghan patients roll up the same way in every report. |
| USCDI and HL7 FHIR US Core | Race, ethnicity, preferred language and an Interpreter Needed flag (LOINC 54588-9). | Pashto (ps) and Dari (prs) coded separately, with dialect and interpreter-gender preference in local fields. |
| NCQA HEDIS stratification | 22 measures can be stratified by race and ethnicity in measurement year 2026, including the Middle Eastern or North African category. | Afghan members reported inside the required categories and as a detailed group for internal action. |
| CMS Hospital Readmissions Reduction Program | From the FY 2027 program, the six readmission measures include Medicare Advantage patients. | Readmissions stratified by language, dialect and interpreter use at discharge. |
| IRS §501(r) and HIPAA | Tax-exempt hospitals assess community health needs every three years; de-identification follows 45 CFR 164.514. | An Afghan community profile for the assessment, with small numbers suppressed in every published table. |
We never ask about or record immigration status.
We never guess a patient’s ethnicity from a name or photograph. Any statistical estimate stays aggregate, labeled and out of the patient record.
We do not publish breakdowns by Afghan ethnic group, such as Pashtun, Tajik, Hazara or Uzbek, that could expose patients; language and dialect carry the clinical signal.
We do not route data or inquiries through channels controlled by the de facto authorities in Afghanistan.
We work alongside your counsel, privacy officer and accreditor, never in place of them.
REaL stands for race, ethnicity and language. It is the self-reported information hospitals, clinics and health plans collect so they can report quality, safety and access for each patient group. For Afghan patients, REaL data should include self-identified origin (Afghan), preferred spoken and written language, dialect, and whether the patient needs an interpreter.
It depends on the standard, which is why Afghan patients are hard to count. The 2024 federal standard, SPD 15, lists Afghan as an example under Asian, while the CDC code set used by many EHR interfaces codes “Afghanistani” as White, under Middle Eastern or North African. Ask patients to self-identify, record Afghan as a detailed answer, and keep every category the patient selects.
No. Dari and Iranian Persian (Farsi) are closely related forms of Persian, but everyday and clinical words differ: an Afghan patient says shafakhana for hospital and dawa for medicine. Record Dari as its own language, and match Afghan patients with Dari interpreters rather than Farsi interpreters.
Pashto and Dari are the most common. Afghan patients may also speak Uzbeki, Turkmeni, Balochi, Pashayi, one of the Nuristani languages, a Pamir language such as Wakhi or Shughni, or Hazaragi, a variety of Dari. Ariana Nexus works in 24 Afghan languages and records each one to the dialect.
We combine signals you already hold, such as language fields, interpreter request logs and country of birth, into a candidate list, then confirm with each patient at the next visit. We do not guess ethnicity from names, and no record is changed without the patient’s confirmation.
Readmissions, emergency department return visits, missed appointments, preventive screening, prenatal and postpartum care, blood pressure and diabetes control, depression and trauma screening with follow-up, interpreter use and wait times, and patient experience, each reported by language and dialect where the numbers allow.
Yes, with the right consent and governance. We analyze screening and follow-up for depression, post-traumatic stress and moral injury, which many families and clinicians call moral trauma, among patients who lived through the war in Afghanistan, and we pair the numbers with interviews in Pashto, Dari and other Afghan languages so care teams understand what the scores mean.
No. Immigration status is not part of REaL data, no analysis we run needs it, and asking about it keeps Afghan patients away from care. We advise every client not to collect it for this purpose.
We sign a business associate agreement before any patient data is shared, use the minimum data necessary, work inside your environment wherever possible, and release findings only as de-identified tables with small numbers suppressed.
Our data standard is system-neutral and maps to HL7 FHIR US Core. We prepare configuration requests for Epic, Oracle Health, MEDITECH and athenahealth teams, and we build reporting in Microsoft Power BI, Tableau or your EHR’s own reporting tools.
A data readiness assessment takes about three weeks at a fixed fee. A full program runs 12 to 16 weeks and is priced by the number of facilities, data sources and languages in scope. You receive a written scope and a fixed price before work begins.
Yes. Outputs are built to support NCQA race and ethnicity stratification, hospital readmission and quality reporting, The Joint Commission’s National Performance Goals, and the community health needs assessment that tax-exempt hospitals complete every three years.
Not one that tests the language. The bilingual exams from CCHI cover Spanish, Arabic and Mandarin, and those from NBCMI cover Spanish, Mandarin, Cantonese, Russian, Korean and Vietnamese. A Pashto or Dari interpreter can earn only a credential that never tests Pashto or Dari, so we assess every interpreter in the exact dialect and train to our own standard, for our team and for the Afghan diaspora interpreters we train.
Confidential from the first conversation. We sign your NDA, and a business associate agreement before any patient data is shared.
Request a data readiness assessment