1The problem we are solving
Most countries in transition still run health on paper, point products, and donor-funded silos. The cost is invisible until a child gets the wrong vaccine schedule, an outbreak is detected weeks late, or a Ministry cannot answer a basic question about its own population.
Across Afghanistan and the broader region, the typical health facility runs on a stack that grew by accident:
- A paper register for outpatient visits, often the only true source of truth.
- A standalone HMIS data-entry tool for monthly aggregate reporting to the Ministry.
- An independent immunization tracker (often Excel) for EPI campaigns.
- A separate lab tool for results and a different system for pharmacy.
- A donor-mandated reporting tool layered on top — usually entered twice.
The consequences compound:
- Patients repeat their history at every visit; allergies and chronic conditions get re-asked, re-missed, re-prescribed.
- Vaccines are duplicated, missed, or expired in cold-chain because no system tracks the dose and the recipient.
- Outbreaks (cholera, measles, COVID-class events) are detected from monthly paper aggregates instead of from real cases.
- Maternal & child indicators live in spreadsheets that disagree with each other.
- Donors and Ministries get late, partial, and uncomparable data — and lose confidence in the system that provides it.
Ghasi-eHealth replaces that fragmented stack with one platform. Identity is shared. The chart, the lab, the pharmacy, the immunization register, the patient app, and the public-health export are views of the same record — governed under one access policy, one audit log, and one consent model.
2National eHealth vision — Afghanistan as the flagship
Ghasi-eHealth is designed, from day one, to be the digital backbone of national health — first in Afghanistan, then across aligned deployments. The vision, drawn directly from the platform's own normative documents, is:
Vision statement (normative): Enable safe, standards-based, equitable digital health for Afghanistan and aligned deployments — one backbone that connects facilities, pharmacies, labs, public-health reporting, and research, under national governance, with FHIR-first interoperability and auditability.
Source: NATIONAL_EHEALTH_VISION.md · Audience: Ministry of Public Health (MoPH), architecture, program management, donors.
2.1 Patient → System → Country: one lifecycle, one record
The platform follows the patient through every meaningful encounter, and rolls those encounters up — responsibly, with consent and de-identification where required — into the indicators a Ministry actually needs to govern public health.
2.2 Why "national-scale" is more than rhetoric
| National capability | How the platform delivers it |
|---|---|
| National patient identity | Single longitudinal patient record per tenant / per program node, with deduplication, identifier policy, and configurable national ID schemes. |
| Standards-first interoperability | FHIR R4 as the canonical exchange model; HL7 v2 adapters for legacy lab/pharmacy systems; IHE profiles where required; DICOM/DICOMweb for imaging. |
| Public-health reporting | De-identified aggregates feed an HMIS-aligned indicator catalog (encounters, immunization coverage, NCD prevalence, screening compliance, e-prescribing volume) — derived from clinical truth, not duplicated. |
| National governance & auditability | Tamper-evident audit on every clinical action; access policy is data-driven, not hard-coded; consent is first-class. |
| Equity & reach | RTL-first UI for Pashto, Dari, and Arabic alongside English; Solar Hijri / Hijri / Gregorian calendars; offline-tolerant clinical surfaces for low-connectivity facilities. |
| Sovereign deployment | Multi-tenant with data residency options; the national program can run in-country while donor partners get the aggregate views they need. |
3Population health, vaccination, and disease surveillance
This is where Ghasi-eHealth changes the equation for a Ministry of Health: clinical truth at the bedside becomes public-health truth at the country level — automatically, governably, and on time.
3.1 Immunizations — every dose, every child, every schedule
The platform's immunizations module is built for the operational reality of national EPI programs:
Full administration record
Every dose captured with vaccine code, date, dose number, lot, manufacturer, route, site, performer, and location — the data a recall or cold-chain investigation actually needs.
Refusals & contraindications
Refusals captured with reason; contraindications captured as a structured workflow distinct from refusal — so coverage denominators reflect reality.
Schedule governance
Forecasts are computed against jurisdiction- and tenant-specific schedule versions. Afghanistan can run its own EPI schedule today, and update it tomorrow without a software release.
Forecast-driven outreach
Due / overdue indicators feed reminder workflows, so the question stops being "did the campaign happen?" and becomes "which child is overdue, in which district, today?"
Registry interoperability
Bidirectional sync with national / regional immunization registries with acknowledgment and reconciliation states — no more parallel paper books.
Digital immunization certificates
Signed, exportable certificates when policy permits — useful for school entry, travel, and donor reporting alike.
3.2 Disease registries & population analytics
The health-population service brings the same modern population-health pattern Western health systems use (HEDIS-style quality programs, chronic disease registries, risk stratification) to a context that has rarely had it.
| Capability | What it enables |
|---|---|
| Chronic disease registries | Diabetes, hypertension, COPD, asthma, CHF, CKD, mental health, and obesity — each with control status (e.g. A1C, BP), last visit, last labs, active medications, risk score, care gaps, and outreach status. |
| Population dashboard | Active patients by node, age and gender histograms, chronic disease prevalence, immunization & screening compliance, high-risk patient counts and missed-appointment indicators — scoped to tenant, facility, department, or provider. |
| Cohort builder | Predicates across demographics, diagnoses, medications, labs, vitals, immunizations, encounters, SDOH, custom tags, and risk; nested AND/OR/NOT; versioned definitions; saved and shared cohorts. |
| Risk stratification | Risk engine computes tier and drivers for eligible patient sets; care-gap engine detects due / overdue interventions using registry and screening rules. |
| Quality reporting | Quality-measure snapshots on configured cadence with numerators, denominators, exclusions, and trend points — exportable, drill-downable, role-constrained. |
| Outreach operations | Auto-generated or curated outreach lists assigned to teams — closing the loop between "the data shows a gap" and "someone is following up". |
3.3 From clinical truth to HMIS reporting — without duplicate entry
Illustrative indicator families the platform feeds (to be nationalized with MoPH):
| Family | Example metrics | Platform inputs |
|---|---|---|
| Service utilization | Encounters per facility · new vs. follow-up | Registration · scheduling · encounters |
| Maternal & child | ANC visits · immunization coverage (aggregate) | Immunizations · health-population |
| Non-communicable disease (NCD) | Diabetes / hypertension registry counts · control status | Health-population registries · problem list |
| Quality & prevention | Screening compliance · care-gap rates | Health-population FR-POP series |
| Medication safety | E-prescribing volume · dispense linkage | E-prescribing-gateway · pharmacy |
3.4 The impact, in plain terms
The platform shifts public health from monthly retrospective paperwork to operational visibility. A district health officer can see today which villages are behind on a vaccine, which clinics have flagged febrile clusters, and which mothers are overdue for ANC — and act on it.
4Care delivery — one chart, one safe story
For the clinician, none of the platform's national reach matters if the bedside experience is bad. So the Core EHR is built to be the always-on clinical baseline that every facility starts with.
4.1 The always-on clinical baseline
Patient management
Registration, identifiers, search, deduplication; demographics and emergency contacts; configurable national-ID handling.
Scheduling
Provider directory, capacity-aware booking, appointment lifecycle, reminders, and walk-in flow.
Clinical chart
Encounters, vitals, problems, allergies, medications, notes, results — assembled as one longitudinal story.
CPOE (orders)
Computerized provider order entry for labs, imaging, procedures, and medications, with safety checks.
Results review
Lab and radiology results flow into the chart with status, abnormality flags, and acknowledgment.
Medication safety
Structured allergies, problem lists, drug interaction surfacing where coding is mature.
4.2 Modular extensions — pay for what you use
On top of the Core EHR, facilities and programs can license what they actually need. There is no duplicate patient index across products; every module is a view of the same record under the same identity, audit, and access policy.
| Module family | Highlights |
|---|---|
| Diagnostics | Laboratory (LIS) and Radiology (PACS): order → specimen / acquisition → result / report → back into the chart, with full status traceability and DICOM imaging where deployed. |
| Medication safety | Pharmacy and e-prescribing gateway for structured prescribing, dispensing, and (where regulated) electronic transmission to pharmacies. |
| Revenue & insurance | Encounters, billing, insurance, and claims as licensed add-ons for facilities that need them — without forcing private-billing logic on public clinics. |
| Patient engagement | Patient portal, secure messaging, and virtual care surfaces (where licensed) so care teams and patients stay aligned between visits. |
| Population health | Cohorts, registries, risk stratification, quality measures, outreach (see §3). |
| Immunizations | Administration, refusal, contraindication, forecast, registry interop, certificate (see §3). |
| Interoperability | FHIR R4 gateway, HL7 v2 adapters, IHE profiles where required — the system other systems can talk to. |
5Who Ghasi-eHealth serves
| Role | Outcomes we enable |
|---|---|
| Clinicians & care teams | Faster rounds, safer orders, complete chart at the bedside; mixed-direction (RTL/LTR) UI that does not fight the language of care. |
| Registration & scheduling | Clean identifiers, capacity-aware booking, fewer duplicate records, fewer turned-away patients. |
| Labs & radiology | Order → result loop with clear status and traceability; results back to the ordering clinician without phone calls. |
| Pharmacy | Structured prescriptions, dispensing, and (where enabled) electronic transmission — instead of illegible paper Rx. |
| Patients & families | Access to appointments, education, immunization records, and the channels you choose to open — in their language, on their phone. |
| Public-health officers | Live coverage and surveillance views; cohorts and care-gap lists they can act on; HMIS exports without re-keying. |
| Ministry & program office | Policy-aligned national reporting, donor-aligned indicators, and a single sovereign data plane. |
| Donors & partners | Trustworthy, governed aggregates that arrive when they're due — replacing "the data is being compiled" with "the dashboard is live". |
| IT & security | Service-oriented architecture, multi-tenancy, fine-grained audit, FHIR alignment, and a documented standards posture. |
6AI in healthcare — useful, governed, never autonomous
Clinical AI on Ghasi-eHealth — documentation support, triage assistance, decision support hints, registry classification, and orchestration — runs through a single governed AI gateway. That gateway enforces:
- Provenance on every output — model, version, prompt category, and decision linkage are recorded.
- Access control and consent — who can use which model for which purpose is policy-driven, not implicit.
- Pre- and post-moderation — PII redaction, safety filtering, and refusal handling are uniform.
- Per-tenant budgets — AI cost cannot run away from a facility or program.
- Review paths appropriate to regulated care — high-impact suggestions require a human decision before they affect the chart.
Automation never replaces clinical judgment or informed consent where the law and your policy require a human. AI is an assistant that drafts, suggests, and surfaces — not an actor that prescribes.
7Equity, language, and reach
The platform was designed with the operational reality of Afghanistan and the wider region in mind — not retrofitted to it.
Languages & direction
English, Pashto (RTL), Dari (RTL), Arabic (RTL) as Phase 1 languages. RTL is a first-class citizen, not a CSS afterthought. Mixed-direction content (Arabic name + Latin MRN + lab units) renders correctly.
Calendars & time
Gregorian, Islamic Hijri, and Solar Hijri calendars configurable per tenant; Asia/Kabul and Asia/Dubai time zones supported as defaults; ISO 8601 storage in UTC.
Currencies & fiscal reality
Multi-currency per tenant including AFN and AED; tax/VAT configurable per tenant and effective date — no hardcoded rules.
Print & portable evidence
Printed artifacts (charts, prescriptions, immunization certificates) support RTL shaping, ligatures, mixed-direction segments, and embedded fonts — so a paper handoff still works.
Communications
SMS, email, and portal templates support all four Phase 1 languages with correct number and date formatting per locale.
Adaptive integration
External integrations (payers, registries, authorities) vary by jurisdiction — handled via adapter-based design, not hardcoded country logic.
8Why Ghasi-eHealth — and why now
| The siloed status quo | The Ghasi-eHealth answer |
|---|---|
| Imported foreign EHR — heavy, expensive, not designed for Pashto/Dari/Solar Hijri or low-connectivity facilities. | Built for the region from day one: RTL-first, multi-calendar, multi-currency, offline-tolerant clinical surfaces, modular licensing. |
| Open-source EHR retrofitted with bolt-on modules — fragile interfaces, weak audit, unclear roadmap. | One platform, one identity, one audit log, one access policy across EHR + immunizations + population health + portal + e-prescribing. |
| Donor-funded "vertical" tools (HIV, TB, EPI) that do not talk to each other and re-collect the same patients. | Vertical programs as views and modules of the same record; aggregates derived once, not re-entered five times. |
| Reporting built on Excel exports compiled monthly. | Live, governed HMIS-aligned indicators derived from clinical truth — and ready when the Ministry asks, not when the spreadsheet finally arrives. |
| "AI tools" that drop into the workflow with no provenance, no audit, and no human-in-the-loop. | Single governed AI gateway with provenance, consent, budgets, and a clear rule: AI assists; humans decide. |
| Vendor lock-in to a closed data model; integration is a billable change request. | FHIR R4 first; HL7 v2 adapters where the world is still legacy; documented standards posture; data the customer owns. |
| "Sovereign deployment" as a marketing slide. | Multi-tenant with data residency options; the national program runs in the country it serves; donor partners get the aggregates they need without owning the patient data. |
9The impact we are aiming for
- Every encounter at a participating facility produces a structured, longitudinal record — not a paper page.
- Every administered vaccine is linked to a specific child, dose, lot, and schedule — usable for recall, coverage, and certification.
- Public-health surveillance moves from monthly retrospective aggregates to operational, near-real-time dashboards.
- Chronic disease registries (diabetes, hypertension, mental health) become actionable outreach lists instead of statistical estimates.
- Maternal & child indicators (ANC visits, immunization coverage) become per-village, per-week views with names attached — under proper consent and access control.
- The Ministry of Public Health gets a single, governable, sovereign data plane — and donors get trustworthy reporting derived from it, not re-collected next to it.
One backbone. Every facility. The health of a nation.
Deep-dive service bundles, EHR module catalog, FHIR-first standards, and the national vision documents live in the Ghasi-eHealth documentation tree — for architecture review, donor narrative, and implementation planning.