Skip to main content
Expert Talk
Amakomaya — Aamako Maya Apps Pvt. Ltd.

Case Study: DHIS2 & Health Information Systems in Nepal

Case studyBy , Co-Founder & CHIO, Aamako Maya Apps Pvt. Ltd.Published 12 min read

The problem​

Nepal migrated HMIS data to DHIS2 in 2011, and the platform has been the national reporting backbone since. But a national platform and a usable one are different things, and three gaps show up immediately at facility level.

The interface language. Health workers recording services at a health post are not working in English. A form they cannot read fluently gets filled in fast, approximately, and defensively.

The calendar. Nepal runs on Bikram Sambat. A reporting system that thinks in Gregorian months asks every health worker to convert dates mentally, every time, and asks every planner to reconcile a Nepali fiscal year against Gregorian aggregates.

The register. The actual system of record in most facilities is the paper register on the desk. Where software is a separate monthly transcription exercise, the transcription is where errors enter.

Architecture​

The work sits in the health information systems layer described in the reference architecture — beneath the products people use, above the exchange layer that moves data between systems.

Four pieces, each addressing one of the gaps above:

Nepali-language data capture — eRecord​

Localized DHIS2 data capture, in Nepali, for community and facility level. The substance is not translation but workflow: enrolment, the master register (मुल दर्ता), which services appear for which age group and sex, and the recording order that determines whether a person is counted correctly in the monthly report.

That last point is the one that costs implementations. Register and complete the master register before recording a service; do it the other way round and the validation warning gets clicked past and the person is missing from age-group counts. The full operational detail is in the eRecord user operations manual, written in Nepali for the people doing the work.

The Bikram Sambat calendar in DHIS2​

Implemented in DHIS2 2.36 across three surfaces:

  • the Capture app
  • the Android Capture app, including offline support
  • aggregate reporting on Nepali months, years and the Nepali fiscal year

This is more than a display concern. Bikram Sambat month lengths vary and are published by authority rather than computed, so a correct implementation carries authoritative period definitions rather than deriving them. Getting it wrong produces reports that are subtly misaligned with the fiscal year every planner uses.

DHIS2 runs as the national HMIS in more than 80 countries, which makes calendar localisation of this depth reusable well beyond Nepal.

Report generation​

Aggregation of facility-level data into the standardized forms used for routine national reporting, with consistency and completeness checks run before a report is confirmed — so problems surface while they can still be fixed at the source. See the HMIS Report Generator.

Digital HMIS registers​

From July 2025, a web-based application digitising the paper registers health workers complete at every facility, improving collection, management and interoperability within the national IHIMS. The Ministry of Health and Population owns the systems; GIZ leads the digitalisation programme under BMZ funding; Amakomaya is technical consultant and development partner.

This is the successor phase to the DHIS2 eRecord work, and it addresses the third gap directly: if the register itself is digital, reporting stops being a transcription exercise.

Implementation​

Municipal support. Technical support for health facility EHMIS implementation at Tokha Municipality — setup, maintenance and user support. This is the unglamorous half of DHIS2 work and the half that determines whether a deployment survives its first staff rotation.

Reporting pathways for other programmes. In 2025, a national reporting pathway for psychosocial counselling services: record sheets for community counsellors, submission forms to the Palika, a Palika reporting template feeding DHIS2, and report templates drawn from the national data. Worth noting what this is — a health informatics engagement, building and supporting a reporting system. It is not service delivery.

Continuity with national systems. The DHIS2 work sits alongside a run of national systems for MoHP from 2020 onward, including the COVID-19 Information Management Unit system and QR-based vaccine verification.

Lessons​

1. Metadata design is the expensive decision. Data elements and category combinations are extremely difficult to change once data exists. A rushed design is paid for every month afterwards. See metadata design principles.

2. The organisation unit hierarchy is not a facility registry. It works as one until another system needs the same list — at which point a hierarchy designed for reporting rollup is asked to be a registry, which it is not.

3. Analytics are scheduled. Data entered after the last analytics run does not appear in dashboards until the next one. A large share of "the system is wrong" reports are this, and the fix is explanation rather than engineering.

4. Validation without follow-up is decoration. A rule nobody acts on trains users to click past warnings — which is worse than having no rule.

5. Federal restructuring is a data migration. Identifiers that encoded geography inherited a problem when boundaries changed, and any indicator series crossing that boundary needs a documented break.

Implementation reality

The most useful diagnostic I know for a DHIS2 deployment is to sit with the person entering the data for one shift.

You learn things no architecture review surfaces: that the mandatory field everyone complains about gets a placeholder value, that a workaround has become standard practice across a district, that the register on the desk is the real system of record and the software is a monthly transcription exercise.

Every one of those has a downstream consequence that will later be described as a "data quality issue", as though it appeared spontaneously in the database. It did not. It was designed in, usually by someone optimising a form for reporting completeness rather than for the person filling it.

Which is why the highest-return DHIS2 intervention in Nepal is not a version upgrade. It is making the recording workflow useful to the recorder — a working patient list, a follow-up reminder, or simply not having to write the same thing twice.

Frequently asked questions​

Does Nepal use DHIS2?
Yes. Health Management Information System data migrated to the DHIS2 platform in 2011, and DHIS2 remains the national platform for routine health reporting across federal, provincial, local and facility levels.
Can DHIS2 support the Nepali Bikram Sambat calendar?
Yes. Bikram Sambat support has been implemented in DHIS2 2.36 across the Capture app, the Android Capture app with offline support, and aggregate reporting on Nepali months, years and the Nepali fiscal year. Because B.S. month lengths are published by authority rather than computed, a correct implementation carries authoritative period definitions.
What is the difference between DHIS2 aggregate and tracker data?
Aggregate data is counts for a period and an organisation unit — what routine HMIS reporting needs. Tracker data is individual-level, following a person through programme stages over time, which is what continuity of care requires. Both can coexist, but a national aggregate reporting platform is not a substitute for a clinical system.
Why do DHIS2 implementations fail?
Rarely on software. The recurring causes are metadata design that is expensive to change once data exists, an organisation unit hierarchy that does not match how the health system is administered, validation rules nobody acts on, and training treated as a one-off event rather than ongoing support.

Sources

  1. Paudel S, Paudel D, Boucher F. Digital Health in Nepal. Kathmandu University Medical Journal. 2025; 91(3): 386–91 — HMIS migration to DHIS2 in 2011. PDF
  2. DHIS2 documentation. docs.dhis2.org
  3. Nepal HMIS. hmis.gov.np
  4. Bikram Sambat implementation, digital HMIS registers, municipal support and the psychosocial reporting pathway: Amakomaya company record.
Last updated on