Case Study: Digital Maternal Health in Nepal
How this project ran
- CARE NEEDPregnancy care is a sequence across several contacts; the record often did not follow the woman.
- CURRENT WORKFLOWContacts recorded separately, content in English, devices shared and low-specification.
- IMPROVED WORKFLOWNepali content reaching women through FCHVs and midwives, with structured capture at the visit.
- DIGITAL SOLUTIONOffline-first design, and two applications — one for mothers, one for health staff.
- INTEROPERABILITYIdentifier and terminology work, which integration forced and which we would now do earlier.
- IMPLEMENTATIONMunicipal deployments, partner-led rollouts, and a midwife-staffed counselling service.
- RESULTA service financed through local government and operating beyond its pilots.
The problem
Maternal health in Nepal has a continuity problem more than an information problem. Pregnancy care is a sequence — registration, antenatal visits, delivery, postnatal care, immunisation — spread across a health post, a birthing centre, sometimes a referral hospital, and a community health volunteer who may be the only constant. Each contact may be recorded. The record may not follow the person.
Three constraints shape everything built for this context:
Language. The audience reads Nepali. Health content written in English does not survive contact with a rural user.
Connectivity. Intermittent outside urban corridors, and expensive. In 2011, when this work found its first funding, Nepal's internet penetration was around 8%.
The device. Frequently shared, frequently low-specification, and frequently belonging to a family member rather than the pregnant woman herself.
Origin: a privacy finding, not a technology plan
The work began in 2004 with a digital information portal delivered through community tele-centres on the Nepal Wireless network in Myagdi and Kaski, established with community development leader Mahabir Pun.
In consultations with Mothers' Groups (Aama Samuha), a finding emerged that redirected everything: many women were uncomfortable seeking reproductive health information in a public tele-centre. Cultural and privacy reasons, not technical ones.
That is the founding insight of the platform. A shared village computer was the wrong place to ask a private question. The personal account, the phone, the offline download and the private call with a midwife all descend from it.
By 2008 the service was a multimedia web portal with audio, video and text guiding women through pregnancy, with medical content developed with Public Health Concern Trust (PHECT) Nepal. Then mothers started asking for something the team had not planned: copies of the videos on micro-SD cards, to watch privately on their own phones.
The users specified the mobile strategy before the team proposed it.
Architecture
Offline-first, and what it actually costs
Deciding that the client works offline is not a feature toggle. It means the client creates records before the server has seen them, which means client-generated identifiers, which means a reconciliation strategy and a conflict policy. Every later design choice inherits that.
It aged well. The alternative — assuming connectivity — fails precisely where the health need is greatest.
Two applications, not one
The 2013 Android work, funded by ISIF Asia, produced two applications rather than one:
- For mothers — stage-appropriate pregnancy content in Nepali, downloadable.
- For health staff — FCHV registration of pregnant women, and ANM access to patient information during clinic visits.
Splitting them was correct. The mother's app optimises for private, offline consumption at low literacy. The health worker's app optimises for fast structured capture during a visit. One application trying to do both does neither well.
The delivery layer nobody designs for
Nepal has roughly 50,000 Female Community Health Volunteers. They were the mechanism that made any of this reach a rural household: trained to use the device, registering women door to door, and playing content for women who would never have downloaded it themselves.
By the 2015 Internet Society case study the platform was recorded across 11 communities, with 1,061 women reached and 90 FCHVs equipped with smartphones and trained. A field survey in April 2017 recorded 1,500+ registered pregnant women, 67 FCHVs and 61 health posts across 5 districts.
Implementation
The first launch
The Android application launched formally at Ramkot Primary Health Centre, Kathmandu, on 3 October 2013 — the first Android maternal health application in Nepal. It was attended by the Kathmandu DHO Administrative Division director and the heads of Community Medicine & Public Health and of Gynaecology & Obstetrics at IOM/TU, alongside health workers, FCHVs and local leaders. Two Android devices were handed over for FCHV door-to-door use.
Launching with the university and the district health office standing behind it mattered more than the software did.
Makwanpur, and the first government money
Four sites followed in Makwanpur — Tistung, Bajrabarahi, Hadikhola and Sisneri — with devices handed to health posts and FCHVs trained on site. Monitoring visits in May 2014 found the video content generating more demand than the project could supply devices for.
Then something more durable than a pilot: Makwanpur District Government formally accepted the initiative and created a budget provision for it in its district plan for 2013/14 — the first instance of government financing, and the precedent for everything municipal that followed. Two VDCs procured mobile devices for all their FCHVs through local government.
Municipal scale
After federalisation the work moved to local government delivery: Bhirkot Municipality (Syangja), Tulsipur Sub-Metropolitan City (Dang), Manahari Rural Municipality (Makwanpur) and Pokhara Metropolitan City (Kaski).
The most complete was Khatyad Rural Municipality in Mugu, Karnali, on 22 March 2019: Amakomaya Care, Amakomaya Content and Vial-to-Child deployed together across 7 health posts, with 63 FCHVs and 60 phones plus 7 tablets, handed over at an orientation attended by a Member of Parliament, the municipal chairperson and the District Health Coordinator. Devices came from ENRD and UNICEF.
Mugu is among the most remote districts in the country. If a design works there, its connectivity assumptions are honest.
Partner-led deployment: One Heart Worldwide in Sarlahi
Municipal delivery is one route to scale. The other is an NGO that already runs a maternal health programme and adopts the platform inside it.
One Heart Worldwide has worked on maternal and newborn health in Nepal since 2010 and reports having impacted over two million mothers and newborns across 36 districts through its Network of Safety model. In Q2 2023 it trained 35 health providers from 13 health facilities across 2 municipalities of Sarlahi to use the Amakomaya platform — a cohort estimated to reach roughly 1,750 pregnant patients a year. The deployment used a translated version of the app to reach and track every pregnant and recently delivered mother in two rural municipalities of the district.
Sarlahi is a different problem from Mugu. It is Terai, not mountain; the constraint is not altitude or connectivity but language and population density. The districts along the southern border are Maithili- and Bhojpuri-speaking, and Nepali-language content that works in Kaski does not land the same way in Sarlahi. That the partner asked for a translated build before scaling is the same finding as the micro-SD cards a decade earlier: users specify the localisation, and they specify it earlier than you expect.
The partnership then did the thing that matters more than the pilot. In February 2024, One Heart Worldwide presented its experience and lessons from tracking pregnancies with the app at a national-level consultative meeting, alongside Pokhara Metropolitan City, Birendranagar Municipality and the Midwifery Association. Their recommendations were to strengthen the data recording and reporting system, and to run workshops with other government partners to scale the approach to further sites.
Two things are worth naming honestly about that. First, the recommendation was a criticism of our reporting layer, made in public, by a partner with a large field operation — which is the most useful kind of feedback available and directly informed later work on structured records. Second, an implementing partner carrying findings to a national forum is a better distribution mechanism than anything a vendor can do for itself. Pokhara appears on both lists — as a municipal deployment of ours and as a partner in that consultation — which is what adoption looks like when it stops depending on the supplier.
The call centre, which began as an emergency
In the first week of Nepal's COVID-19 lockdown in March 2020, women could not reach facilities, so midwives reached them by phone. That emergency response was made permanent: a toll-free line at Paropakar Maternity and Women's Hospital, co-initiated with MIDSON and NESOG, with running costs covered by Beyond Beijing Committee Nepal.
In July 2023 it became a contracted national service under a memorandum of understanding covering Sexual, Reproductive, Maternal, Newborn and Adolescent Health across all 77 districts of Nepal. The cost-sharing explains why it survives: PMWH owns and staffs the call centre with midwives; Amakomaya provides the platform, the connection, toll-free billing and staff training. Neither party pays the other.
Support for the hospital's website and online ticket system closes the loop — a mother calls the helpline, a midwife counsels and refers her, she books a ticket, and the outcome returns to the counselling record.
The application has passed 100,000+ downloads, with around 1,600 mothers engaged daily.
Lessons
1. The constraint is rarely the application. Language, privacy, device ownership and connectivity decided more outcomes than any feature did.
2. Human delivery beats distribution. FCHVs and midwives are why the technology reached anyone. A maternal health app with no human layer is a download statistic.
3. Government financing is the sustainability test. The Makwanpur budget provision in 2013/14 predicted which work would still exist a decade later.
4. Digitisation is a resilience argument. In the 2015 Gorkha Earthquake some health facilities lost all their paper records; data held on Amakomaya servers was recovered. Nothing in a business case was as persuasive.
5. Emergencies build permanent things — but only if the emergency build is treated as the start of a system rather than a stopgap.
What I would do differently
Registries before features. We solved identity when integration forced us to, which is later than optimal. Building the identifier model before the second system existed would have avoided reconciliation work that was entirely predictable.
Terminology earlier. Every free-text clinical field eventually becomes a mapping project. Binding to a code system at design time costs a week; doing it retrospectively costs a quarter and never fully succeeds for historical data.
Instrument the outcome, not the output. Downloads and registrations are outputs. The questions that matter — did the referral happen, did the visit follow — have to be designed into the record from the start, or you end up years later with impressive usage data and no ability to answer them.
Frequently asked questions
- What is digital maternal health in Nepal?
- It is the use of digital tools — mobile applications, health worker systems, telephone counselling and health information systems — to support pregnancy, delivery and postnatal care in Nepal. In practice it must handle Nepali-language content, intermittent connectivity and shared low-specification devices before it can address anything clinical.
- Why is offline capability important for maternal health apps in Nepal?
- Connectivity outside urban corridors is intermittent and expensive. Content designed to be downloaded and used offline reaches women that an online-only service cannot — which is precisely the population with the greatest need.
- What role do Female Community Health Volunteers play in digital maternal health?
- FCHVs are the delivery layer. Trained to use the devices, they register pregnant women door to door and play content for women who would not otherwise access it. In these deployments FCHVs were the mechanism that made the technology reach rural households at all.
- Which organisations have deployed the Amakomaya platform?
- Deployments have run through local government — Bhirkot Municipality, Tulsipur Sub-Metropolitan City, Manahari Rural Municipality, Pokhara Metropolitan City and Khatyad Rural Municipality in Mugu — and through implementing partners. One Heart Worldwide trained 35 health providers across 13 health facilities in two municipalities of Sarlahi in 2023, using a translated build to track pregnant and recently delivered mothers.
- Does Nepal have a national maternal health helpline?
- A toll-free counselling service for Sexual, Reproductive, Maternal, Newborn and Adolescent Health operates from Paropakar Maternity and Women's Hospital, staffed by midwives and covering all 77 districts under a memorandum of understanding.
Related
- Digital Health in Nepal — the national picture
- Health information systems in Nepal
- Case study: DHIS2 and health information systems
- Case study: HL7 FHIR and healthcare interoperability
Sources
- Amakomaya: mobile application to support pregnant women in Nepal. ISIF Asia Technical Report No. 2 (2013–14). PDF
- Internet Society Asia-Pacific case study (2015) — 11 communities, 1,061 women reached, 90 FCHVs equipped and trained. Project profile
- Deployment sites, municipal implementations, the 2013/14 Makwanpur budget provision and the toll-free service terms: Amakomaya company record and the signed memorandum of understanding with PMWH.
- Paudel S, Paudel D, Boucher F. Digital Health in Nepal. Kathmandu University Medical Journal. 2025; 91(3): 386–91. PDF
- One Heart Worldwide — Sarlahi training figures (35 providers, 13 facilities, 2 municipalities, ~1,750 pregnant patients a year) and the translated build. Quarterly Report Q2 2023
- One Heart Worldwide — February 2024 national consultative meeting, partners and recommendations on data recording and reporting. Quarterly Report Q1 2024
- One Heart Worldwide — organisational scale: founded 2010, over two million mothers and newborns impacted across 36 districts. oneheartworldwide.org