IMPALA is continuous vital-signs monitoring for children's and newborn wards. Bedside monitors feed a central screen that alerts health workers when a child deteriorates, and local GOAL 3 teams handle installation, training and service. Over the past few years it has moved from pilot to routine use:
We think the evidence is now strong enough to scale while we keep testing it, and we scale through two routes. GOAL 3, a Dutch social enterprise founded in 2019, sells IMPALA directly to hospitals, governments and NGO programmes with a service contract. That contract pays for the repairs, software updates and refresher training that keep monitors in use, which donated equipment often lacks (Perry & Malkin, 2011). Investors, not donors, cover the company's losses while it grows. The GOAL 3 Foundation, set up in 2023, raises philanthropic money for where the need is greatest but the money isn't: wards in hospitals that can't afford the upfront cost, evidence, and entry into new countries.
In 2027 and 2028 we aim to grow from about 70 to about 400 hospitals. Most of that growth will come through direct contracts and programme funding. The Foundation's part is to reach the hospitals that can't pay the full cost, and to fund the evidence that helps everyone else decide.
For this, the Foundation's fundraising target is US$5.3 million: US$2.15 million in 2027 and US$3.15 million in 2028. This would cover about 150 wards, an independent evaluation, first steps in Uganda, Ethiopia and Nigeria, and the integration of Smart Triage and Smart Discharge into IMPALA. Nothing for 2027–28 is contracted yet, so this is also our current gap. The target is not a ceiling: our absorption capacity is higher, and how much more we can take on depends on what is funded.
| Amount (US$) | Time to live wards | What it pays for | Modelled deaths averted a year once live: RP central estimate (observed effects)[1] |
| ~US$31,850 [2] | 3–6 months (3–4 in hospitals already using IMPALA) | One children's ward: 10 monitors, a server, a central screen, installation, training and five years of support | ~14 (48–57) |
| ~US$96,000 | 3–6 months | Three children's wards in Malawi, Rwanda or Tanzania | ~40 (144–171) |
| ~US$300,000 (about US$700,000 with control sites) | 18–24 months | An independent evaluation of IMPALA in routine care across about 55 hospitals; protocol by Q2 2027, first results by end 2028 | Not applicable; evidence creation |
| ~US$500,000 | 6–18 months | About 20 wards in public and mission hospitals that can't pay the full cost (about 70% children's wards, 30% newborn units), ~47,000 admissions a year | ~220 (~800) |
| ~US$2.55 million | Hospital by hospital in Rwanda | Rwanda's national plan: newborn and children's wards in all 52 public hospitals | Not estimated |
| ~US$5.3 million | 2027–2028 | Our full plan: about 150 wards, the evaluation, first steps in Uganda, Ethiopia and Nigeria, and integration of Smart Triage and Smart Discharge | Not estimated |
Most of the gap is here. The Foundation pays for wards in public and mission hospitals that treat very sick children but can't pay for IMPALA upfront, even where it would lower their costs over time. Each hospital still pays at least 10% of the five-year cost and takes on the yearly service fee, so it has its own reason to keep the ward running. The Foundation board approves every allocation, and a project only starts once it is fully funded.
The first wards ready are children's wards in about 20 hospitals that already use IMPALA in their newborn unit. Each would add 6–8 monitors to the existing server and could be in use three to four months after funding.
Our evidence comes from a handful of studied hospitals. About US$300,000 would fund an independent evaluation of how IMPALA performs in routine care across about 55 of the hospitals already using it, or about US$700,000 with control sites.
The evaluation is observational. With research partners, we are also scoping a randomised control trial, including whether randomisation can be built into a funded rollout, which would be cheaper and faster than a stand-alone trial. If you fund or design trials and would like to be involved, please get in touch.
Rwanda's Ministry of Health, Rwanda Biomedical Centre and GOAL 3 have agreed a national plan to equip newborn and children's wards in all 52 public hospitals, costed at US$2.55 million and fundable one hospital at a time. Several large referral hospitals we work with started with 10–20 monitors but have room for 200–400 or more; covering a whole hospital costs roughly US$0.5–1 million and can be done ward by ward.
Part of the full plan also goes to integrating Smart Triage and Smart Discharge into IMPALA. Both tools were developed at the University of British Columbia. Smart Triage helps health workers decide which children need care first when they arrive at hospital, and Smart Discharge identifies children at high risk of dying after they go home, so follow-up can focus on them. Built into IMPALA, they would cover the moments before and after a child's time on the ward.
Wards that go live sooner start helping children sooner. Beyond that, three things make funding in the coming months worth more than the same funding later:
You can give to the GOAL 3 Foundation here: https://www.goal3.org/foundation#donate-now
The GOAL 3 Foundation is a Dutch stichting with ANBI status (RSIN 864979976), so gifts are tax-deductible in the Netherlands. It also holds an NGOsource Equivalence Determination and uses fiscal sponsorship programs for tax-deductible donations from other geographies such as Canada, the United States, United Kingdom, Switzerland, Germany, Australia, and New Zealand.
If you're interested in our work, whether or not you're considering a donation, please reach out at jelle.schuitemaker@goal3.org. We're happy to share the model, walk through the evidence, or hear where you think we're wrong. We'll answer questions in the comments as well. Our evidence, costs and governance are all on our public knowledge hub. If you are curious to get a quick deep-dive about the work of GOAL 3 and the GOAL 3 Foundation, feel free to attend our end-of-year webinar where we elaborate further; register here.
The content comes from our own written public knowledge hub, Rethink Priorities’ report and our internal planning documents. We used AI to help combine and condense that material, then went through several rounds of revision, checked each number against its original source, and rewrote the parts that didn’t sound like right. Any errors are ours, and we’d be grateful if you point them out.
[1] Modelled, not counted. Single-ward rows assume a children's ward with 3,000 admissions a year and the 3.8% baseline mortality in RP's model. The central figure applies RP's discounted 12% reduction; the bracket applies the 1.6 and 1.9 percentage-point reductions observed in two Malawian hospitals, which describe those hospitals rather than forecast new ones. The US$500,000 row uses our planned 70/30 mix of children's wards and newborn units, with RP's discounted effect for each.
[2] A ward costs US$31,850 over five years: a US$24,850 one-off package plus five years of the US$2,000 yearly platform and support fee, which includes a US$3,000 discount when the five years are paid upfront. Hospitals pay at least 10% of this, and 20% on average. Single-ward rows show the full price. The US$500,000 and US$5.3 million rows assume the average hospital share (about US$25,480 per ward from the Foundation), which is why US$500,000 covers about 20 wards rather than 16. Amounts exclude in-country transport and, where exemptions don't apply, VAT.