Global health · Family caregiving
Noora Health
A nonprofit that turns hospital wards into classrooms and trains the relatives who will do the caring at home. Its newborn study, 133,733 families across 28 government hospitals, is among the largest ever run on caregiver education.
The ward as a classroom
In August 2025 an Associated Press photographer spent two days inside a government hospital in Gurugram, outside New Delhi. One frame shows pregnant women on plastic chairs while a nurse works through a flip chart. In another, patients queue outside a doctor’s room and watch the session while they wait.
The AP followed one woman home. Antima Kumari’s sister, Lilavati Devi, had just delivered twins and was too weak to care for them. Kumari stepped in. She had never handled a baby before. She had taken the hospital class.
The photographs show her settling a newborn against his father’s bare chest: kangaroo mother care.
The class has a name: the Care Companion Program, run by Noora Health, a US registered nonprofit, inside public hospitals across South Asia since 2014.
The gap it is aimed at
In a crowded Indian public hospital the average patient gets two or three minutes with a doctor, and discharge instructions arrive when the family is counting medicines and collecting bags. Co-founder Edith Elliott puts the retention problem in numbers: “Research shows that patients forget as much as 80% of what their doctor tells them as soon as they leave the hospital or a doctor’s visit.”
Meanwhile the corridors are full of families who have moved into the hospital for the length of an admission. Shahed Alam described it from the ward: “They get to see their family member for a couple of hours each day”. Then they are handed the patient and the entire job. The Audacious Project, which later funded the organization, notes that in South Asia alone, “nearly 1.5 million children under five die every year from causes that could be averted through health practices actionable at home.”
A class project that refused to end
Elliott and Alam met at Stanford in 2012. He was in medical school, she was studying global health policy, and both took the same course in human centered design. Each had a reason. When Elliott was 13 her mother was diagnosed with a brain tumor. “My mother had a very long surgery and an extremely long recovery.” Alam had watched his own mother care for his grandmother through a degenerative neurological condition.
Their team, which also included Katie Ashe and Jessie Liu, was sent to a low cost cardiac hospital in Bangalore to study patient flow. The answer they kept hearing was something else. Security guards, who stand in the wards all day, said the people doing the most for patients were the families. “So slowly the light came on that, wow, like this is an amazing resource within the system,” Alam said. In an intensive care unit they found a nurse, Anand Kumar, showing photographs to relatives so patients would not panic and pull out their tubes on waking. He later became Noora’s director of training.
By summer 2012 they had a pilot curriculum filmed in Kannada. It was supposed to end there. “We had no intention of starting an organization,” Elliott said. The hospital chairman asked them to roll it out across his network. They named the organization after a young mother whose baby needed heart surgery; Noora means light in Urdu. Y Combinator took them into its winter 2014 batch, unusual for a nonprofit.
Into state health systems
Scale came from governments. Punjab began with six district hospitals and went to all 22, run by Dr Baljit Kaur at the state health department. Across 2018 and 2019 the curriculum broadened from newborn care into general surgery and chronic disease, which made it relevant to most district hospital patients.
Noora now lists nine Indian state governments as partners: Andhra Pradesh, Goa, Haryana, Himachal Pradesh, Karnataka, Madhya Pradesh, Maharashtra, Odisha and Punjab. Add the national government of Bangladesh, seven regencies in Indonesia, and Nepal from 2024. In Bangladesh about 3,000 senior nurses and midwives from 565 hospitals are master trainers. In May 2025 the World Health Organization signed a three year memorandum of understanding with Noora Health.
What is proven, and what is still claimed
| Evidence | What the record shows | Source type |
|---|---|---|
| Newborn mortality | PLOS Global Public Health, May 2023. Pre and post design, 28 public tertiary facilities, four states, Oct 2018 to Feb 2020. Outcomes for 33,599 births before and 60,078 after. Mortality 41.3 then 33.6 per 1,000. Risk ratio 0.82, 95% CI 0.76 to 0.87; hospital-clustered 0.82, 95% CI 0.71 to 0.93. | Public record |
| The same study, fully adjusted | Adjusting for state as well as hospital clustering gives 0.93, 95% CI 0.82 to 1.06, not significant. No state estimate reached significance; Punjab trended the wrong way at 1.06. Maternal mortality did not move. The authors did not randomize and all data are self-reported. | Public record |
| The 18% headline | Noora Health, WHO and the Audacious Project all repeat an 18% reduction in newborn mortality without the qualifier. It comes from the hospital-clustered model, not the state-adjusted one. | Differs from the paper |
| Cardiac complications | Journal of Global Health Reports, 2019. One Kolkata cardiac institute, 2014 to 2015, 105 control and 110 intervention pairs, not randomized. Any at home adverse event at 30 days: adjusted odds ratio 0.29, 95% CI 0.13 to 0.66, the source of the 71% figure. Limitations: small sample, single site, self-reported. | Public record |
| Newborn readmissions | BMJ Open Quality, May 2022. Pre and post in 11 district hospitals, 1,474 before and 3,510 after. Readmissions risk ratio 0.44, 95% CI 0.31 to 0.61, the 56% figure. Complications down 16% for newborns and 12% for mothers, skin to skin care up 78%. | Public record |
| Randomized evidence | Bulletin of the World Health Organization, April 2025. Cluster randomized trial, 15 intervention and 9 control hospitals, 21,937 participants. Significant gains in 7 of 11 practices. Behaviors, not mortality. | Public record |
| Randomized evidence, COVID | Clinical Epidemiology and Global Health, 2023. Pragmatic randomized trial in Punjab. No effect on 21 day mortality, odds ratio 0.64, 95% CI 0.19 to 2.12; hospitalization down 48%, odds ratio 0.52, 95% CI 0.31 to 0.90. | Public record |
| Trial registration | One registered trial in the record set: NCT07232901 in Bangladesh, 1,029 participants, completed. The headline studies are not registered there. | Not found |
| Reach figures | 24 million+ trained by 2024 on the tenth anniversary page, over 30 million cited by WHO in July 2025, 65 million+ and 16,400+ facilities in 2026. None independently audited. | Company-stated |
Read plainly: the behavior evidence is strong and in one case randomized; the mortality evidence is large, real and softer than the marketing. A pre and post design cannot rule out secular decline, and the paper says so. What is unusual is that the authors published the weaker model too.
What to watch
- The full evaluation. The 2023 paper was a midpoint analysis paused from March 2020 to June 2022.
- The Bangladesh newborn unit results, presented to health officials in September 2025 but not yet peer reviewed.
- Whether reported reach and verified reach converge; the claim roughly doubled between mid 2025 and 2026.
In their words
“And you realize that family members, who are the people who care the most about helping their loved ones recover, are completely left out of the healing process.”
Edith Elliott, co-founder and co-CEO, SSIR podcast, 2021 · Interview
“So slowly the light came on that, wow, like this is an amazing resource within the system.”
Shahed Alam, co-founder and co-CEO, SSIR podcast, 2021 · Interview
“And even I did not know what I was doing was, you know, something that big.”
Anand Kumar, ICU nurse, later Noora director of training · Interview
“We had no intention of starting an organization.”
Edith Elliott, on the Stanford class project, 2021 · Interview
“We all feel that this was the need, and health education is the key to bringing about health-seeking behavior.”
Dr Baljit Kaur, Department of Health and Family Welfare, Punjab, 2021 · Government partner
“We have to look at being more efficient in addressing this critical gap,”
Seema Murthy, head of Noora Health’s India program, to the Associated Press, 2025 · Independent
“We want the Care Companion Program to be included as an integral part of health service delivery and owned by the government,”
Dr Arefin Amal Islam, Noora Health Bangladesh, The Business Standard, 2025 · Independent
“Noora is so effective that even if you measure their costs in the most conservative way, by dividing their entire budget by the number of lives saved, the cost of saving a life is the lowest I’ve seen.”
Paul Graham, donor and NFT buyer, 2021 essay · Donor
Related companies
Sources
- Public recordImproving neonatal health with family-centered, early postnatal care
- Public recordImpact of family-centred postnatal training in district hospitals in two states in India
- Public recordA mobile messaging service for families on postnatal knowledge and practices
- Public recordImpact of a family caregiver training program in Kolkata, India
- Public recordCan training over phone calls help improve outcomes for COVID-19 positive patients?
- Public recordWHO and Noora Health begin collaboration on family caregivers
- Public recordIRS Form 990 filings, EIN 46-4746592, fiscal 2014 to 2024
- Public recordClinicalTrials.gov NCT07232901
- IndependentPHOTO ESSAY: Newborn caregiving programs help ease parenting in India
- IndependentHospital-based health education improves maternal, neonatal care: Study
- InterviewThe Healing Force of Family, full transcript
- IndependentNoora Health case study with founder quotes
- DonorAn NFT That Saves Lives
- IndependentNoora Health profile, funding and grant history
- FunderNoora Health grantee page
- CompanyNoora Health recognized for its impact in transforming the way care is delivered
- CompanyImpact, Where we work, What we do and tenth anniversary pages
Profile researched and written by Healthcare Discovery. Last updated September 29, 2026.
