Rotavirus vaccines: half the protection, still a huge win

After India added its home-made rotavirus vaccine, Rotavac, to the Universal Immunization Programme, the share of hospitalised children with diarrhoea who tested positive for rotavirus fell from 40% to 20% ⁷. Three doses protected 54% of children against severe rotavirus disease ⁷. That is lower than in rich countries. At the scale of India's birth cohort, it is still a large gain for child survival.

A new pooled analysis now answers the question that matters most: do these vaccines stop children dying? Across hospital studies in 16 countries, at least one dose cut deaths from rotavirus diarrhoea by 75.8% ¹². This article sets out what evidence from September 2011 to September 2026 shows, where the vaccines fall short, and what ministries and funders can do next.

The burden

Rotavirus is still the single largest cause of diarrhoea deaths in young children. The Global Burden of Disease (GBD) 2023 study, published in 2026 and funded by the Gates Foundation, estimates that rotavirus causes 40.2% (95% uncertainty interval 32.5–48.5) of diarrhoea deaths in children under 5 ¹. The next causes, Shigella (24.0%) and adenovirus (23.4%), are well behind ¹.

Across all ages, enteric infections killed an estimated 1.27 million people in 2023, down from 3.69 million in 1990 ¹. Diarrhoeal diseases caused 1.11 million of those deaths ¹. South Asia (599,000 deaths) and sub-Saharan Africa (501,000) carry most of the load ¹.

Gavi, the Vaccine Alliance, cites about 443,832 deaths from diarrhoeal disease each year in children under 5, and names rotavirus as the main cause of infant deaths from diarrhoea ².

The trend is down. The WHO position paper of 2021 records that rotavirus caused more than 500,000 child deaths and more than 2 million hospital admissions in 2000 ³. Between 2013 and 2017, the estimate was 122,000–215,000 child deaths each year, a fall of 59%–77% since 2000 ³. An analysis of GBD 2021 data estimated about 170,000 rotavirus deaths at all ages in 2021 ⁴.

Where the burden stays high. GBD 2023 found that 141 of 204 countries met the global target of fewer than 20 diarrhoea deaths per 100,000 children under 5 by 2025 ¹. In the countries that missed it, rotavirus was consistently the first or second cause of diarrhoea deaths in young children ¹.

Part of the fall is due to vaccines. A team at the London School of Hygiene & Tropical Medicine (LSHTM), with PATH and other partners, modelled 186 countries. They estimate that rotavirus vaccines prevented 139,000 under-5 deaths (95% UI 98,000–201,000) from 2006 to 2019 ⁵. In 2019 alone, vaccines prevented 15% of under-5 rotavirus deaths. Global use, at the coverage of other infant vaccines, could prevent 37% ⁵. The gap between 15% and 37% is the size of the task that remains.

How it works

The virus. Rotavirus infects the cells lining the small intestine. It damages them, so the gut cannot absorb fluid, and the child gets watery diarrhoea and vomiting ³. Severe cases lead to dehydration, shock and, without treatment, death ³. Before vaccines, almost every child was infected by age 3 to 5 ³. Treatment is fluid replacement (oral rehydration salts) and zinc ³.

The vaccine. All current rotavirus vaccines are live, weakened viruses given as drops by mouth ³. The weakened virus multiplies in the child's gut and trains the immune system, without causing disease ³. There is no injection.

The four WHO-prequalified products are ³⁶:

• Rotarix (GSK, Belgium): one human strain (G1P⁸); 2 doses.
• RotaTeq (Merck, USA): five strains; 3 doses. Merck left the Gavi market in 2020 ⁶.
• Rotavac (Bharat Biotech, India): one strain (G9P¹¹); 3 doses.
• Rotasiil (Serum Institute of India): five strains; 3 doses.

The schedule. WHO recommends the first dose as soon as possible after 6 weeks of age, so that the child is protected before first exposure ³. Doses should be at least 4 weeks apart ³. In practice, doses are given four weeks apart starting at six weeks of age, alongside other infant vaccines ³⁶.

Practical points for health workers, from the WHO position paper ³:

• Catch-up is allowed up to 24 months. If a child under 2 misses a dose, give it. Do not restart the series. WHO does not recommend vaccination after 24 months, because most severe disease happens before then.
• Give drops first. The sugar in the oral vaccine reduces pain, so give it before injections.
• Switching products is fine. Complete the series with the same product when possible. If not, use any licensed product. If any dose was from a 3-dose product, or the product is unknown, give 3 doses in total ³⁶.
• Local strain data should not drive product choice, because prequalified vaccines protect across strains ³.
• Malnourished and HIV-positive children can be vaccinated on the standard schedule. Premature babies follow their age from birth ³.
• Do not vaccinate a child with a past intussusception, severe allergy to a previous dose, or severe combined immunodeficiency ³.

WHO also stresses that vaccination is one part of diarrhoea control. Breastfeeding, handwashing, clean water, sanitation, oral rehydration and zinc remain essential ³.

What the evidence shows

2014–2017: the trials behind the lower-cost vaccines

India's 2014 phase 3 trial of the 116E vaccine, later known as Rotavac, randomised 6,799 infants. Efficacy against severe rotavirus gastroenteritis was 53.6% (95% CI 35.0–66.9) ²⁵. That established meaningful protection despite incomplete efficacy; the newer Indian surveillance tests performance after rollout. Bharat Biotech, the manufacturer, was among the trial's funders ²⁵.

In Niger, the 2017 placebo-controlled trial of the heat-stable vaccine later known as Rotasiil found 66.7% efficacy against severe rotavirus gastroenteritis (95% CI 49.9–77.9) in the per-protocol population of 3,508 infants. Deaths were 27 versus 22, and serious adverse events did not differ significantly ²⁶. The trial demonstrated disease protection, not a mortality reduction. These trials support using affordable oral vaccines while keeping protection against severe disease distinct from later evidence about deaths.

India's national programme (Rotavac)

The Christian Medical College (CMC) Vellore network, led by Gagandeep Kang's group, ran hospital surveillance across India from 2016 to 2020 ⁷. It enrolled 24,624 children with diarrhoea and included 8,372 in the effectiveness analysis ⁷. The design combined a test-negative comparison (vaccination rates in rotavirus-positive children against rotavirus-negative children) with a before-and-after comparison.

Three doses of Rotavac gave a vaccine effectiveness of 54% (95% CI 45–62) against severe rotavirus gastroenteritis ⁷. Against the G1P⁸ genotype, effectiveness was 81% ⁷. The rotavirus share of diarrhoea admissions fell from 40% to 20% ⁷.

Caveat: effectiveness varied by strain, and hospital surveillance does not measure deaths at home.

Afghanistan (Rotarix)

The Afghan Ministry of Public Health worked with LSHTM and the US CDC on a test-negative study of 1,172 cases and 2,173 controls ⁸. Two doses of Rotarix gave 45% (22–62) protection against hospitalisation ⁸. Rotavirus positivity among admitted children fell from 51% to 39%, a 39% adjusted reduction ⁸.

Caveat: this is observational data from a fragile state. It is also some of the only data from such a setting.

Mali

CVD-Mali and the CDC measured effectiveness in children under 2 from 2015 to 2018, using a test-negative design ⁹. It was 60%, though the confidence interval was very wide (3–84) ⁹. Mali has one of the highest under-5 mortality rates in the world, so any usable data from there count.

Caveat: the wide interval means the true effect could be small or large.

Ghana's switch to a cheaper vaccine

In 2020 Ghana switched from Rotarix to Rotavac. The Noguchi Memorial Institute for Medical Research and Yale fitted a transmission model to surveillance data from Navrongo ¹⁰. Overall effectiveness against moderate-to-severe cases was 21.7% (17.4–26.1) ¹⁰. After the switch, from 2020 to 2023, the mean relative overall effectiveness was 37.1% (27.5–46.5) ¹⁰.

This is the first real-world evidence on an Indian-made rotavirus vaccine in Africa after a product switch. Ghana did not lose ground.

Caveat: these are model-based estimates, not a trial.

Protection against death: the MNSSTER-V pooled analysis

This is the most important new result. A CDC-led team, with partners including the SAMRC Wits Vaccines and Infectious Diseases Analytics unit in Johannesburg, pooled child-level data from test-negative studies in 24 countries, enrolled from 2007 to 2023 ¹². The analysis included 27,252 children under 5 from 22 countries ¹².

Sixteen countries reported at least one in-hospital death from acute gastroenteritis, covering 21,522 children. In total there were 183 deaths from all-cause gastroenteritis, and 25 deaths in children who tested positive for rotavirus ¹².

In children aged at least 3 months who had received other routine vaccines, at least one dose of rotavirus vaccine gave an adjusted effectiveness of 75.8% (95% CI 28.4–91.8; n=13,630) against death from rotavirus gastroenteritis ¹².

Against death from gastroenteritis of any cause, effectiveness was 20.8% (−47.0 to 57.3; n=20,005), which is not statistically significant ¹². That fits the fact that many diarrhoea deaths have other causes.

Caveat: the estimate rests on 25 rotavirus deaths, all in hospital. The study reported no funding source ¹².

Safety in Africa

Intussusception, a rare bowel blockage, is the main known safety concern with rotavirus vaccines ³. A CDC and WHO Regional Office for Africa study of 318 infants in five African countries used a self-controlled case series ¹¹. It found no increased risk after RotaTeq. After dose 1, the relative incidence was 2.71 (0.47–8.03) in days 1 to 7 and 0.77 (0.0–2.69) in days 8 to 21, neither statistically significant ¹¹.

WHO states that all prequalified rotavirus vaccines have a good safety profile, and no serious adverse event other than intussusception has been linked to them ³.

Deaths averted at global scale

Two models give the scale. The independent LSHTM-led model estimates 139,000 under-5 deaths prevented from 2006 to 2019 ⁵.

An analysis by Amaris Consulting, funded by GSK, the maker of Rotarix, estimates that about 431 million children completed a Rotarix course between 2004 and 2023 ¹³. It estimates more than 220,000 rotavirus deaths avoided by 2023, with a forecast of 259,000 by December 2025 ¹³. Because the manufacturer funded it, treat this as a context figure, not an independent estimate.

Results at a glance

VE = vaccine effectiveness; RI = relative incidence; n.s. = not statistically significant.

LimitsWhat didn't work (or the limits)

The performance gap. Swiss TPH and Yale researchers pooled efficacy and effectiveness studies in a meta-regression ¹⁴. Predicted efficacy was 69.6–94.3% in the Americas, Europe and the Western Pacific. In the African, South-East Asian and Eastern Mediterranean regions, where most rotavirus deaths occur, it was only 18.6–85.3% ¹⁴.

The Cochrane review cited by WHO found the same pattern: about 90%–95% efficacy against severe disease in low-mortality countries, against about 44%–70% in high-mortality countries ³. In high-mortality countries in Africa and Asia, the four vaccines had similar efficacy at 1 year, 48% to 57% ³.

Malnutrition blunts protection. Aga Khan University in Karachi ran a test-negative study with 517 cases and 1,310 controls ¹⁵. Two doses of Rotarix gave 33% (12–49) effectiveness in infants. In better-nourished children (weight-for-age z-score of −2 or more) it was 45% (21–62). By age 1, effectiveness was 24% (−20 to 52), which is not statistically significant ¹⁵. The children most at risk of dying from diarrhoea are the malnourished ones, and they get the least protection.

An injectable vaccine failed. One idea was to bypass the gut with an injected vaccine. PATH tested a three-dose injectable subunit vaccine, TV P2-VP8, against two doses of oral Rotarix in a phase 3 trial in Ghana, Malawi and Zambia ¹⁶. The trial enrolled 4,055 infants and was stopped early at interim analysis ¹⁶.

Severe rotavirus gastroenteritis occurred in 4.51% of children on TV P2-VP8 and 2.57% on Rotarix ¹⁶. The relative efficacy of the new vaccine was −77.68% (95% CI −162.38 to −21.54), so it protected worse than the existing oral vaccine ¹⁶. It was safe and produced antibodies, but antibodies did not translate into protection ¹⁶. This is an important negative result. It shows that Rotarix, despite its limits, is a hard benchmark to beat.

Protection varies by strain. In India, effectiveness was 81% against G1P⁸ but 54% overall ⁷. Mixed genotypes pull the average down.

Thin data on the cheaper vaccines. The meta-regression included 56 Rotarix studies and 35 RotaTeq studies, but only 3 for Rotasiil and 2 for Rotavac ¹⁴. These are the cheaper products, so more independent evaluation is needed.

Supply shocks. UNICEF reports that supply disruptions have marked the rotavirus market since 2018 ⁶. A 2021 fire at a Serum Institute plant disrupted one presentation. Bharat's liquid presentations were delayed in 2022, so five countries took the frozen one instead. Reduced GSK supply in 2022 and 2023 forced seven countries to switch product ⁶. Each switch costs time and retraining.

Wide uncertainty. The Mali estimate (3–84) and the non-significant Pakistan estimate at age 1 show how hard precise measurement is in these settings ⁹¹⁵. Ghana's overall effectiveness of 21.7% also comes from a model, not a trial ¹⁰. The mortality estimate rests on only 25 rotavirus deaths ¹².

Cost and scale

Countries and coverage. WHO reports that 136 countries had introduced rotavirus vaccine by the end of 2025, with global coverage estimated at 61% ¹⁸. That is up from 123 countries and 55% coverage as of 2023 ⁶.

Gavi support. By the end of 2024, more than 406 million children had been immunised with rotavirus vaccine with Gavi support ². Coverage across 57 Gavi-supported countries reached 71% in 2024 ². Chad and Viet Nam introduced the vaccine in 2024. Somalia, South Sudan and Cambodia planned launches in 2025 ². Gavi pays for vaccine procurement plus a one-time introduction grant ².

UNICEF procured 329 million rotavirus vaccine courses from 2011 to 2023, rising from 900,000 courses in 2011 to 39 million in 2023, for 57 countries ⁶. In 2023 those 39 million courses (101.4 million doses) cost USD 127.3 million. That was about 44% of global volume but only 5% of global revenue ⁶.

Price per dose. UNICEF's awarded prices for Gavi-eligible countries include ¹⁷:

• Rotavac (Bharat Biotech), main presentation: US$0.85 per dose from 2019, falling to US$0.70 per dose for 2025–2028.
• Rotasiil (Serum Institute of India), 2-dose lyophilised vial: US$0.95 per dose. Serum Institute applies the same price to lower-middle-income countries outside Gavi.
• Rotarix (GSK), single-dose tube: €1.88 per dose, rising to €2.05; blow-fill-seal tubes €1.70–1.79 per dose.

India buys its own vaccine; Rotavac was priced at US$0.95 per dose for India from 2018 to 2021 ¹⁷.

The middle-income squeeze. Countries outside Gavi pay far more. Manufacturers offer up to US$18.00 per course to lower-middle-income countries and US$27.00 per course to upper-middle-income countries that self-finance through UNICEF ⁶. Bharat's frozen 5-dose Rotavac costs US$2.55 per course for Gavi-eligible countries and US$4.50 for lower-middle-income countries ⁶. Twelve eligible countries have not applied for Gavi support, citing programme readiness, cold chain capacity and concerns over long-term pricing ⁶.

Switching saves money. When GSK supply fell, Gavi partners found that countries could save 18–60% of their vaccine co-financing cost, and up to 70% of cold chain volume, by switching to other options ⁶. Ghana's experience suggests protection can hold after such a switch ¹⁰.

Manufacturers. Three suppliers now serve the Gavi market: Bharat Biotech and Serum Institute of India, both in India, and GSK in Belgium ⁶. In 2023, 71% of UNICEF demand was for GSK's Rotarix single-dose tube, largely because it needs only 2 doses and less cold chain space ⁶. Viet Nam (POLYVAC) and China (Lanzhou Institute) make vaccines licensed in their own countries but not prequalified by WHO ⁶.

What to watch

A vaccine given at birth. RV3-BB is an oral vaccine made by PT Bio Farma in Indonesia, which sponsors its phase 3 trial ⁶¹⁹. The Murdoch Children's Research Institute in Melbourne sponsored its phase 2 study in Malawi (NCT03483116) ²⁴. It is based on a neonatal rotavirus strain that replicates well in the newborn gut in the presence of breast milk, so the first dose can be given in the first days of life ²³. A phase 3 trial in Indonesia (NCT04185545) enrolled 1,400 neonates to test efficacy, safety and immunogenicity ¹⁹. It started in October 2020 and was completed in May 2023. No results were posted on the registry by September 2026 ¹⁹. Publication of these results is the key event to watch.

A preprint from Yale and Noguchi, not yet peer-reviewed, modelled RV3-BB in Malawi and Ghana ²⁰. It found RV3-BB would be preferred over Rotavac if priced below US$1.2 per dose in Malawi and US$2.5 in Ghana ²⁰. Price will decide whether a birth-dose vaccine is worth switching to.

Boosters and mixed schedules. A 2026 review by CMC Vellore, WHO AFRO and Emory University highlights alternative dosing schedules, booster doses in late infancy and mixed vaccine regimens as ways to improve performance ²¹. The Pakistan data on waning by age 1 point in the same direction ¹⁵.

New presentations. UNICEF expected GSK's blow-fill-seal Rotarix, Bharat's 5-dose liquid and Serum Institute's 1-dose liquid to return or arrive in late 2024 and early 2025 ⁶. Serum Institute's lyophilised Rotasiil presentations were due to be phased out in 2025, so countries using them must switch ⁶.

Other candidates. UNICEF lists candidates from Bio Farma (Indonesia), Lanzhou Biologicals and Wuhan Institute of Biological Products (China) in clinical development ⁶. UNICEF does not expect any new product to be prequalified in the near future ⁶.

Guidance. The 2021 WHO position paper remains current ³. The published highlights of the March 2026 meeting of SAGE, WHO's Strategic Advisory Group of Experts on Immunization, contain no new rotavirus recommendation. Rotavirus remains one of 14 diseases with WHO recommendations for all immunization programmes ²².

What it means

Half the protection seen in Europe is not a failure when the disease is common and often deadly. In India, Afghanistan, Mali and Ghana, rotavirus vaccines consistently reduced severe disease and hospital admissions ⁷⁸⁹¹⁰. Pooled data now show they cut rotavirus deaths by about three-quarters ¹². The safety data from Africa support continued use ¹¹.

For ministries of health:

• Introduce or keep going. Rotavirus remains the leading cause of diarrhoea deaths in the countries that missed the 2025 target ¹. Countries that have not introduced the vaccine are leaving the biggest single cause on the table.
• Close the coverage gap. Global coverage is 61% ¹⁸. Vaccinating on time, and catching up missed children up to 24 months, is the cheapest gain ³.
• Choose on price and logistics. WHO advises choosing a product on supply, price and programme fit, not strain data ³. Rotavac and Rotasiil cost under US$1 per dose for Gavi countries ¹⁷, and Ghana switched without losing ground ¹⁰.

For funders:

• Protect middle-income access. Prices up to US$27 per course keep some countries out ⁶.
• Fund independent studies of the cheaper vaccines. There are only 5 efficacy or effectiveness studies of Rotavac and Rotasiil combined in the global meta-regression ¹⁴.
• Back the next generation carefully. The injectable TV P2-VP8 failure shows that antibodies are not enough ¹⁶. The RV3-BB results are the next test ¹⁹.

For clinicians and health workers:

• Give the drops at 6 weeks, before injections, and complete the series ³.
• Vaccinate malnourished and HIV-positive children on the usual schedule ³.
• Keep oral rehydration and zinc at hand. The vaccine does not replace treatment ³.

The frontier for research has three parts:

• Nutrition. Pakistan's data suggest that better-nourished children respond better ¹⁵. Links between nutrition programmes and vaccine response deserve study.
• Cheaper products. India's Rotavac worked at national scale, and Ghana's switch held up ⁷¹⁰. More independent effectiveness studies of Rotavac and Rotasiil are needed ¹⁴.
• Waning. Protection in Pakistan weakened by age 1 ¹⁵. Schedules that extend protection through the second year are worth testing ²¹.

The honest summary is the title: rotavirus vaccines give about half the protection in the places where children die of diarrhoea, and that half is still worth a great deal.