“I’m not talking to you about Ebola or Zika virus but about a virus that everyone in this room has had and everyone of your children and probably all children in the world get in their first few years of life,” said Roger Glass, director of the Fogarty Institute, as he began his Wolfson Lecture in London last month.

In 1973 85% of cases of diarrhoea were of unknown aetiology. It was that year that Ruth Bishop, an Australian virologist, and her colleagues described rotavirus, although somebody in the audience pointed out that a Birmingham virologist, Tom Flewett, had identified it at about the same time and given the virus its name because it is shaped like a wheel.

Glass began his work on rotavirus in 1977 when he was working at the International Centre for Diarrhoeal Disease Research, Bangladesh, and his wife Barbara Stoll, now dean of the McGovern School of Medicine at the University of Texas in Houston, analysed cases of diarrhoea admitted to Dhaka Hospital and found that 6-9% were cholera and almost 30% were caused by rotavirus.

We now know, said Glass, that rotavirus is the commonest cause globally of severe diarrhoea, all children get it, and improvements in sanitation don’t prevent it (as they do cholera). It is a democratic virus—“or as we call it in the US,” joked Glass, “an equal opportunity virus”—infecting everybody.

In 1970 diarrhoea caused 4.8 million deaths but now it’s about 700 000, almost all of them in children under 5. Rotavirus was estimated to have caused around 873 000 deaths in 1985, but now it’s between 200 000 and 250 000. Globally, there are about 4.6 million hospital admissions or emergency room visits a year for severe diarrhoea.

Working out the burden of the disease in the US was not easy because rotavirus is usually not diagnosed in the laboratory, has no ICD code, has the same symptoms as other causes of diarrhoea, and a diagnosis is not needed for treatment. But analysis of hospital discharge data showed high peaks in children aged 7 months to 3 years in January—and the peaks are caused by rotavirus. It’s now clear that all children get rotavirus, about one in seven sees a doctor, one in 80 is admitted to hospital or attends an emergency room, and one in a million dies, which means about 20-40 deaths a year in the US.

The rotavirus is a double stranded RNA virus made up of 11 segments that come in every combination. It was soon discovered, however, that producing a vaccine against just four or five of the segments would prevent most infections. The young of cows, monkeys, pigs, and other animals also are infected with rotavirus, and non-human viruses are a common source of vaccines. The first vaccine was produced in February 1998 from a monkey strain of the virus and was used in the US.

But the deaths occurred in low and middle income countries, and Glass wanted to do trials of the vaccine in those countries. Bill Gates—who wondered how he had never heard of a virus that caused 600 000 deaths—agreed to fund the project to allow global use of the new vaccines. Unfortunately, it then became clear from the US that about one in 10 000 children who got the vaccine developed intussusception. The vaccine was pulled from the market in the US. Glass made an attempt to argue that the vaccine was still worth testing in India where one in 200 children died from rotavirus, but, although seeing the numerical logic, the Indian authorities thought that it would not be politically possible to test a vaccine banned in the US.

Despite having no vaccine to test, Glass managed to convince Gates to continue funding on rotavirus. The failure of the first vaccine left the market open for new manufacturers, and two new vaccines appeared in 2006. Safety trials were acceptable, and the vaccines were used in the US. There was widespread uptake, and the annual spike in hospital admissions in children disappeared. There was a 90% reduction in hospital admissions or emergency room attendances, saving about 50 000 hospital admissions a year. Younger children were vaccinated, but there was also a reduction in infections in older children—the effect of herd immunity. Intussusception still occurred, but with only about one to five cases for every 100 000 children vaccinated. (There was discussion at the meeting about why intussusception should be less common with the newer vaccines, and it could be something to do with younger children being vaccinated: intussusception increases markedly at 3 months of age.)

A trial of the vaccine was also conducted in Mexico and showed a 35% reduction in deaths. The vaccine was introduced in the UK in 2012 and led to an 84% reduction in hospital admissions and attendances at casualty. Yet a trial in Bangladesh showed efficacy to be only about 42%. There is now a consistent picture of higher efficacy in high income countries and lower efficacy in low income countries; in other words, said Glass, the vaccine “works less well where it’s needed most.” Nobody knows exactly why, but maternal antibodies from breast feeding and a different microbiome are suspected causes.

Nevertheless, the World Health Organization has recommended vaccination for all children, and the vaccine is being used in routine programmes in 82 countries today thanks to funding from GAVI to subsidise vaccine costs in the poorest countries. A new vaccine made in India was licensed in 2015 by Prime Minister Modi, and India plans to vaccinate half of the infants by the end of 2017.

The big challenges are to increase the efficacy of the vaccine and to bring down the cost. The vaccine costs $120 in the US and $15 in PAHO countries—and GAVI has reduced the cost to $5. But that’s still too much for many low income countries. Roger Glass continues with what has been a “lifelong passion.”

Richard Smith was the editor of The BMJ until 2004.

Competing interest: RS spoke at the summer school and had his expenses paid.