Do measles vaccine campaigns work?

Before the 1980s when mass vaccinations began, measles killed nearly 3 million people - most of them children - every year. Where vaccines are rigorously used, the numbers of deaths are now very low. But, in some places, uptake has been more patchy and also the absence of actual data proving that vaccination campaigns save lives have made them harder to implement, particularly in resource-poor settings like India. Speaking with Chris Smith, Prabhat Jha, from the University of Toronto, has used data from a mortality survey in India, called the million death study, which he leads, to compare measles fatality rates in areas of the country with vaccination campaigns and those without…

Prabhat: Do measles campaigns save lives? This was the central question we sought to answer in India where, measles deaths remain quite common. And what we found is that the government of India's national measles campaign avoided something like 50,000 deaths over a three year period. So this study very clearly answers that question, that special efforts to eliminate measles deaths work where measles deaths remain common.

Chris - I must admit Prabhat, when I read your study I was pretty surprised to learn that we don't actually have data supporting doing these sorts of campaigns. It seems a little bit like it's assumed knowledge.

Prabhat - You're right, Chris, that public health has acted on measles without information on whether it actually made a difference. But we know from all sorts of experience, including our kids being immunized against measles in high income countries, that it does save lives. Just the recent controversy and cases that have been reported in the US, where measles immunization has tailed off, is a demonstration of that. But the big picture globally is of deaths from measles in places where it remains common, such as in India or parts of Africa. And importantly, in these settings, most deaths don't occur in the hospital; but rather they occur at home without medical attention. So the million deaths study obtains a random sample of all of the deaths. And in this case we studied in particular 27,000 child deaths from 2005 to 2013.

Chris - And what's the significance of that date window?

Prabhat - The significance is that in around 2009, the government of India launched a special campaign to try to reduce measles deaths, in the states of India where measles was more common. And what we were able to do is to see a reduction in measles deaths in the places where the campaign had started.

Chris - In essence then, if you're looking at a range of different causes of mortality, and you see a reduction in measles, but other things like diarrhea for example remain stable, you can be reasonably sure that the intervention - i.e. the vaccination campaign - has had something to do with that difference.

Prabhat - That's absolutely right Chris. What we did is look at whether the decline was greater in the special states, that campaigns states, and it was - it fell by about 27% measles deaths in those states, versus the non campaign states where it fell 11%.

Chris - So it obviously is achieving what it sets out to do. If you look in more detail, though, at who it's saving and who it's helping, is it across the board? Or are there any exceptions or unusual outcomes from it?

Prabhat - What was welcome news in this study is that the reduction in girls’ mortality from measles was greater than boys’. Now generally, mortality from measles in India is higher in girls than boys. And we also found that particularly in the districts where the campaign was most active was the biggest reduction. So it does suggest that the campaign - which doesn't rely on families bringing their children to a clinic, but rather goes out to the house, finds unvaccinated children or under-vaccinated children, and whacks them with a second dose of measles - was actually quite effective.

Chris - Why do you think this study is actually necessary though? Because could the WHO have not told you that, just look at the data for any country, and if you vaccinate against this disease you get lower mortality rates. It saves lives, this vaccine, therefore it's a good idea; and it doesn't matter which country you live in, you're going to benefit.

Prabhat - There's two reasons why we need direct evidence. The first is that the WHO or model-based estimates are probably overestimating the numbers of actual measles deaths. And the second is local believability. Governments and media, and particularly the public, can be skeptical of sometimes foreign labelled models. They think, “oh well, these are just do-gooders from outside.” And in India, like in the UK or in Canada or the US, there is a strong anti-vax movement. So getting good local data that are generated by the government, rigorously evaluated, and saying, “well, these are your own numbers that are showing that measles vaccine saved lives,” is important for the public discourse. It will change minds, we hope.

Chris - Have you sent your results to Mr. Modi, and said, “well done, have a pat on the back for a campaign well done, now make sure you keep funding it?”

Prabhat - My approach is to always rub the nose of politicians in the data and nothing else. We had co-authors from the Indian government, and they've expressed a strong interest in saying, yes, they would like to think about potentially a measles death elimination effort. The hope is that they will also implement more decentralized simple systems like this to count the dead and describe the causes, and then they can relate it to the amount of spending that is ongoing, and use that mortality surveillance as a way of monitoring the progress. It's good news, and we hope there will be more.