How Toronto-based epidemiologist Prabhat Jha is counting the dead to save the living
As a medical intern at Winnipeg’s St. Boniface Hospital, Prabhat Jha earned the nickname Dr. Discharge for his heroic life-saving efforts. But when he started to see the names of some of his former patients in the obituaries months later, he realized there was a disconnect between his fast-acting techniques and keeping people healthy.
So the Indian-born medical student traded dreams of being a “hotshot clinical doctor” to study the patterns of disease and global health events as an epidemiologist — where there’s no heroism, just doing “small things for millions of people,” he said.
After attending Oxford University on a Rhodes Scholarship to study public health and landing a job at the World Bank, the now 47-year-old became the director of the Centre for Global Heath Research at University of Toronto-affiliated St. Michael’s Hospital.
With much of his previous work focused on HIV/AIDS, malaria and tobacco control, he launched the 12-year Million Death Study in India in 2002.
Since, unlike the west, many Indians die at home, Jha’s is the first study to systematically monitor the causes of death across rural and urban areas in hopes of changing policy and approaches to public health not only in India, but across the developing world.
His latest breakthrough as part of the study, published in The Lancet medical journal Wednesday, reveals the prevalence of tobacco-related and cervical cancers in India, killing mainly middle-aged citizens and highlights the need for a focus on prevention strategies.
Jha spoke with the Star from Singapore, where he received an American Cancer Society Luther L. Terry award for his ongoing work related to tobacco control. This is an edited version of the conversation.
Q: What is the goal of the Million Death Study?
A: It’s quite simple. It’s to count the dead. In India, like most parts of the world, most deaths occur in rural areas, without medical attention. In Canada, most deaths occur in hospital and usually a physician is able to certify why the person died. That simple statistic has been extraordinarily useful for understanding things like lung cancer. When lung cancer rates were going up in Canada and the U.S. and the U.K. in the 1930s, the statistics led to the research that identified smoking as a big cause.
So these statistics, one kind of assumes, are done in developing countries also, but it’s just not true. This is where the Million Death Study came in.
Q: How do you collect the data for the study?
A: It’s also a simple method. The Indian government and the census department splits India into a million areas and randomly select about 7,000 of those. So it’s a true snapshot of India. They send a team around every month just to ask who’s born and who’s died. And we’ve trained those 800 staff to go and collect information. That’s converted into electronic records. Each record goes to two physicians independently and the two physicians have to agree on the cause of death if they disagree they get each other’s forms anonymously. And if they continue to disagree it goes to a senior physician.
Q: What kind of new information are you able to draw from the data?
The World Health Organization thought there (were) only 15,000 deaths from malaria at all ages. Well, we found 200,000 just below age 70. And we found that malaria was not just a killer of kids, but also of adults and that malaria was common where malaria transmission is reported. And this has spurred a global rethink on what is happening with malaria numbers. That’s what this research does. It creates unexpected results and shakes up assumptions. If you don’t look, you’re just guessing as to what’s going on and the people are invisible to the system. And we’ve hopefully removed that obstacle.
Q: Why did you choose India?
A: It’s an extraordinarily diverse place. You get parts of India that literally look, in their statistics, in their poverty levels and in their health statistics, like the world did 500 years ago. And then you have really modern diseases, such as . . . big increases of smoking, big increases in body weight and diabetes and heart attack is the leading cause of death in urban areas. It’s an epidemiologists dream to work in areas where there’s so much diversity.
Q: What can we learn from results in India?
A: In the northeast of India . . . next to Burma and to China, cancer death rates are four times that of just the adjacent states. And we have no clue as to why that is. Some of that is tobacco related and that’s important because the study has also documented that a million Indians are killed from smoking a year. If we find what those factors are, that’s going to be relevant not just for India but for global cancer. It could be relevant to the cancers we see in Canada. So there (are) extraordinary benefits.
Q: And how have people in India responded to being part of the study?
A: Because we work with close partners and the registrar general, the communities are very much behind it. They’re told at the outset this isn’t going to do you any benefit. But it is going to create statistics that help your community and it will help you help the community down the stream. They react very favourably to that.
In some of the places where we did a sub-study of blood pressure, for example, we didn’t have just the houses we recruited — the neighbours showed up and said, ‘Oh, why aren’t you taking our blood pressure? We want to be part of the study.’ So we had 105 per cent participation rate, which you don’t see in any kind of Canadian research.
Q: Do friends or coworkers ever tell you that it seems like you’re in a fairly morbid line of work?
A: I also work on HIV and tobacco, so the joke is I’m professionally obsessed with sex and death.
It’s important and you can learn a lot from a dead person, as I’ve said. And we are. The statistics are not about the dead. They’re about how to help the living.
Cancer in India: Latest results from the Million Death Study
• Of 122,429 deaths studied across 1.1. million homes, 7,137 were found to be from cancer. Those numbers correspond with the 555,400 national cancer deaths recorded in India in 2010.
• The majority (71 per cent) of cancer deaths were in people aged 30-69.
• In the 30-69 age range, the most common cancers in men were oral, stomach and lung. For women it was cervical, stomach and breast.
• Tobacco-related cancers represented 42 per cent of male deaths and 18.3 per cent of female deaths.
• Twice as many tobacco-related deaths were from oral cancer compared to lung, suggesting prevalence for chewing tobacco over cigarettes.
• Cervical cancer was less common in Muslim women than in Hindu women, suggesting less transmission of the HPV virus known to cause cervical cancer. It’s likely attributed to high circumcision rates among Muslim men which protects against HPV.