Skip navigation

Million Death Study documents how people are dying in the developing world

The Canadian-led study has upended many assumptions about the causes of death, which are often poorly tracked in developing countries.


Four years after Prabhat Jha and his parents moved from Bihar, India to Winnipeg in the early 1970s, his grandfather died. Like many of India’s nine million deaths a year, Dr. Jha’s grandfather died at home, without medical documentation. Lacking the money to visit regularly, and with conflicting reports from family members back home, his mother returned to Bihar to speak to family members about her father’s death sometime later. In 1989, when Dr. Jha was in medical school, he did the same thing.

“I asked [my grandmother] what happened on that day,” said Dr. Jha, “and she could still recount everything in very vivid detail. I was able to make something of a diagnosis that he’d had a stroke, a left-sided one, and it might have been hemorrhagic, a serious bleed that killed him quickly.”

That experience of determining the cause of a death through family memories stayed with Dr. Jha during his early career: as a medical intern at St. Boniface Hospital in Winnipeg, a Rhodes Scholar of epidemiology at the University of Oxford, and a research team leader at the World Bank. While studying smoking-related illnesses and deaths in India in the mid-’90s, he had the idea to apply the concept of “verbal autopsies,” as they’ve since come to be called, to tracking causes of death using census visits throughout India, where this kind of information often is not recorded in rural areas.

Then in 2002, as an associate professor at University of Toronto’s Dalla Lana School of Public Health, he launched the Million Death Study, or MDS, one of the largest studies of mortality ever conducted and one that would occupy at least the next 14 years of his life. Now wrapping up the data-collection of its first phase, Dr. Jha’s work into the causes of death in India has upended preconceived notions about public health issues in that country.

“We do know broadly, at the global levels, what the big killers are,” said Dr. Jha, now a full professor who heads the Centre for Global Health Research sponsored by St. Michael’s Hospital and U of T. “But when you get to the country level, that’s where you have fundamental problems in really understanding how people die, and therefore being able to act upon it.”

To illustrate, he pointed to two examples of huge information gaps about deaths worldwide. First, the overwhelming majority of the nearly 60 million deaths that occur each year happen in developing nations. In these countries, hospital deaths are often the exception and many deaths go unrecorded, misidentified, or reported with limited information. (In India, for example, nearly 70 percent of deaths occur without medical care.) The second example concerns child deaths: as few as three percent of child deaths worldwide are documented with a death certificate. “To achieve the big gains we want to in global health,” said Dr. Jha, “it’s indispensable to have information on how people die.”

To begin, the MDS set out to record and analyze one million deaths throughout India that occurred between 1997 and 2014. The gargantuan undertaking fortunately had a framework to build upon: Twice a year, surveyors visit preselected homes across the country to gather data on various census topics, including births and deaths. Dr. Jha worked with the census commissioner to develop a standard verbal autopsy that surveyors could conduct on households that had recently experienced a premature death (that is, before the age of 70). Families are asked 12 symptom-related questions and make supporting notes about details of the circumstances of the event. These are electronically documented, sent to a team of doctors to determine a medical cause and then codified.

While the work involves a lot of knocking on doors and talking to people, it also is simple and relatively cost-effective, a boon to government departments with limited budgets. “We’ve spent a lot of time thinking about how you can make every step in this process cheaper, better, faster,” said Dr. Jha. “We’ve been designing computer systems that any country can use and organized the surveys very simply.”

While all one million of the recorded deaths won’t be codified and analyzed for another few years, the data has already produced dozens of journal publications, some of which have turned global health assumptions on their head:

  • One finding, published in the Lancet in 2010, found that death rates related to malaria were much higher than previously thought – close to 200,000 a year, compared with the World Health’s Organization’s estimate of 15,000 at the time – largely because WHO’s numbers relied on deaths in hospitals, where malaria can be treated more effectively.
  • A finding that smoking-related illnesses were killing up to one million Indians per year prompted government approval of more widespread use of warning labels on tobacco products.
  • From 2006 to 2011, Dr. Jha’s team published findings that sex-selective abortion was on a steady increase, starting a national conversation about the issue.
  • More recent findings in the last couple of years have focused on such topics as neonatal, maternal and traffic-related deaths.

Perhaps the biggest success of the study to date has been the compelling case it has made to government departments about the value in counting the dead. The verbal autopsy will continue to be used as a survey tool in India’s census when the process begins again in 2015.

The Toronto-based Centre for Global Health Research is now working with Ghana, Ethiopia, Nigeria, Bangladesh, South Africa and the Philippines to adopt similar surveys. Having this data, said Dr. Jha, empowers countries to determine their own public-health priorities, which may differ from those of foreign donors.

“It’s different than you and I coming in and saying, ‘Well, your biggest problem is HIV, and we’re here with money for it.’ We’d like a health minister to be able to say, ‘That’s good, but I’ve got twice as many people dying from chronic lung disease than AIDS – what are you doing about that?’”

Post a comment
University Affairs moderates all comments according to the following guidelines. If approved, comments generally appear within one business day. We may republish particularly insightful remarks in our print edition or elsewhere.

Your email address will not be published. Required fields are marked *

Click to fill out a quick survey