Data-driven autopsies aim to revamp developing world health care

Register now

(Bloomberg) -- In Myanmar, most deaths occur at home and only a quarter are certified by doctors. That’s worrisome, since not knowing why people die can undermine efforts to save lives.

Now, innovative methods for counting the dead are being rolled out in the Southeast Asian nation, revealing previously under-appreciated patterns of disease that may change health policies and strategies. Known as verbal autopsies, the data-gathering relies on midwives quizzing family members. The answers to a questionnaire are then entered into a computer tablet that uses a software algorithm to come up with a probable cause of death.

Results from the first 5,000 deaths collected from more than a dozen townships or districts are already challenging perceptions that pneumonia and diarrhea are among the biggest causes of death, according to Alan Lopez, the Rowden-White Chair of Global Health and Burden of Disease Measurement at the University of Melbourne, who is leading the research. Stroke and heart disease are instead coming up among the top killers.

“It’s a kind of light bulb moment,” said Lopez, who has been collating and analyzing mortality data for decades and is a co-author of the Global Burden of Disease Study, a research program funded by the Bill & Melinda Gates Foundation. “In Myanmar, they’re really captivated by this and are committed to rolling it out.”

Data For Health

Similar verbal autopsy and data-collection programs are being undertaken in Sri Lanka, Bangladesh and 16 other low- and middle-income countries under a four-year, $100 million initiative funded by Bloomberg Philanthropies in partnership with the Australian Department of Foreign Affairs and Trade. Bloomberg Philanthropies was set up by former New York Mayor Michael Bloomberg, founder and majority owner of Bloomberg News parent Bloomberg LP.

About 35 million deaths go unrecorded globally, leaving dozens of nations unable to reliably track what kills their citizens. The Bloomberg Data for Health Initiative aims to help more than 1 billion people by providing the tools, systems and training to enable countries to record more accurate birth and death data.

In the case of Myanmar, the 20-minute verbal autopsy process needs to be expanded to collect nationally representative samples that will generate reliable cause-of-death data. From that, evidence-based health policy decisions and prevention strategies for diseases such as type-2 diabetes can be formulated, Lopez said.

‘Surprise to Governments'

The metabolic condition is already turning up in 5-to-8 percent of deaths in poor populations, Lopez said, “and that’s a surprise to governments.”

Drugmakers, artificial intelligence companies, health insurers and pharmacy chains are also investing in data collection and analysis -- a market that research firm Research N Reports predicts will be worth more than $34.27 billion by 2022.

CVS Health Corp. agreed this month to buy Aetna Inc. for about $67.5 billion. That will create a health-care giant that will be better equipped to track health information and drive down drug prices, said Thomas Laur, president of SAP SE’s Connected Health group in Boston.

“Some of the large transactions we have seen over the last couple of days, weeks and month are really focused on creating large health-care delivery platforms by bringing all this data together and trying to drive inefficiencies out of those systems,” Laur said in an interview with Bloomberg TV last week.

Routinely collected data can boost the efficiency of a health system by identifying strategies for producing “good outcomes at an acceptable cost,” said Luke Slawomirski, a health economist and policy analyst with the Organisation for Economic Co-operation and Development in Paris. “However, no one data set can tell us everything.”

Privacy Risks

Consequently, there’s a need to link different sources of data, including from hospitals, clinics and pharmacies, to yield new insights and knowledge, Slawomirski said. “Of course, linkage raises patient privacy risks, so we need sound data governance frameworks that manage the risks while maximizing the benefits of secondary use of these data.”

One of the largest health threats that data from around the world are showing is obesity and its consequences, like diabetes -- a disease that has prompted lawmakers in countries from the U.K. to South Africa to pursue sugar taxes.

“We will see the drive for timely and very local information continue,” said Christopher Murray, professor of global health at the University of Washington and director of the Institute for Health Metrics and Evaluation in Seattle. “I think we will also see this approach to highly standardized measurement spreading into allied areas of health -- things that are big determinants of health, like education.”

Quantifying Disease

Murray co-founded the Global Burden of Disease Study with Lopez in the 1990s, when the research involved fewer than a dozen people. There are now almost 3,000 collaborators on the study, which aims to quantify the comparative magnitude of health loss due to diseases, injuries, and risk factors by age, sex, and geography over time.

Murray’s team is undertaking more disease modeling work to anticipate alternative health scenarios, and has received funding from the U.K. government, Wellcome Trust, and the Gates Foundation to analyze trends related to antimicrobial resistance, he said.

“Data is at the core of every decision we make,” said Trevor Mundel, president of global health at the Seattle-based Gates Foundation, the planet’s largest private supporter of global health. “Good data and smart analytics help us extract maximum value from our investments.”

--With assistance from Andy Clarke

Bloomberg News