This year more than 29 million people will die without a known clear cause. While this may sound like a plot for an Avengers movie, it's actually a real-life, real-world problem that billionaire philanthropist Mike Bloomberg wants fixed.
As Bloomberg explained last week at the New York City headquarters for Bloomberg, LP, "half of all deaths worldwide have no associated death certificates that indicate the cause of death." He described this as "a huge problem that results in a lot of preventable deaths and diseases because governments just don’t have the information that they need to target the causes." In other words, as they say in many settings ranging from the financial world to the set of the Real Housewives, you can't fix what you don't understand.
Bloomberg also mentioned a similar problem with births. There's no official documentation of nearly 40% of the 128 million babies born each year worldwide. While the reason for births is usually pretty clear (if it isn't for you, then you may need to re-take sex ed), this means that each year over 40 million infants are entering the world virtually invisible. The answer to the question "how does it feel to be a complete unknown" is not simply "like a rolling stone." Being an "undocumented birth" means that you may not get the appropriate health care and other support necessary to survive and thrive.
This is not a Big Data or even a Little Data problem. It's a No Data problem. No data on the most basic of information that's central to decision making for so many different things. Birth and death data are known as vital statistics for a reason. Without accurate tallies for the beginnings and endings of life, it is very difficult to get a full picture of what is happening in between these events. For example, you may not know what diseases and health conditions are the biggest killers and thus which to target for prevention and control. And you may not know where births are occurring and where best to allocate health care services and economic resources. This affects decision making not only for governments and health care entities but also a wide variety of businesses and industries. There's another word for making decisions without data. It's called "guessing" or "rolling the dice" or "wearing a blindfold" or in some cases "repeatedly smacking your head with a frying pan."
It's not surprising that this problem would resonate with Bloomberg. In 1981, he founded Innovative Market Systems (IMS) the precursor to Bloomberg, LP, to bring more data to financial decision-making. Nowadays, compared to the 1970's and early 1980's, there's a buffet of financial data available to investors. But if investors have a buffet, many parts of public health by comparison have cups of potato chip fragments.
Therefore, in March 2015, Bloomberg Philanthropies teamed with the Government of Australia to launch the four-year Data for Health Initiative, which has aimed to work with 20 low-and-middle-income countries (that have more than 1 billion people) to improve their public health data systems. The initiative has consisted of three components: the Civil Registration and Vital Statistics (CRVS) Program, which has aimed to improve the recording of births, deaths and causes of death, the Data Impact Program, which has aimed to expand the use of data for public health planning and policy making, and a program to develop new tools for noncommunicable disease (NCD) risk factor surveillance.
Kelly Henning, MD, who has led the Bloomberg Philanthropies Public Health program since it began in 2007, said that "this initiative made a lot of sense to Mike Bloomberg. Governments have to have their own information to make important decision." Henning also previously worked in the Division of Vital Statistics in the New York City Department of Health and Mental Hygiene when Bloomberg was the Mayor of New York City.
A major challenge at the beginning of the CRVS Program was that many countries just did not have the systems in place to collect vital statistics. As Adam Karpati, MD, MPH, Senior Vice President for Public Health Programs for Vital Strategies, a major implementing partner for the Data for Health Initiative, explained, "a lot of public health initiatives had focused on particular diseases or risk factors. By contrast, the CRVS focused on more fundamental systems building and capacity building." This basically means that the program had to help countries build the systems and expertise necessary to collect, manage, and analyze vital statistics data.
According to Philip Setel, PhD, Vice President and Director of the CRVS Improvement Program at Vital Strategies, "three reasons made improving vital statistics a particularly challenging problem." The first in his words was that "coverage and completeness rates were so low, and systems were so fragmented." In other words, because a number of the countries did not have adequate systems (e.g., personnel, procedures, processes, and machinery) in place, large swathes of their populations were overlooked.
A second reason has been what Setel described as a "fascination with statistical estimates and survey-derived estimates. Surveys could give quick approximate answers but were lousy in really determining causes of death." You don't need to take a survey of experts to realize that surveys have numerous limitations, simply revealing the opinions of a sample of people (if those sampled are even willing to reveal their true opinions) and not giving the full picture of a population. In survey results can even be misleading. (See the Ford Edsel, New Coke, and election polls.) Using surveys to determine what is really happening in the overall population is like asking a couple players on the Cleveland Cavaliers, "did you score a basket, did you think you won, do you think you are better than them" rather than keeping the actual scores and records during the NBA Playoffs. Moreover, as Setel said, "surveys to determine causes of death are prohibitively expensive and don’t factor into the governmental machinery."
The third reason was that building the necessary data systems was not previously the focus of many governments and the international community. Alan Lopez, PhD, currently a Professor at the University of Melbourne who previously had a 22-year career at the World Health Organization (WHO), has long been trying to expand the focus of health programs from just specific diseases to broader statistics such as births and deaths. As he explained, "the historical disease-focus of health has been annoying. Disease-specific programs all want info on their specific diseases so this has inhibited progress." Incidentally, his efforts pushing the need for better death data earned him the nickname "Dr. Death," which would rank pretty high on the list of intriguing public health nicknames (and possible names of heavy metal bands).
But as Karpati related, people not paying attention to this area also meant that it was fairly open space without the politics, obstacles, and navigation problems involved in entering a crowded, competitive arena. Once the possibilities were discussed with countries, "this was of great interest. Countries wanted this," according to Karpati. Setel added that "having such data could provide numerous societal and individual benefits. For example, a death certificate could help with insurance. A widow who is married into a patriarchal system could then accrue the benefits that she deserves."
Another challenge has been overcoming the "friends, Romans, lend me your data, so-that-I-can-either-sell-it-or-write-a-bunch-of-papers-which-will-help-me-get-promoted" concerns. Setel emphasized that "low and middle countries are not the laboratories and workbenches of high income countries. Low and middle income countries had to be confident that they would retain ownership of the data and take the reins of the process and that higher income countries were there just to assist with training and capacity building and not control things."
Additionally and more specifically, determining the causes of death is not a simple matter. Bloomberg Philanthropies, the Government of Australia, and Vital Strategies have been working alongside many government officials and partners such as the U.S. Centers for Disease Control and Prevention, the Johns Hopkins Bloomberg School of Public Health, the University of Melbourne, and the WHO. As Jennifer Ellis, PhD, who directs the Bloomberg Data for Health Initiative, related, "the Initiative has had to work across many different health systems in 20 different countries. Plus, many deaths occur outside the hospital so there has been a ton of investment into the verbal autopsy and its methodology." A verbal autopsy does not mean figuring out why certain phrases like "cruisin’ for a bruisin’" or "cool beans" are no longer used. Rather, it's the range of techniques (e.g., interviews) used to determine the cause of death when official records are not available.
Despite these daunting challenges, the Data for Health Initiative has been making substantial progress. Bloomberg indicated that the initiative has "so far collected or improved data on over 1 million deaths." Of the 20 countries, 17 countries now have international standard death certificates, 12 for the first time collect data on rural deaths, and 10 are developing new NCD mobile phone survey approaches. The initiative has also resulted in data behind 36,000 rural deaths, 61 manuals and tools on how to collect birth and death data, 29,022 health professionals trained in birth and death data, 46 data-driven policy briefs, 1064 people trained in the analysis of health data for policy, and nearly 200 people trained in NCD mobile phone surveys.
Here are some examples of country-specific progress. By the end of 2016, Zambia began coding death data for the first time, Rwanda becomes first Data for Health country to begin collecting data on rural deaths, Brazil held its first-ever training for journalists on using health data, and Ghana launched its first-ever government-issued scientific health bulletin. By the end of 2017, Brazil scaled their existing rural death data collection efforts to a national level, 9 additional countries began collecting data on rural deaths for the first time, Peru began requiring every physician to be trained in certifying cause of death, and the first-ever NCD mobile phone survey was launched in Zambia. So far in 2018, Peru has committed 7 million in US dollars to ensuring its electronic death certificate is reaching rural populations, and Bangladesh demonstrated a 40-fold increase in the registration of rural deaths
That's progress in just over three years, after years of comparatively little progress. Setel explained that it is "hard to overstate the courage that Mike Bloomberg had to tackle this intractable problem. This initiative has shown in 3 to 4 years that improvements can be made even to places that are ossified."
Think of this as thousands and thousands (and eventually millions and millions) of people appearing and finally being heard after remaining relatively invisible for so long. Helping the invisible to appear benefits people who are already visible as well. With millions of more voices available, our understanding of health risks and diseases could change dramatically in the coming years. Getting better data on births and deaths could be the death of outdated decision making and the birth of many new ways to make decisions and improve health around the world.
