The American way of death data
The CDC is using emerging IT tools to improve the way mortality data is reported, collected and used.
While Benjamin Franklin famously said the only sure things in life are death and taxes, the Centers for Disease Control and Prevention is working to make data serve both.
Information about mortality is valuable to governments in many ways, from tallying what caused it, updating tax and voter rolls and other basic functions of governmental housekeeping.
Death registration is also frustratingly difficult to gather accurately and efficiently nationwide because it can come from an almost dizzying array of sources, including funeral homes, physicians, medical examiners, coroners, and local and state health departments.
The CDC has been working to make gathering mortality data across the country easier, more secure and efficient under a program called the CDC Surveillance Strategy.
The initiative, announced in February 2014 is aimed at accelerating use of emerging IT tools and other efforts across public health organizations to help detect and monitor disease, injury and collect other crucial data. Specifically, it calls for modernizing and transforming the National Vital Statistics System (NVSS) into one capable of supporting near real-time public health surveillance, including mortality data. The strategy set a performance objective of having 80 percent of death reporting performed electronically within one day in at least 25 states by 2016.
Paula Braun, a data scientist with Charlottesville, Va.-based consulting firm Elder Research who has been embedded at the CDC since last December under its Entrepreneur in Residence program, said she sees the agency on the right track, citing an upward trend from 2011 when 10.6 percent of reporting was within 10 days, to 2014, when 30 percent reported in 10 days.
‘A modernized death reporting system’
Gathering accurate and timely information about the deceased -- including cause of death, surviving relatives and basic identification information -- is a scattered process, with data flowing from morticians, doctors and others to a wide array of state systems, according to Braun.
Braun is helping CDC devise better electronic death registration systems support so the agency can hone its analytics and predictive modeling capabilities to assist federal agencies that depend on mortality data, such as the Social Security Administration and Treasury Department.
An improved electronic death registration system would result in:
- Faster reporting of final annual mortality data.
- Real-time surveillance of deaths of public health importance.
- Quick matching with birth certificates, assuring that birth certificates of people who recently died are not issued to others for fraudulent purposes.
In the coming months, Braun wants to have some basic plans drawn up for a renovated national system.
“Within the next 90-120 days, we hope to have the conceptual roadmap and skeleton for the IT infrastructure for a modernized death reporting system,” she said.
Braun is looking to set up trials of several reporting systems around the country to test ideas. She has extended her initial year-long stint at CDC under the entrepreneur program to work on the issue beyond the end of the year.
Any national mortality reporting system would be dependent on local and state health departments, she said. Their electronic death registration systems are uneven, ranging from state-of-the-art technology to basic data processing capabilities, including some paper-based operations.
And there are a lot of them -- the 50 states, five U.S. territories, the District of Columbia, and an independent system for New York City.
Braun is working to develop a standard interface for the mortality data repository, with help from the National Association for Public Health Statistics and Information Systems. The national nonprofit organization represents state vital records and public health statistics offices, making it a logical point of contact for the program, as well as a thought-leader for health IT, Braun said.