Monday, August 25, 2014
Ind. Gov't. - "In Indiana, state government tries using big data project to reduce infant mortality"
That is the headline to this story today by Mohana Ravindranath of the Washington Post. Some quotes:
Over the past year, Indiana’s governor, Republican Mike Pence, has spoken frequently about a persistent public health issue: a higher-than-average infant mortality rate. * * *
Pence’s staff suspected that the solution to Indiana’s infant mortality problem lay hidden in the state’s records, stored in disparate data sets throughout various government agencies. Analyzing these data sets — pairing causes of death from the Department of Health with socioeconomic data from the Family and Social Services Administration, for instance — could help the state determine which preventative programs to fund, they surmised.
In July, the Indiana state government signed a contract with European software company SAP to use its data management system HANA, which stands for High-Performance Analytic Appliance, to crunch data about infant mortality. This is the first major public-health big data application in Indiana, according to state officials, and if successful, the state plans to use the approach to solve other problems, such as recidivism. Indiana is investing about $9.1 million in the overall big data project. * * *
The state selected SAP — whose main customers are large commercial businesses — out of several other software vendors, according to Indiana’s chief financial officer and management and budget director Chris Atkins.
About a month into the project, the state government is still establishing links between the mortality rate of infants and other factors such as the parents’ level of education. Atkins declined to share the pilot’s findings until after they have been subjected to more tests.
“Maybe it’s education, [but] maybe it’s something else we need to look into,” Atkins said, adding that once the state identifies factors tied strongly to mortality, it plans to invest more heavily in programs addressing those factors.
If a low maternal education level is found to be a strong predictor of infant mortality in Indiana, the state may invest more heavily in education programs or in outreach for expectant mothers. Other factors, such as environmental pollutants or crime, could also be linked to mortality.
“The government gives us the ability to shift resources — whether they be personnel, or dollars — into programs or areas that demonstrate better results, and away from programs that don’t demonstrate results,” Atkins said.
Atkins declined to share how much the state was paying SAP. He noted that the majority of fees were paid upfront, with a small annual maintenance fee. The state also has had to hire a handful of ad hoc technology consultants to set up the effort.
“The state of Indiana didn’t have a lot of data scientists on the payroll before we started this project, but by and large we’ve been able to do it within existing [staff],” he said.
Posted by Marcia Oddi on August 25, 2014 09:14 AM
Posted to Indiana Government