AAMI News May 2018

Hospital Uses Big Data to Keep Even the Smallest Patients Safe

EHRElectronic health records are one of the main sources of pediatric data.

After the wave of digitization that has swept healthcare over the last two decades, hospitals are “swimming in data,” according to one expert, but not all of them are sure what to do with it.

“You hear people talking about how valuable data is, but then people don’t talk about how hard it is to process it, how hard it is to understand it,” said Eric Kirkendall, MD, associate chief medical information officer at Cincinnati Children’s Hospital Medical Center. “Data is hard. To be able to transform it into knowledge and insight takes a lot of work.”

For more than a decade, Cincinnati Children’s has been building its infrastructure and staff to be able to tackle this challenge, not just for the benefit of its own patients but also for those being treated across the country and even in other parts of the world.

Building ‘Learning Networks’

Cincinnati Children’s plays a central role in the data collection and analysis for 11 pediatric learning networks. These networks bring together 558 teams across 286 pediatric care organizations in 43 states, Washington, DC, and five countries, including Belgium, Canada, Qatar, and the United Kingdom, to leverage data for clinical care, improved outcomes, and research.

“For the most part, with a couple of exceptions, most of the data for these networks is coming through Cincinnati Children’s in some fashion to be analyzed and presented back to participating centers,” said Daniel Jeffers Jr., a project manager in Cincinnati Children’s Division of Biomedical Informatics. “We are really becoming the hub for this data, as well as for project management, quality improvement, and potential clinical trials for this kind of work.”

One of these learning networks is the Solutions for Patient Safety (SPS), which includes more than 120 children’s hospitals in North America that care for approximately half of all children hospitalized in the United States.

Using high-reliability concepts and quality improvement science methods, SPS is focused on preventing readmissions, serious safety events, and 10 hospital-acquired conditions (HACs). It does this through a process that starts with clinical research into a safety issue that could be mitigated through technology. Once a potential solution has been “pioneered” by a small number of participating hospitals, it is then successively rolled out as a technology “bundle” through the entire network.

“Our quality improvement work differs from more traditional research in that we generally work with a narrower set of data and focus on improving a handful of key outcome measures, whereas research is often more exploratory,” said David Purcell, a research associate at the James M. Anderson Center for Health Systems Excellence based at Cincinnati Children’s. “In quality improvement work, we align from the beginning on the key variables that we will collect and monitor over the next couple of years to measure improvement.”

Participating hospitals enter data on both outcomes and process reliability each month into a set of web forms. The system then produces a number of charts automatically, and the data also undergoes some custom analytics by the Cincinnati Children’s team.

“Because of our continual quality improvement approach, we are constantly monitoring the data,” Purcell said. “We’ll look to see how well hospitals are doing implementing the bundle in terms of reliability. We’re also looking to see where the rate for that particular HAC is.”

Between 2012 and February 2017, this approach saved 7,282 children from serious harm and led to estimated savings of more than $130 million in healthcare costs, according to Cincinnati Children’s statistics.

Preventing Unintended Consequences

Part of Cincinnati Children’s focus on using data to improve patient safety extends to ensuring that the technological initiatives they implement are, in fact, improving outcomes.

“When you put something in place to try to make just one thing safer, it may make a different process or situation a little less safe. So, there are always these trade-offs that we have to evaluate, and that’s where data helps us,” said Kirkendall. “What you don’t want is your solution to become another problem.”