With more than 5 million patients a year entering intensive care units, researchers are investigating news ways to use vast amounts of ICU data to help providers and patients achieve better outcomes, according to The Wall Street Journal.
Previously, ICUs have worked to reduce obvious risks to patients with more basic techniques, such as checklists. Now, many hospitals are testing big data approaches to sift through years of medical records and information to find correlations no one previously knew existed, thus discovering more trouble spots and more potential solutions.
Here are three thoughts regarding big data's role in improving quality and outcomes, from industry professionals, as outlined by WSJ.
1. According to Kenneth Sands, MD, chief quality officer and senior vice president at Beth Israel Deaconess Medical Center in Boston, checklists work for predictable risks, but "it's the less-predictable, unexpected event that we are most concerned about." For that, Beth Israel is working with Massachusetts Institute of Technology and human factors experts on a project called Risky States to be able to gauge risk levels in a particular ICU at any given time.
2. Daniel Talmor, MD, interim chair of anesthesiology, critical care and pain medicine at Beth Israel said, "We are using data you wouldn't necessarily think of in predicting harm in the ICU." For example, a nurses' level of experience "is typically not factored into risk." Now, Beth Israel is analyzing numerous factors, like nurse experience, to identify situations with increased risk.
3. Johns Hopkins Medicine in Baltimore is another organization trying to improve its ICU's ability to compare current patient data with historical data for faster analysis and diagnosis. Peter Pronovost, MD, PhD, senior vice president for patient safety and quality at Johns Hopkins Medicine, says the lack of connectivity among medical devices in the ICU forces physicians and nurses to piece together information from individual devices, which can be painstaking and puts patients at risk.
"We need software that links the medical record with devices to predict which harms patients are at risk for, recommend what therapies you need to prevent harms, display whether they got the therapies and then monitor to see how the patients did," Dr. Pronovost told the WSJ.