Machine learning shows promise in optimizing ICU blood tests

A computational approach has the potential to help clinicians in intensive care units make better decisions about ordering common blood tests.

A computational approach has the potential to help clinicians in intensive care units make better decisions about ordering common blood tests.

Results of their study, presented earlier this month at the 2019 Pacific Symposium on Biocomputing, showed that using a machine learning algorithm developed by Princeton University researchers could have reduced the number of lab orders for white blood cell tests by as much as 44 percent.

In addition, researchers demonstrated that their approach would have helped inform clinicians to intervene sometimes hours sooner when a patient’s condition began to deteriorate.

“With the lab test ordering policy that this method developed, we were able to order labs to determine that the patient’s health had degraded enough to need treatment, on average, four hours before the clinician actually initiated treatment based on clinician ordered labs,” says Barbara Engelhardt, senior author of the study and associate professor of computer science at Princeton.

In their study, researchers leveraged the MIMIC III database—which includes detailed medical records of 58,000 critical care admissions at Boston’s Beth Israel Deaconess Medical Center—and selected a subset of 6,060 records of adults who were admitted to the ICU between 2001 and 2012.

“Our experiments show that we can estimate a policy that reduces the frequency of lab tests and optimizes timing to minimize information redundancy,” states the study. “We also find that the estimated policies typically suggest ordering lab tests well ahead of critical onsets—such as mechanical ventilation or dialysis—that depend on the lab results.”

“There is a scarcity of evidence-based guidelines in critical care regarding the appropriate frequency of laboratory measurements,” says Shamim Nemati, an assistant professor of biomedical informatics at Emory University, who was not involved in the study.

“Data-driven approaches such as the one proposed by (electrical engineering graduate student Li-Fang) Cheng and co-authors (Engelhardt and computer science graduate student Niranjani Prasad), when combined with a deeper insight into clinical workflow, have the potential to reduce charting burden and cost of excessive testing, and improve situational awareness and outcomes,” adds Nemati.

Also See: Boston Medical Center trims unneeded tests with EHR-based prompts

“Our results suggest that there is considerable room for improvement on current ordering practices, and the framework introduced here can help recommend best practices and be used to evaluate deviations from these across care providers, driving us towards more efficient healthcare,” conclude the authors.

“Furthermore, the low risk of these types of interventions in patient healthcare reduces the barrier of testing and deploying clinician-in-the-loop machine learning-assisted patient care in ICU settings,” they add.

Towards that end, the Princeton researchers are working with data scientists at Penn Medicine to introduce an optimal policy in the clinic for ordering lab tests.

“This is one of the first times we’ll be able to take this machine learning approach and actually put it in the ICU, or in an inpatient hospital setting, and advise caregivers in a way that patients aren’t going to be at risk,” says Engelhardt. “That’s really something novel.”

“Having access to machine learning, artificial intelligence and statistical modeling with large amounts of data” will help clinicians “make better decisions, and ultimately improve patient outcomes,” adds Corey Chivers, senior data scientist at Penn.

More for you

Loading data for hdm_tax_topic #better-outcomes...