Cheap, data-driven tool identifies sickest hospital patients

An automated tool that uses electronic health records.

Cheap, data-driven tool identifies sickest hospital patients
Medical workers moving patient on gurney through hospital corridor

A team of Yale researchers developed and tested an data-driven uses electronic health records to identify patients most at risk of deteriorating while in the hospital.

Led by associate professor of medicine Robert Fogerty, the team used criteria originally established to rapidly detect patients with sepsis, one of the most expensive and potentially deadly medical conditions in the United States. The researchers created software that directs the electronic health record system to notify an attending physician, via pager, as soon as a patient meets the criteria. When key vitals such as heart rate, blood pressure, and temperature change for the worse, the attending receives an automated text message.

The tool — designed from the outset to be low-cost, easy to use, and highly sustainable — was used to monitor more than 15,000 patients in real time over one year and identified individuals at increased risk for admission to the ICU and for mortality. It is a cost-effective yet powerful strategy for spotting seriously ill patients who might otherwise get missed, the researchers noted.

“We made things easier for the providers and safer for the patients, and we did it on a really small budget,” said Fogerty.

The study was a collaborative effort of investigators at Yale School of Medicine, Yale New Haven Health, and Yale Center for Analytical Sciences. The study is published in the Journal of Patient Safety.