Alert system uses predictive analytics to combat sepsis

Jersey City Medical Center monitors patient data, sends alerts to clinicians to head off issues.


Tapping into electronic health record data and predictive analytics, hospitals are enlisting a new technology solution in the fight against sepsis, a leading cause of death and hospitalization in the United States.

According to the Centers for Disease Control and Prevention, there are more than 1 million cases of sepsis each year, killing more than 258,000 Americans annually.

However, Jersey City (N.J.) Medical Center, a 325-bed urban tertiary care facility, is implementing a sepsis alert system that constantly scans patients’ EHRs and leverages predictive analytics to determine which patients are septic or at risk for septic shock and alerts appropriate care team members.

Called Sentinel, the system from Newark-based collaboration and communications vendor Uniphy Health tracks more than 70 clinical features in real-time, including body temperature, heart rate, respiratory rate, white blood cell count, medication lists, blood culture reports, diagnostic imaging, as well as clinician progress notes.

Vital signs combined with serum white blood cell count and other information can accurately predict sepsis, an immune system response to infection that can damage organs and cause permanent physical and mental disabilities and even death.

Also See: CDC Funds Zika surveillance systems in 40 states, territories

Uniphy Health’s Sentinel is a commercially available version of a sepsis and failure-to-rescue alerting system called the Military Acuity Model, which was originally developed in conjunction with the U.S. Air Force. The Military Acuity Model, the precursor to Sentinel, is credited with reducing by 87 percent deaths resulting from infection by identifying patients at risk for sepsis—as well as other diseases—and providing automated high-value task prompts or notifications only to care team members in a position to intervene.

“We are deploying Uniphy Sentinel to provide our care team physicians and nurses with secure, real-time alerts and put them in the best position to act on high-value messages that, if missed or delayed, pose a high risk to our patients’ lives,” said Kenneth Garay, MD, chief medical officer for Jersey City Medical Center, a member of RWJBarnabas Health.

According to Stuart Hochron, MD, co-founder and chief medical officer of Uniphy Health, many hospitals do not have a dedicated sepsis alerting system, relying instead on alerts via ad hoc systems combining automated and manual processes, which he contends often respond too slowly, may not recognize the signs of sepsis soon enough, or send false alerts to hospital staff.

However, Hochron contends that Sentinel addresses the problem of alert fatigue by providing 75 percent fewer alerts than a typical EHR system, delivering only those related to high-value tasks and to care team members best positioned to respond to a high-value issue. “Those tasks are sent to individuals who are not saturated with tasks,” he says.

The Military Acuity Model, on which Sentinel is based, was successfully tested at Johns Hopkins, where tasks performed by physicians and nurses were evaluated and then identified jobs that could be safely assigned to support staff, according to Hochron.

“Coordinators, hospitalists, and nurse practitioners can in effect become the de facto sepsis monitoring response team with this system,” he adds.

More for you

Loading data for hdm_tax_topic #better-outcomes...