Top 5 videos from HIMSS17

We have compiled the top 5 most viewed videos from HIMSS17 TV below.


Top 5 videos from HIMSS17

NorthShore University Health System leverages predictive modeling through EHR data

Directors will discuss at HIMSS16 how they've been able to automate clinical workflows as a result.
By Jack McCarthy
10:07 AM

NorthShore University Health System is easing and, in certain instances, automating clinical workflows within electronic records to pinpoint high-risk patients and gaps in care using predictive modeling.

“We’re taking that analytics information and integrating it into the clinical workflow,” said Chad Konchak, NorthShore University Health System’s director of clinical analytics. “And putting it in the hands of clinical administrators who need to make clinical decisions.”

Konchak, along with NorthShore vice president of clinical analytics Ari Robicsek, MD, leads a team to improve workflows in a number of systems at NorthShore to advance care efficiencies at affiliated hospitals.

[Also: Doing predictive analytics? You might be doing it wrong]

Robicsek and Konchak will discuss their work at HIMSS16 in Las Vegas, which starts in late February with the presentation “Actionable Analytics: From Predictive Modeling to Workflows.”

Konchak and Robicsek will outline how analytics projects are transforming clinical data into predictive models and analytical tools that have been integrated back into the EHR and injected into clinical workflows for both inpatient and ambulatory settings. Their projects include predictive modeling to target high-risk patients and find harmful gaps in care.

They have also developed programs to integrate data into clinician workflows through EHRs, thereby enabling knowledge derived from large datasets and advanced modeling to be used at the point of care.

See all of our HIMSS16 previews

The NorthShore team’s work includes developing data models that can be integrated into the required testing of every patient for contagious and antibiotic-resistant staph bacteria MRSA, or Methicillin-resistant Staphylococcus aureus.

Their model, integrated into an EMR workflow, calculated with greater precision if a patient was likely or not to pass the bacteria on. The new workflow cut the number of patients who needed to be tested in half with no increase in MRSA.

“The most important implication of programs like this is that actionable intelligence, with thousands or millions of data points, can be brought to bear to improve someone’s health,” Robicsek said.

Likewise, the team developed a program to streamline an ID system for high-risk patients that involved reaching out often to such patients with reminders about taking medications and other communications. The NorthShore team’s program automatically gathered input on patients from diverse sources and updated  the patients’ records every day, so case managers would have it during daily record review. The program dramatically cut the work case managers had to perform to keep the records current.

“In the past, hospitals had to have an army of caseworkers to keep up with this record keeping,” Konchak said. “Every day, this new system looks at hundreds of thousands of patients and updates their records.”

The session “Actionable Analytics: From Predictive Modeling to Workflows,” is slated to take place on March 1, 2016 from 10-11 a.m in the Sands Expo Convention Center Human Nature Theater. 

Twitter: @HealthITNews