Do you know your patient harm 'risk score'?

Software crunches numbers to warn of high-risk situations

Some hospitals' ICUs are going beyond simple checklists to improve outcomes by leveraging software and big data to identify opportunities to improve care, Laura Landro reports for the Wall Street Journal.

The death rates in ICUs—which treat more than five million patients per year—range from 10% to 29%. Many ICUs have made progress reducing preventable errors by using best-practice checklists, but some have taken further strides by turning to years of data from electronic health records (EHRs) and other sources.

Hospitals try to make ICUs less chaotic for patients, families

For example, Beth Israel Deaconess Medical Center, along with researchers from the Massachusetts Institute of Technology, developed a system that mined 2012-2014 data from the hospital's seven ICUs to identify situations that increased the risk of patient harm, including a:

  • High number of admissions;
  • Larger number of sicker patients;
  • Greater percentage of nurses with less than one year of experience in the ICU; and
  • High patient-to-nurse ratio.

The team identified approximately 30 harms—such as medication errors and ICU readmissions post-discharge—that are linked with these risky situations.

To combat such scenarios, the hospital's system collects data in real time from providers and from EHRs, which it uses to generates a risk score. That score is then displayed on monitors and hand-held devices in the ICU.

"We can now tell people ahead of time when they are entering a risky state," says Kenneth Sands, chief quality officer and SVP at Beth Israel. Patricia Folcarelli, senior director of patient safety at the hospital, explains the ICU staff can respond to an elevated risk score by changing staffing levels, postponing elective procedures, or shifting patients to less-busy ICUs.

Massachusetts adopts nurse staffing mandates for ICUs

Johns Hopkins Medicine is also culling data from EHRs and bedside sensors to arm its providers with a "harms monitor" on tablet devices and in patients' rooms. The monitor reminds ICU staff when patients should receive care and of risks to patients, such as the need to adjust the angles of certain patients' beds (Landro, Wall Street Journal, 6/25).

Takeaway: More hospitals are using big data and predictive analytics in the ICU, in an attempt to improve outcomes and identify the riskiest patients.

Next in the Daily Briefing

You're fired: How managers can do it nicely

Read now