Skip to content Skip to navigation

Aggregating Clinical Data at the Point of Care

June 23, 2011
by Jennifer Prestigiacomo
| Reprints
Providence Health rolls out system-wide clinical decision support tool

As a part of Providence Health & Services’ clinical safety and quality strategic plan from the early 2000s, the system is now rolling out an early clinical warning system to engage clinicians at the point of care through technology to enable informed clinical interventions. Providence Health & Services, based in Renton, Wash., operates 27 hospitals and other non-acute facilities across five states, and will be using the Redmond, Wash.-based Microsoft Amalga UIS Modified Early Warning System (MEWS) to monitor early clinical signals among all inpatients to help prevent infections, escalations to intensive care, and ultimately fatalities.

At the end of 2008, Providence started in earnest its assessment phase of the early clinical warning system, and in July 2009 decided on a proof of concept with Amalga, which uses a complex algorithm published in the medical journal Lancet involving many data elements that predict early signs of heart failure. The Amalga system aggregates admission, discharge, transfer (ADT) feeds; lab orders and results; radiology orders and results; medication orders; clinical vitals; and nursing documentation, and based on the algorithm creates a score from 1 to 5 indicating the criticality of the patient. The patients are sorted in decreasing order, with more acute patients on top, and displays current scores, previous scores, and changes in the score over the entire inpatient stay.
“[Clinicians] can double click on the patient row and see the whole pantheon of clinical data available from the system, and they can drill down see all laboratory values, all microbiology values and more detailed information on vitals,” says Paul Tittel, system director, enterprise Amalga and data services, Providence Health & Services.

For the pilot, Providence picked two smaller regions, one Alaskan hospital, Providence Alaska Medical Center, which had been operating an electronic health record (EHR) from McKesson (San Francisco), and four hospitals in California on Meditech (Westwood, Mass.). The entire process of launching the Amalga infrastructure and connecting it to Providence’s clinical systems took three months, and as John Kenagy, Ph.D., vice president, chief information officer, notes, it took only a short time to provide value. The Amalga initiative has since grown to 90 data feeds from eight different regions, managed by a project team in five states. Now the team is working on a system-wide implementation, with a goal of putting the MEWS system in 23 of Providence’s inpatient facilities by the end of this year.

Roy Davis, M.D., chief medical officer, Providence Alaska Medical Center, integrated this warning system into the responsibilities of the hospital’s existing rapid response team. A nurse or other clinician from the team monitors the Amalga reports several times a day to check on patient criticality scores and intervene if necessary. “Even without pro-active alerting out of the Amalga platform—which won’t be supported till the next product release—they really saw marked improvements in certain key outcomes that included decreased code blue events, increased rapid response team calls, decreased care escalation events, less frequent transitions of patients from standard inpatient units into the ICU because they caught impending inpatient problems earlier and were able to intervene,” says Tittel.

Increasing Accuracy, Timeliness
Even though it’s too early to report quantitative clinical improvement statistics, says Kenagy, qualitative surveys from clinicians have provided positive results that the MEWS system has become an evaluative tool to aid their clinical decisions. “The thing that we found that really improved was our transaction system, timeliness and accuracy, even though that wasn’t really our target,” says Kenagy.

Kenagy says that even early on during the proof of concept phase he saw the value of investing in the analytical solution, which was having the ability to take data from multiple systems and do comparisons and analytics across the entire system, and have the ability to share best practices. “If sepsis management is statistically significantly better in one of our ministries, this [system] would illustrate that and [we would have] the ability to then share that best practice across all of our organizations,” he says. “We had no way to do that with our multiple-sourced systems here.”

Other possible benefits from MEWS include potential cost savings, including unreimbursed expenses associated with escalations of care to ICU and avoidance of future capital investments because of better ICU capacity management. According to Microsoft and Providence, in the first two months of use, the technology has increased the desired interventions by 40 percent, and Providence Alaska Medical Center estimates that it could save as much as $450,000 annually from unreimbursed expenses that are associated with preventable intensive care admissions.

Challenges of Physician Documentation
With a system-level approach that engaged the CMIO in each of the eight regions that have adopted MEWS, accountability and ownership of this initiative has been high. This resulted in few challenges along the way, says Kenagy, except for the large one of physician documentation “I think the biggest challenge was the lack of timely, structured data on which to base the analysis—or even in some cases, even the existence of such data,” Kenagy says. “The physician’s note is a lot more like a glob of data, rather than data that is in an atomic, molecular form to be able to reformulate.” With clinician leadership and training this challenge was eventually overcome.

Pages

RELATED INSIGHTS
Topics