ACC: Decision-support tool improves PCI outcomes, lowers costs

Implementing a real-time clinical decision support tool to identify PCI patients at high risk of bleeding events reduced complications and costs, according to a study presented March 10 at the American College of Cardiology (ACC) scientific session in San Francisco.

Craig Strauss, MD, MPH, a researcher at the Minneapolis Heart Institute Foundation and a cardiologist at t he Minneapolis Heart Institute at Abbott Northwestern Hospital in Minneapolis, and his colleagues developed the decision-support tool that incorporated a validated pre-PCI bleeding risk score. The tool was designed to guide physicians by stratifying risk and helping them choose appropriate strategies, such as treating at-risk patients with the antithrombin drug bivalirudin (The Medicines Company), which has been shown to reduce peri-procedural bleeding.

Estimates of PCI bleeding events range from 3 to 9 percent, depending on the definition of bleeding, and their occurrence increases the risk of short- and long-term morbidity and mortality, hospital length of stay and costs. To assess the tool, Strauss and his colleagues compared rates of complications, red blood cell transfusions, bleeding within 72 hours, mortality, length of stay and variable costs pre- and post-protocol implementation: July through December 2009 and August through September 2012.

They found across-the-board improvement:

  • Complications decreased from 22.8 percent to 14.9 percent;
  • Transfusions from 11.3 percent to 6.4 percent;
  • Bleeding within 72 hours from 7.7 percent to 2.1 percent;
  • Mortality from 5.7 percent to 5 percent;
  • Mean length of stay from 2.9 days to 2.3 days; and
  • Costs from $15,092 to $14,217.

Adjustments after the health system closed its fiscal year showed more dramatic cost savings, Strauss told Cardiovascular Business, from a baseline of $15,092 to $12,641.

The decision-support tool has changed practice within their system, he said. In March 2012, they started using the tool in two of the three PCI centers across the Allina Health System and added the third hospital in May.

“We tracked how many cases actually received a risk score and it runs about 93 percent of all PCI cases,” he said. “Since May, we have gotten the adoption for bleeding avoidance strategies for high-risk patients from roughly 25 to 30 percent up to 65 to 70 percent.” He added that strategies include the use of radial access and closure devices as well as bivalirudin.

The study has shown that a validated risk score can be calculated in real time and successfully implemented in a high-volume PCI system. “This can achieve changes in physician practice patterns that reduce variation and improve outcomes,” he said.

Other institutions can implement a similar tool, according to Strauss, because the bleeding risk score is publicly accessible and several software companies offer products that use bleeding risk scores.