Paul Nagy, PhD: Improving quality with data

Forget the 98,000 preventable deaths that occur in hospitals: The latest number is between 200,000 and 300,000, making medical error the third leading cause of death behind heart disease and cancer.

“For me, this is the cause of what I spend my time doing,” Paul Nagy, PhD, director of quality, Johns Hopkins University radiology department, told a crowd at SIIM during the session “Quality Improvement, Harnessing Informatics,” in Long Beach on May 17.

Nagy shared tips on creating a culture of quality, engaging employees, and using informatics to make it happen. “A lot of our tools provide great power, but they also can be contributory factors,” he cautioned.

Some hospitals employ a cadre of experts with clipboards to implement quality, but Nagy believes that can be detrimental to creating a quality culture. “People on the front lines have the capability to solve their own problems,” he said. “The people doing the work are best able to improve it.”

When implementing a culture of quality, keep in mind the following four principles:

  1. Focus on the patient.
  2. Empower the frontline.
  3. Help bridge boundaries.
  4. Do it digitally.

“I believe that informatics is the Archimedes lever of quality improvement,” he said.  “We are missing a major opportunity to improve quality and create new value by leveraging informatics.”

At present, information technology is powerful, but too slow. “Making a change in RIS or PACS can take six months,” he said. “Even getting data out can take weeks.” Quality improvement teams need access to data, and imaging informatics teams need to provide it, he said.

“What would happen if you combined the powerful change engine of a quality improvement program with a rapid application development lab?” he asked. “Crossing informatics with quality is really what you need: they are two side of the same coin and the coin is change.”

Mission-critical Tools

To provide the clinical teams on the frontline with the tools they need to make change, Nagy and his informatics team did a lot of work trying to glue systems together.

The first key piece is a context manager to make RIS and PACS talk to each other, integrated on the desktop. “If you are not doing this, you are causing medical errors,” he said. “If the radiologist has to transcribe the medical record number to the PACS, I guarantee that you will be (periodically) reporting on the wrong patient.”

Nagy also siphoned the message of each patient study they were addressing onto a website and then that web site was able to control the desktop as well. "Being able to get into the frontline was really important to us," he said.

Next, they bought a $49 forms engine (Survey Monkey is an example of a forms engine) to enable on-the-fly construction of questionnaires and put it onto the web site, hosted internally in the department and behind a firewall to safely host PHI.

If technologists wish to do peer review on each other, Nagy can build a form in five minutes, and the quality project is launched. “Because we are able to siphon this field between the RIS and PACS, whenever I go to the web site, it fills in the fields automatically for them,” Nagy explained.

Before this tool was launched, Nagy may have had to tell the technologists, “We don’t have that field in the RIS, I have to go talk to the vendor.” What may have taken a few weeks or months happens in a half-hour meeting.

“You are talking about weeks of process,” Nagy enthused. “Here, I can build the form with the user, and it actually constructs the database on its own, because you are building the forms directly. When people submit issues it can be emailed to the user.

In a one-year period, he launched more than a dozen quality projects using this tool, including:

  • CT technologist image quality audit,
  • MRI quality visual checklist,
  • MR weekly ACR quality control tool,
  • radiologist protocol correction form, and
  • radiologist IT training evaluation form.

People come together, they form a small team, they identify a problem, they build the forms, collect the data to see how they are doing and then start using Lean, Six Sigma, and other aggressive change management techniques to drive change in their environment, Nagy said.

“Instead of taking 6 months, you can really run at the same speed as your QA team,” he said.

Nagy celebrates rather than bemoans the fact that his department has the greatest number of quality issues being reported.  “It’s called a surveillance bias,” he said. “The more you look for, the more you find.”

The bigger pool of data enables root cause analysis, quality improvement plans, training and ultimately the development of better processes.

“What's the great invention that is going to happen over the next 20 years?” Nagy asked in conclusion. “We know how to treat patients, but we are not doing it in a consistent effective way for all patients. Our real invention will be trying to deliver this in a consistent fashion through these quality improvement methods and reengineering healthcare.”