Structured Reporting: The Cardiologist’s View on the Value of Quality Data Collection

After pushing for years to build a unified cardiology database powered by structured reporting, several members of the heart team at 237-bed Providence St. Patrick Hospital in Missoula, Mont., finally got their moment around six years ago. It came when their primary aim—keeping up with changing requirements for quality reporting—meshed with the institution’s decision to acquire and install a new cath lab, which also required a leading edge CVIS and a dictate to make physician adoption a priority.

“We hadn’t been satisfied with the quality of the data we were getting,” recalls Mark Sanz, MD, an interventional cardiologist and medical informaticist who helped lead the charge on quality reporting. “Clinical data abstractors were joining and leaving, and doctors weren’t taking ownership of their own data.”

Sanz recalls it well, since he served as director of vascular medicine at St. Patrick, one of 34 hospitals in the multistate Providence Health & Services system. He remembers the team reaching consensus on the idea of using the opportunity to add structured reporting which they’d spoken about but not yet moved on.

“We decided we would attempt to do structured reporting,” he says. “But it would have to not be an onerous burden. It would have to be built into the workflow we were already doing. At the time, that second part seemed like a tall order.”

The team had been using an early-generation CVIS as a back-end addition to the National Cardiovascular Data Registry (NCDR). The approach worked as a basic way to manage data, but, due to the lack of physician ownership of data, many datasets were incomplete.

The door was open to turn a minor problem into a major opportunity for quality improvement.

“Along with structured reporting, we also wanted to include echoes and have everything in one database, including images,” says Sanz. “We had multiple products from various vendors. We wanted to build one solution for everybody.”

They went shopping and, with its single database design, the McKesson Cardiology CVIS caught their eye. The system would offer every member of the heart team access to images, reports and waveforms.  It would also support workflows used not only for cardiac- and peripheral-cath procedures but also for hemodynamics monitoring, electrophysiology, echocardiography and vascular ultrasound.

For Sanz, one feature sealed the deal. “We looked at all the cardiology IT vendors,” he says, “and McKesson was the furthest along with building structured reporting.”

Big buy-in by open invitation

At Providence St. Patrick Hospital, as at all provider organizations tapping into new technology, selection, acquisition and installation were one thing. Getting people to make the most of the new CVIS could have been another. But Providence set their sights and focused a persuasive team on adoption.

“We had an age range of physicians from mid-40s to early 60s,” Sanz says. “That’s important because adoption by people who didn’t grow up with computers typically is not very high.”

Concerned that the older docs might resist, Sanz and other departmental leaders involved these colleagues “early and often,” meaning from the start of the selection process through to implementation and everything that followed.

“Once we chose McKesson, because it could provide a single source of everything — reports and images, echo and cath—we found that our echo docs had very little trouble,” he says. “They were used to structured reporting. Cath lab docs were another story,” but they too came around.

Working with McKesson support staff, Sanz built the first model report. All physicians were encouraged to comment, and, before long, all agreed on a single report format. Next came bringing the team together to talk about the need for each physician to fully “own” his or her own data.

“That was a pretty novel discussion,” says Sanz, adding that, for him, the experience provided a reminder of an age-old principle. “Relationships matter. If you just go to people with a solution that a small group made up but everybody didn’t buy into, you’re going to get a lot of push-back. And if you’re in a hospital with multiple competing groups, you must have representation from all of those groups when you start a design process.”

Documentation: Done

Asked to elaborate on the design process, Sanz describes the scenario by which he and other physician-CVIS power users adapted the system to the team’s needs and preferences.

He points out that, by the time a patient is in the cath lab, for example, a wealth of information should already be available to the interventional cardiologist. For starters, there’s pre-hospital data and clinical data.

“Is the patient diabetic? Does the patient have hypertension? Is this for unstable angina or MI? What is the indication? And of course, you have the clinical history of the patient,” he says. Before deploying the CVIS, such key clinical insights were received by nurses speaking with patients in the observation unit prior to the start of the procedure.

Most data entry was handled by an abstractor, and the data entered may or may not have been accurate, says Sanz. “Or a tech would enter it during

a procedure while a physician called out things like, ‘right coronary, 50% lesion.’ They didn’t always fill in the TIMI flow. They never filled in the complications, when there was a significant dissection or perforation or whatever.”

The physician would have to take care of such details later, he says, reiterating his point about struggles over physician ownership of data.

“With this system, you’re in control of your own data,” says Sanz. “Now that we have appropriate use criteria as well as structured reporting, you have the opportunity to make sure you have documented that the patient entered the cath lab with cardiogenic shock or had pre-existing class III heart failure, or really did have unstable angina versus stable angina.”

“All of these things are not really done well by people who are not medical, so that’s the way it’s done now,” he adds. “When appropriate use criteria came out, my docs quickly realized that they were in control of how their risk adjustment occurred: by proper documentation. And it wasn’t hard for them to understand the need to properly document urgent versus emergent versus salvage and so on.”

“They learned the definitions very quickly,” says Sanz, “when they got their first reports showing their risk adjusted complications and mortality.” As some people note: No outcome, no income. This is important stuff.

Quality data, happy hospital

The reports included NCDR data, showing comparisons of Providence St. Patrick’s outcomes with those of peer hospitals in the Providence

system as well as hospitals throughout the country. This provided a powerful incentive for any stragglers to get with the new system.

The new system’s ease of use did the rest of the incentivizing. McKesson Cardiology’s reporting tools are customizable, making it simple for users to quickly document and analyze their care practices in a way best suited to their personal preferences. They also can complete reports and other tasks remotely, via workstation-quality web access. 

It is clear that Providence Health’s decision of six years ago to acquire topnotch cardiology IT capabilities along with a new cath lab has paid off handsomely.

“When we looked at risk-adjusted rates of complications and mortality, it became really clear that we needed to have clean data—and our hospital has done very well,” says Sanz. “Hospital leadership were happy. They saw that the data were clean once the physicians took ownership.”

And the physicians themselves? “Sharing the reports, in the end, is really important,” says Sanz. “It’s important that everybody gets to see the benefit of their work”—including their work learning, using and making the most of CVIS.