CVIS: Improving Departments from Within
Properly utilizing and disseminating data gathered from a CVIS requires a systematic approach that should extend throughout a healthcare network, says Mark Watts, imaging systems manager at Iasis Healthcare. The deployment of a widespread IT project needs complete administrative support, and should be bolstered and reviewed by a physician champion. He installed a CVIS at his previous employer, Provena Healthcare, a six-hospital system based in Illinois, and currently is undertaking a similar, but larger scale implementation with McKesson’s Horizon Cardiology CVIS at Iasis, a 15-hospital chain that spans six U.S. states.
Once deployed, hundreds of reports can be acquired via the CVIS, and certain providers are choosing to change processes based on these data, with the intention of improving quality measures and patient outcomes within their facility.
Methodist DeBakey Heart & Vascular Center in Houston was an early adapter of CVIS when in 2003 it implemented a Digisonics system, says director Marti L. McCulloch. The center has 10 operating rooms, eight cardiac catheterization labs, 154 acute-care beds, 48 ICU beds and 30 transplant beds.
Methodist tracks quality measures, such as study turnaround times and patient report turnaround times, which can be pulled from the database to assess physician performance. Then, the provider disseminates the data through a quality metrics dashboard, which includes report turnaround times and stat turnaround times, to ensure processes are completed within the provider’s mandated time frames. For instance, STAT EKGs need to be done within 15 minutes.
“Regardless of the superiority of the clinicians or their interpretation, if the report doesn’t get sent off to the referring physician in a timely manner, its relevance is lost,” explains McCullough.
In fact, when the CVIS was first installed, the facility discovered its vascular lab had an average turnaround time of 10+ days. Now average turnaround time is 24 hours.
“With CVIS, we are building transparency into our system to identify opportunities for improvement, not to admonish under- performers” Watts says.
McCullough concurs: “It’s important to have performance very transparent because physicians and their staffs tend to be competitive, and each team wants to perform the best.”
Barry Szakolczay, data specialist for the cardiovascular services at Scottsdale Healthcare, a three-hospital network with a total of about 850 beds in Scottsdale, Ariz.—which went live with its GE Healthcare’s Centricity DMS [Data Management System] in late 2006—says his institution generates between 70 and 90 reports per month from the CVIS.
Scottsdale Healthcare frequently monitors length of stay in both the cath lab and cardiac surgery. With those data, length of stay from admission to procedure, length of stay from procedure to discharge, and total length of stay can be reviewed by cardiologists and CV surgeons. Also, surgery data are entered into the Adult Cardiac Registry of the Society of Thoracic Surgery (STS), allowing STS to evaluate outcomes on a national level.
One of the key data elements the society tracks is patients with a hospital length of longer than 14 days. Data collected in the cath labs also is entered into an American College of Cardiology (ACC) registry, allowing the provider to compare outcomes with those on a national level.
“If a surgical patient had an extended stay of 14 days or beyond, our surgeons would review that case to assess why the patient stay was prolonged. But the only way we are able to track these factors on a frequent basis is through reports issued from the DMS,” Szakolczay says.
Also, the Joint Commission requires providers to track their critical results, and much of those data collected from CVIS need to be sent to Performance Improvement. On a quarterly basis, a random sampling is aggregated, McCullough says. In addition, a critical results indications component of the CVIS allows easy querying of critical results data. “As a result, we can easily query which indications are applicable and produce the appropriate report for the commission,” she says.
Scottdale’s CVIS has a statistical risk expected report which extracts data from an STS module, and examines the outcome/expected outcome (O/E) ratio of patients for such events as mortality, renal failure, re-op, prolonged ventilation, deep sternal wound infection rate, and so on. Also on this report, patient characteristics and co-morbidities are available from the database. The STS module assigns a risk score to each patient for each of the categories, and that information is pulled onto a report from the CVIS. Biannually, Szakolczay and his colleagues meet with the surgeons to review expected outcomes, compared with actual outcomes.
Szakolczay says that the “surgeons rely on these data very heavily, and look forward to seeing the report and where they stand, as well as where they can make improvements. For example, at one time we had higher rates of deep sternal wound infections, and these reports allowed the surgeons to see how many of their patients were acquiring these infections. As a result, we changed our processes, and greatly reduced our adverse event rates with deep sternal wound rates to next to nil.”
Likewise, at Iaisis Healthcare the CVIS produces “hard facts that allow managers to have serious, productive conversations with their physicians, who in turn, have tremendous clout within the community and within the organization,” Watts says.
Within the cath lab, there has been a tremendous push to reduce door-to-balloon times associated with percutaneous coronary intervention (PCI), especially for patients who have experienced an acute heart attack.
For the ACC/AHA databases on door-to-balloon times, figures can be captured in real-time via hemodynamic equipment and fed back into the CVIS while the case is ongoing. The report also can indicate when the ED calls the cath lab, ED triage times, the time of the first EKG, when the cardiologist was called, when the patient arrived at the cath lab and when the first balloon was inflated.
“All these different times are captured in the CVIS report, clearly indicating which areas need improvement, and therefore, have helped facilities greatly improve door-to-balloon times,” Szakolczay says. “In fact, any case that takes longer than 90 minutes is reviewed at Scottsdale to determine where the delay occurs.”
J. Brent Muhlestein, MD, professor of medicine at the J. L. Sorenson Heart & Lung Center, Intermountain Medical Center in Murray, Utah, also speaks to the importance of CVIS data in tailoring procedures to produce the best outcomes for patients. Muhlestein and his colleagues have been conducting small trials to determine best practices based on CVIS data from across Intermountain’s 21 hospitals.
For example, Muhlestein used the CVIS to help determine which anticoagulants are beneficial in the cath lab based on risks associated with bleeding and ischemic complications.
Using these data, “we were able to more effectively define our institution’s protocols, and also published our findings in journal articles to disseminate the information to other departments.” Based on data from the CVIS, Muhlestein and colleagues were able to show that bleeding risks associated with glycoprotein IIb/III inhibitors is significantly reduced, if the heparin dose is reduced.
Leveling outcomes across a network
The changes in quality of care can even occur from a distance, such as when a CVIS pulls data from multiple facilities, and is made transparent throughout the network.
“With these data, the hospital can be assured that the person on third-shift at a small hospital in a rural location is providing the same high quality of care that is being administered in the flagship hospitals mid-day,” Watts explains.
Watts also notes that even if a physician is not on-site, the CVIS can still help improve patient care across a network. For instance, if a cardiologist is called upon to perform a primary PCI case, they also may have a patient who had an echocardiogram performed at another facility who is waiting to be released, pending diagnostic results. The physician can return home after the PCI, finalize the report in the CVIS remotely and the patient can be released without waiting for the cardiologist to return to the facility.
“If a facility performs about 500 operating room procedures annually, then a half-day reduction in length of stay results in about $250,000 in savings—because the average daily cost is approximately $1,000,” Watts says. “Not only is it good medicine, compassionate and appropriate to send well patients home, if we can, but it’s also fiscally beneficial to the organization to reduce length of stay.”
To monitor expenses, Scottsdale Healthcare issues an inventory report via the CVIS to its cath lab supervisors on a monthly basis. In addition, Szakolczay runs a weekly device implant report for electrophysiology to examine vendor market share, assessing how many devices are being used from which vendor and which percentage of the hospital’s market share they are consuming. That report gets issued to the director of the cardiovascular service line, as well as to the medical director for electrophysiology. Then, on a monthly basis, a similar report is distributed to the cath lab managers and supervisors and contract management.
“This information allows contract management to negotiate better prices with the vendors,” Szakolczay says. “The report plays a big part in negotiations and keeping contract management abreast of what is being used, and therefore saves on the bottom line.”
Muhlestein and his colleagues at the J.L. Sorenson Heart & Lung Center use CVIS data to determine outcomes of various bare-metal and drug-eluting stents within their department, and having found very few different outcomes among the various brands, the managers negotiated better prices based on their clinical outcomes. Likewise, through gathering data at Intermountain’s facilities with the two highest interventional volumes, they also evaluated drug-eluting stents and found similar restenosis outcomes compared with the cobalt chromium bare-metal stent. The latter, less expensive stent provides a nice alterative if a patient is contradicted for a drug-eluting stent.
Watts agrees that inventory management through CVIS reports has been a major source of cost savings. “We could perform a statistical analysis on all the stents used in our 15 hospitals,” he says. We could first gather how many are being used annually, and then drill down into that data to discover how many each physician uses, allowing the organization to see if a particular physician needs to be educated about overuse of such expensive equipment. This opportunity to target potential waste through longitudinal data over time can lead to cost savings across a department.
For example, data can be collected on stenting procedures performed by 40 different cardiologists. The resulting data can indicate an average of 2.2 stents are used during each procedure. “Overall, this isn’t an alarming figure, but when those data are drilled down per stent, per physician or whether the case was planned or emergent, then department managers can begin to evaluate physician practices which could be costing the department unnecessary expenses,” he says.
However, McCullough says, after adding CVIS, Methodist reaped its largest cost savings from a reduction in clerical manpower needed to process the results in the diagnostic labs, now that physicians are responsible for their own reports.
Outside the walls of cardiology
So the data prove it. CVIS not only change cardiology departments, but change perceptions and process throughout the institution. “Unfortunately, for years, cardiology departments have been viewed as islands of data onto themselves. The cardiologist, therefore, would work within the hospital, but not necessarily with the hospital. Now, cardiovascular information systems make available data, which was previously stored on paper or videotape, to be shared in a collaborative effort with the physicians throughout a health system,” Watts says.
To realize these possibilities, Muhlestein says that the CVIS of the future should be able to generate procedure reports and incorporate more graphical and still images with better interfacing with the other hospital information systems, in addition to existing as a complete database system for the cardiology department.