Critical Path Network: Project hinges on top quality hospital data
Critical Path Network: Project hinges on top quality hospital data
Mortality rate for AMI drops 36%
A quality improvement project in Dayton, OH, achieved a 36% drop in mortality from acute myocardial infarction (AMI) among a group of hospitals cooperating on the effort, and participants say it could not have been done without high-quality data collection.
Recently, the Dayton consortium won a Codman Award from the Joint Commission on Accreditation of Healthcare Organizations for outstanding quality improvement projects.
The project was one of the first tackled by the Greater Dayton Area Hospital Association, which in 1998, formed a consortium of 20 local hospitals, area employers, physicians, and quality management professionals to support the development of accurate and comparable measures of cost, quality, and patient satisfaction.
As a result of this consortium, competing hospitals began working together to raise the quality of care they provide to the community, says Joseph M. Krella, FACHE, president of the association. Participating hospitals agreed to annually release aggregate cost and quality indicators to local business leaders.
Beginning in year three, hospital-specific cost, quality, and patient satisfaction measures were released to local business leaders.
The three-year time frame for the release of hospital-specific performance provided enough time for individual organizations to benchmark results against the aggregate data and share best practices in a collaborative effort that would benefit all, Krella says. Setting this time frame from the outset was key to obtaining the cooperation of the hospitals involved, he says, but the hospitals compared data among themselves in the first year of the project.
"Our initial report listed a number of diagnoses and comparative data for mortality, length of stay, and costs," Krella says. "The report compared hospitals in Dayton to each other, and also counties within the state and some state-to-state comparisons."
Following that initial report, the participants realized that Dayton and Montgomery County were outliers with a higher rate of mortality for AMI than the risk-adjusted model would have predicted. So AMI was targeted as one of the initial indicators for improvement. The aggregate AMI rate for Dayton was above the Ohio state average and significantly above the predicted rate. As a result of this collaborative effort, there has been a 36% drop in AMI mortality rates in the city over a three-year period.
The work of the consortium has evolved from a report card focus to a true collaborative approach to process improvements, says Joseph Cappiello, vice president of accreditation field operations for the Joint Commission on Accreditation of Healthcare Organizations.
"Without a doubt, this multiorganization team effectively used performance measures — process and outcome — and performance improvement to elevate the level of care. The performance of all of the hospitals and providers in the group improved significantly while the amount of variation between them was minimized," he says.
"The level of care provided by each is at a comparable level and continues to be improved. The entire community is practicing evidence-based medicine, and the quality of care in the community has improved as a result," Cappiello adds.
Data must be transparent to participants
To address the AMI outliers, the consortium formed a quality council made up of hospital CEOs, medical directors, and leaders from the business community. This quality council meets quarterly and is responsible for overseeing the entire project. The next tier down is a committee of medical directors, and then a steering committee made up of hospital quality managers and business leaders. Beneath that level is a process-of-care committee made up of clinicians and others involved in the particular quality issue being addressed, such as AMI.
"This has truly been a collaborative effort among organizations that have set aside their competing interests in order to improve the quality of health care provided to area residents," Krella says. "Our accomplishments would not have been possible without the commitment and dedication of hospital management, quality management professionals, physicians, and area business leaders."
Cooperation among the participating hospitals was crucial to the success of the project, says Rick Snow, DO, MPH, a physician in the community who worked with the hospital association on the project. A large part of his job with the project was to encourage cardiologists to participate by sharing data and taking a critical look at their own performance. Initially, they were concerned about whether there were problems with the data that would explain the differences in AMI outcomes, but the consortium took great pains to ensure that the data were reliable.
"One factor that came up was DNRs [do-not-resuscitate orders], which are not included in the administrative data we were using, so there were questions about how that might have affected the data," he says. "But we looked at some factors that might be associated with DNRs, like Parkinson’s disease, dementia, and stroke, and we incorporated those into the data because they might be indicative of a DNR. They were predictive and had some effect on the risk of death, so we included them in the model."
Snow says it is important that physicians see the data as "transparent," meaning they fully understand where the data came from, along with any shortcomings or omissions. Otherwise, they will wonder about the real cause for the variance in mortality rates and not focus on what can be improved.
"When the data are transparent, you’re able to move them beyond the data and the model to get them asking questions about the processes of care, to start a dialogue," he says. "A real lesson is that you have to encourage participants to continually ask questions of the data and have the infrastructure to answer those questions."
Quality of data is key
The quality of the data was key to achieving a 36% reduction in AMI mortality over three years, Snow says, and he credits quality improvement professionals with gathering the valuable data.
"The QI professionals were very instrumental as the project evolved from a peer-review model to a process improvement model," Snow says. "They are the experts in data collection, and we turned to them for that key part of what we were doing."
Data were culled from administrative data sources such as the discharge abstracts usually sent to state databases, but the consortium did not automatically accept those data as reliable. Striving to give physicians the most confidence possible in the data, the consortium had physicians compare information in actual medical records to what was in the database for a specified period. When the data checked out, the participants accepted them as true indicators of what was happening at the facilities.
Quality professionals also were instrumental in providing peer-review protection to the consortium’s meetings, working through the Ohio Hospital Association to get a special law passed in the state legislature granting peer-review protection. That helped the physicians discuss their modes of care and particular cases openly, Snow says. "They would not be as willing to bring that kind of detail to the table without some protection, and rightly so. That gets very close to issues of liability."
The data proved important not only in determining which hospitals needed to improve AMI care, but in showing what methods could improve mortality. A lack of national standards meant the consortium had to develop their own local benchmarks, using risk-adjusted mortality to identify the hospitals with the best outcomes and then look to their processes for benchmarks. Reperfusion turned out to be a significant factor, with the better performing hospitals defining ideal populations for reperfusion and carefully timing reperfusion. Smoking cessation and early use of beta blockade also were identified as best practices.
The three hospitals identified as outliers saw their combined AMI mortality rate fall from 9.88% in 1999 to 6.32% in 2001, a 36% reduction. For comparison, the state of Ohio’s AMI mortality rate fell from 8.29% to 7.48% in the same period, a reduction of 9.8%. Congestive heart failure, pneumonia, and patient safety are among the next targets for the consortium, Krella says. The group has moved from annual reports to giving participants quarterly data, which the institutions then share with individual physicians.
"We’re collecting process measures, related very closely to the Joint Commission’s core measures," Krella says.
"We’re feeding that information back to the institutions in almost a real-time manner, which leads to quicker analysis and a quicker implementation of improved processes," he adds.
A quality improvement project in Dayton, OH, achieved a 36% drop in mortality from acute myocardial infarction (AMI) among a group of hospitals cooperating on the effort, and participants say it could not have been done without high-quality data collection.Subscribe Now for Access
You have reached your article limit for the month. We hope you found our articles both enjoyable and insightful. For information on new subscriptions, product trials, alternative billing arrangements or group and site discounts please call 800-688-2421. We look forward to having you as a long-term member of the Relias Media community.