Thomas Jefferson University Hospital in Philadelphia spends almost $1 million annually on collecting and documenting data for CMS’ Hospital Quality Incentive Demonstration (HQID), but when it came time for CMS to distribute bonus payments, the hospital—like many other participants—saw very little.
Nonetheless, Geno Merli, FACP, chief medical officer at the 957-bed hospital, considers the money and effort well spent. “It's critical that the data is accurate so we can strategize an education program to improve our performance,” he said.
Results from the first three years of the CMS demonstration, which paid out nearly $9 million in incentive payments to 115 top performers, suggest that while bonuses were something to shoot for, they weren't the primary motivators for most participants. An analysis of the results published in the June 6 Journal of the American Medical Association, for example, compared HQID hospitals to other hospitals that participated in a public-reporting program that did not use financial incentives and found similar improvement in the care of patients with acute myocardial infarction (MI).
“It's the idea of actually receiving feedback and comparison to peers that drives better performance,” said Eric D. Peterson, MD, a cardiologist at Duke University School of Medicine and one of the study's authors. “It wasn't the money driving better performance.”
Instead, some hospitals viewed their participation in the HQID as an opportunity to implement a tracking system, identify areas for improvement and see how they stacked up against other hospitals. Many hospitals had never quantified how well they complied with basic evidence-based measures—such as aspirin at arrival and discharge for patients with acute MI—and discovered that they were less than 100% compliant.
“There is a gap in what clinicians assume is being done and the actual reality of the practice,” said Richard Bankowitz, FACP, vice president and medical director at Premier Inc., a nonprofit hospital alliance charged with managing the demonstration project for CMS.
The HQID tracked the performance of more than 260 hospitals in five clinical areas: heart failure, pneumonia, acute MI, coronary artery bypass graft and hip/knee replacements. Premier reported that, overall, participants improved quality in those areas by 11.8% in the first two years of the project. And, in contrast to the JAMA analysis, an article in the Feb. 1 New England Journal of Medicine found that participation in the HQID was associated with 9.5% overall improvement in composite process scores on 10 measures compared with a 5.2% overall improvement among hospitals that voluntarily reported information through a public-reporting initiative with no financial incentives.
Perhaps most significantly, many HQID participants saw the investment in quality improvements as worthwhile regardless of whether they ended up qualifying for financial rewards.
Jefferson University Hospital, part of the 12-hospital Jefferson Health System, embraced a system-wide mandate to make the pay-for-performance (P4P) program work, said Dr. Merli. The hospital's $1 million investment went toward paying the salaries of seven documentation specialists and two data analysts hired to work with and ensure the accuracy of the HQID data.
Making staff accountable for successful implementation of the measures was another key element in Jefferson's strategy. The chief medical officers and chief nursing officers from all Jefferson Health System hospitals meet once a month to review their performance measures. Within each hospital, departments routinely present their individual HQID scores to a hospital-wide committee.
The dean of the medical school has mandated 100% compliance with the HQID measures, said Dr. Merli, and department chairs are asked for explanations if their departments fall short. While individual physician performance is not revealed when the chief medical officers and chief nursing officers convene, department chairs know who is not hitting the targets and are expected to get those physicians on board.
That's not easy, Dr. Merli said, because physician buy-in is one of the main obstacles to implementing a P4P program. Physicians often complain that there is not enough time to comply with HQID documentation requirements—such as explaining contraindications that justify not giving a beta-blocker to a patient with acute MI at admission—when they are juggling patient rounds with other responsibilities.
“They would say this is added work, and how are you going to help us do it?” Dr. Merli said of Jefferson physicians' response to the HQID initiative.
To support physicians, the hospital's electronic medical record system has been modified to prompt them on several measures. For example, a pop-up box reminds physicians to provide smoking-cessation counseling and immunizations for influenza and pneumonia, as well as providing discharge instructions.
At Mission Hospital, a 317-bed hospital in Orange County, Calif., Fara Kardan, ACP Member, president and medical director of Advanced Hospitalist Medical Group (AHMG), knows exactly how his group—five full-time hospitalists and nine part-time physicians who mostly cover night shifts—scores on a host of performance measures. Each month Mission Hospital administrators provide a list of measures that includes some, but not all, of the quality indicators collected for the HQID project. Dr. Kardan said the subset represents the performance measures over which physicians are believed to have the most control.
The report also shows how the group compares to internal peer groups. However, Mission decided not to continue in the project after the initial three-year period. Although administrators declined to be interviewed about this decision, Dr. Kardan said that Mission's HQID scores were very good and that its continued monthly monitoring of physician performance shows that it values measurement and reporting.
Dr. Kardan's group first contracted with Mission, part of the St. Joseph Health System, in December 2005, and he takes pride in the group's scores, which are mostly in the 90% to 100% compliance range. He attributes that performance to three things:
- Multi-disciplinary morning rounds—seven days a week—in which the medical director, pharmacists and care managers help make sure the patient is receiving high-quality care.
- A cardiac nurse practitioner assigned to track patients with congestive heart failure and acute myocardial infarction diagnoses subject to performance measurement. “She is involved automatically and, for example, maybe puts a note in the chart for the doctor [saying] ‘Document ejection fraction and make sure to consider angiotensin-converting enzyme inhibitor or beta-blocker on discharge,’ “ Dr. Kardan said. “For the physician, this is like a double-check for prescribing appropriate medications at the time of discharge.”
- Monthly meetings in which he coaches the members of his group to embrace performance improvement.
“If [performance scores] look good, I say, ‘Keep doing the same thing,’” he said. “And if there's an area that needs improvement and reinforcement, I will mention it with the numbers so they know to pay more attention to that. It is reinforcement that our work is about improving quality of patient care.”
The AHMG does not receive bonus pay for achieving high scores, but Dr. Kardan believes the scores help document the value that the group brings to the hospital. For that reason, he keeps quality indicators a high priority.
Improved patient care
At South Miami Hospital, participating in the HQID actually paid off in dollars. It is part of Baptist Health South Florida, which received more than $143,000 in bonus payments for its Year Two performance.
South Miami, with 445 beds, scored in the top 20% of hospitals for heart failure, pneumonia, coronary artery bypass graft and hip/knee replacement outcomes. Shortly after the demonstration began, the hospital contracted with a hospitalist group. Rudolf Gausling, ACP Member, a partner in that group, said the move to the hospitalist model was not specifically for HQID but rather for the bigger goal of building a quality-oriented infrastructure.
Indeed, South Miami administrators say they did not incur any costs specific to HQID because the hospital had positioned itself to participate in other quality improvement initiatives. The extra nurses hired to support the Joint Commission's Hospital Core Measures program, for example, handle data gathering and reporting for HQID.
Dr. Gausling believes his group of five hospitalists, who admit most of South Miami's patients, are one key to the hospital's HQID success—but only one.
“You can have the best physician in the country going to a hospital that does not have a support system, and you will not succeed,” he said. “You are only as good as your infrastructure and your management and planning.”
At South Miami, that includes an electronic medical record system that gives physicians quick access to a patient's history and test results, making it easier to focus attention on performance measures. Even more helpful is a team of nurses, case managers and others who continually check patient charts to make sure the performance measures are being hit and documented.
“Over time it became much easier because everybody in the hospital understands what the targets are,” Dr. Gausling said.
But there is no rest for the HQID winners, said Ellen Redick, South Miami's director of quality management.
“Every single hospital looks at the data and says, ‘If we're in the second decile, here's where we need to be,’ so they are very busily improving whatever needs to be fixed,” she said. “The bar is being raised, which is a great thing for the patient. And you better keep up.”
Lola Butcher is a freelance writer in Springfield, Mo.
CMS to draw upon lessons learned
As CMS prepares to expand pay-for-performance (P4P) to all hospitals—currently planned for 2009—it will draw on the lessons being learned in the Hospital Quality Incentive Demonstration (HQID) project, Mark Wynn, director of payment demonstrations, said.
Richard Bankowitz, FACP, vice president and medical director at Premier Inc., a nonprofit hospital alliance charged with managing HQID for CMS, summed up four lessons from the project's first three years:
- The greatest benefit to patients comes when hospitals reliably deliver a bundle of appropriate care, rather than a single measure. Thus, programs should reward hospitals for a bundle of measures instead of individual indicators.
- Because complications and medical errors are strongly associated with the high cost of care, readmissions, mortality and disability, P4P programs should provide incentives to avoid them.
- Systemic improvement is nurtured more effectively by incentives and transparency than penalties.
- P4P programs should address the fragmentation, duplication and after-the-fact inspection currently common in today's health care delivery system.
Dr. Bankowitz believes these lessons will particularly resonate with hospitalists, who, by the nature of their work, share the big-picture goals of the HQID.
“Hospitalists are in a unique position to lead the institutions by taking a systems view of hospital-based care delivery—a view that is not generally possible in the more traditional setting of treating each patient in a ‘one-off’ manner,” he said.
More good news for hospitalists: Dr. Bankowitz believes the forward march of P4P will strengthen the business case for the hospitalist model.
“The increasingly frequent linking of performance to reimbursement will now provide a new potential value because hospitalists can demonstrate a real impact on the revenue of the institution,” he said.