Case Study: Improving Performance at Charleston Area Medical Center

November 29, 2007

Overview


By Martha Hostetter

Multidisciplinary improvement teams at a medical center in Charleston, W. Va., implemented new care processes, supported by "change agents" and rapid-response data reporting, that led the hospital to achieve benchmark performance across multiple indicators of health care quality.


This article first appeared in the November/December 2007 issue of the newsletter "Quality Matters."

Introduction: The Hospital Quality Alliance (HQA), a public/private collaborative effort to measure and publicly report on the quality of hospital care, now measures performance on 22 process-of-care measures. Specifically, it examines what percent of patients receive evidence-based processes of care for acute myocardial infarction (AMI), heart failure (HF), pneumonia, and prevention of infection related to surgery. Because the Centers for Medicare and Medicaid Services (CMS) provides higher reimbursements to hospitals that report HQA data, nearly all do so. The quality results are posted on the CMS Web site, Hospital Compare.

As discussed in this month's In Focus article, analyses of the HQA data have found that performance on these indicators varies widely and that better performance is associated with lower risk-adjusted mortality for AMI, HF, and pneumonia. [1] An analysis for the December 2005 issue of Quality Matters, completed by Harvard University researchers, used HQA data to identify the top-performing hospitals on composite measures of quality across the three clinical conditions then being measured (AMI, HF, and pneumonia). The case study for that issue looked at how one top performer, Reid Hospital & Health Care Services, was able to achieve high-quality care for multiple conditions.

For the current issue, Quality Matters asked the same Harvard researchers to once again identify the top-performing hospitals, this time using HQA data from July 2005 to June 2006 for the 21 indicators of quality care hospitals reported on during that time period (for AMI, HF, pneumonia, and surgical infection prevention). Their analysis placed Charleston Area Medical Center among the top performers. [2]

Issue: Public reporting of standard health care quality measures helps hospitals to understand how well they are performing compared with their peers, gives them targets for improvement, and provides external motivation to improve. In addition, the reports enable consumers to compare hospitals and select the best institutions for their particular conditions.

Organization: Established in 1972, Charleston Area Medical Center (CAMC) is West Virginia's largest medical center, with more than 5,000 employees. CAMC is a nonprofit, 893-bed, regional referral and academic medical center composed of three facilities: General Hospital, Memorial Hospital (which includes the nation's fourth largest cardiology program), and Women and Children's Hospital.

Dale Wood, M.B.A., M.H.A., vice president for system performance and chief quality officer, and Glenn Crotty, M.D., executive vice president and chief operations officer for CAMC Health Systems, Inc., lead the hospital's improvement efforts.

Target Populations: The medical center's improvement initiatives focus on various care processes and/or patient populations, including those targeted by the Hospital Quality Alliance reports (patients with AMI, HF, pneumonia, and those who undergo surgery).

Implementation Timeline: CAMC began to focus on performance improvement in 2000. Since then, it has initiated more than 100 quality improvement projects across its three facilities.

In 2003, CAMC joined the CMS/Premier Hospital Quality Improvement Demonstration, a voluntary program in which more than 260 facilities submit data on process-of-care and outcome measures for AMI, coronary artery bypass graft, pneumonia, HF, and hip and knee replacements. As part of the demonstration, hospitals performing in the top two deciles were promised higher payments for their Medicare patients.

In 2004, CAMC began reporting data through the HQA initiative.

Key Measures: The CMS/Premier demonstration includes all of the process-of-care measures reported through HQA. In addition, participating hospitals report on patient safety indicators and on patients' perspectives of their care, using items from the Hospital CAHPS (Consumer Assessment of Healthcare Providers and Systems).

Process of Change: Overall, the medical center's improvement process includes the following steps:
  • Implementing process improvement tools. CAMC uses the Six Sigma methodology—Define, Measure, Analyze, Improve, and Control—to examine quality problems and standardize care processes.
  • Collaborating in multidisciplinary improvement teams. The teams focus on areas for which there is good information on evidence-based care, including AMI, HF, pneumonia, and surgical infection prevention. Other teams focus on dispensing antibiotics for hysterectomy patients, reducing central line infections, and providing end-of-life care. The teams report to the evidence-based steering committee, which examines improvement efforts across clinical areas.
  • Focusing on evidence-based processes of care. CAMC created recommended courses of treatments and tests for patients admitted with particular conditions, called standing orders. Color-coded "order sheets" for AMI, HF, and other conditions are placed in patients' charts to alert providers to the recommended care. Also, to ensure that all patients receive advice on smoking cessation—one of the HQA quality indicators for AMI, HF, and pneumonia—the medical center incorporates questions about smoking and patients' willingness to undergo counseling into its registration and discharge procedures. Further, each CAMC provider carries a pocket card on which they record their department's performance targets and what those targets mean to them. For example, the pocket cards might remind them to wash their hands, provide discharge instructions to HF patients, or put red socks on patients at risk for falls.
  • Supporting improvement through clinical "change agents." Each of the improvement teams includes a clinical "change agent," typically an advanced practice nurse, whose role is to educate staff members, coordinate change processes, and ensure compliance. These change agents work on the hospital floor, reviewing patient charts and querying providers if certain recommended procedures have not been followed.
  • Collecting and rapidly feeding back performance data. CAMC's Clinical Quality Management Department collects data on an ongoing basis for all process-of-care measures reported under the CMS/Premier demonstration, and for the HQA reports on AMI, HF, pneumonia, surgical infection prevention, CABG, and hip and knee replacements. It is able to analyze and feed back performance information to department heads within two months after the data are collected. (Patient chart reviews enable real-time oversight and identification of defects.) Analysts work with providers to ensure they are accurately reporting the care they provide to patients, for example, by documenting that a patient's condition indicates certain drugs should not be prescribed.
More specifically, CAMC took the following steps to improve care for acute myocardial infarction patients:
  • The AMI team developed and implemented order sets for all patients who are admitted with AMI. It also developed standing orders for discharging cardiovascular patients.
  • Nursing staff are educated on AMI care processes, and performance targets were posted in nursing units.
  • Nurses track adherence to the eight HQA indicators of quality care for AMI patients, using purple-colored order sheets.
  • On a weekly basis, 10 patient charts are randomly pulled for review by change agents and case coordinators. The results are recorded and shared with members of the AMI team.
  • A quality improvement specialist tracks all AMI patients by going on nursing rounds and using chart review to monitor compliance with indicators.
These changes were based on evidence that giving heart attack patients proven medications, when indicated, as well as smoking cessation counseling, reduces their risk of recurrent heart attacks and death. [3] Studies also show that patients who do not receive a prescription for recommended medications at discharge are less likely to take these medications. [4]

The changes implemented by the medical center led to improvements on several AMI indicators. However, CAMC lagged behind in one area—providing percutaneous coronary intervention (PCI) to heart attack patients within 90 minutes of arrival, a benchmark that many hospitals have struggled to achieve. Performing PCI procedures as quickly as possible improves blood flow and can lessen heart damage but, as of June 2006, CAMC met this goal only 25 percent of the time.

In July of that year, the AMI team formed a special group to focus on shortening the delivery of PCI procedures by moving patients quickly from the ambulance to the emergency department to the catheterization lab. The team members reviewed all the processes that occur along this continuum of care, pinpointed opportunities to improve their delivery times, and researched what other hospitals were doing. They made the following changes:
  • EMS teams were asked to perform EKGs on patients and report the findings to care teams before their arrival at the hospital.
  • A system was put in place to alert providers and the catheterization lab to the arrival of cardiac patients. Cardiologists receive a special page that distinguishes AMI patients who need to go the catheterization lab from those who don't.
CAMC cardiologists and emergency department physicians have been instrumental in ensuring prompt response times for PCI procedures. In particular, a new group of cardiologists began to practice in the community in the spring of 2007. These providers are frequently on unassigned call in the emergency department and their performance on this indicator has improved overall response times.

Results: In the fourth quarter of 2003, CAMC's performance was at 77 percent for the HQA measures of pneumonia care. By August 2007, performance across the pneumonia indicators had risen to 95 percent.

Performance in the fourth quarter of 2003 for AMI and HF care was already high—at 90 and 93 percent, respectively. By August 2007, CAMC had achieved performance of 98 percent for measures of AMI care, and 96 percent for heart failure measures (Table 2).

Table 2: Charleston Area Medical Center: Composite Scores on Process-of-Care Indicators of Health Care Quality

 Q4
2003
2004200520062007*
Acute Myocardial Infarction9095989798
Heart Failure9392949596
Pneumonia7782898689
Surgical Infection******9395
*As of August 2007 ** data not collected for these years

Note: The 21 indicators for these conditions include process of care measures initially reported for Premier/CMS Hospital Quality Improvement Demonstration, and later incorporated into HQA measure set.


In addition, CAMC achieved a more rapid delivery of PCI procedures. By the summer of 2007, performance levels exceeded statistical control limits, demonstrating real improvement as a result of the process changes (Chart 1).

QM November07 image


For internal reporting purposes, CAMC calculates monthly composite performance scores for AMI, HF, and other conditions, which show the average performance level among all the indicators for a particular condition. It also tracks an "all or nothing" score to see what proportion of patients receive all of the recommended care for that condition. In October 2004, the medical center's compliance with the eight HQA indicators for AMI treatment was 50 percent—meaning that patients received all recommended care just half of the time. By August 2007, patients received all eight of the recommended care processes 95 percent of the time.

Implications: CAMC's leadership chose to focus on performance improvement in response to the needs of the community it serves. West Virginia's population is older than the rest of the nation (15 percent were 65 or older in 2005, compared with 12 percent nationally), so demand for services is high and is expected to continue rising. In addition, West Virginia is one of two states where medical service rates are set by state regulators. This forces hospitals to control costs because they are not able to raise prices.

The financial incentives offered through the Premier pay-for-performance demonstration and the public reporting of performance through the HQA program served to reinforce CAMC's improvement efforts, focusing them on evidence-based measures of care for common conditions. In January 2007, CMS announced that CAMC received the second-highest incentive award, $701,000, for their performance. This demonstration has been extended and CAMC continues to participate.

CAMC's administration and medical staff officers have led the medical center's improvement process, both by allocating necessary resources and removing barriers to change. Yet, system improvements could not be achieved without clinical leaders and agents of change to implement the data collection and process improvement tools, and to monitor providers' progress on a daily basis. "If you have people at the top who completely support and want these changes to occur, you can still fall flat on your face," says Wood. "You need a group of networkers who can carry change across an organization."

Benchmarking performance enables organizations to focus their improvement efforts on achievable goals; once a certain level of reliability is achieved, they can aim for higher performance or even perfection. "Certain types of improvements will only work if you have already achieved a certain level of performance," Wood explains. "If you have a 20 percent defect rate, you don't have a process—or it's a chaotic process, at best—so you can't focus on improving the process. You have to focus first on creating one. If you get to where you're performing at a higher level, you can then consider how to automate your system, so that, for example, you identify CHF [congestive heart failure] patients every time and match them with the recommended elements of care."

While hospitals may see rapid progress in some areas, it takes a high level of reliability to consistently deliver all aspects of care. CAMC's tracking of "all-or-nothing" scores helps it monitor how reliably its patients receive all recommended care.

Performance data that demonstrate steady improvement have helped to achieve buy-in among CAMC providers. "I meet with new physicians each quarter and I tell them: 'The hospital, physician leaders, and faculty have worked to develop systems and processes to support the care process. The systems and processes we've designed together will produce better results than one can do working independently,'" says Wood. "It's hard to argue with the outcomes we've seen."

Some worry that public reporting could cause health care providers to focus only on those aspects of care being evaluated, to the neglect of others. Wood argues that the systems and processes CAMC has put in place have created a foundation for improvement, leading to positive change across all conditions. CAMC leaders are now asking how much further they can go: What if they were tracking 40 conditions, and 400 quality indicators?

For Further Information: Contact Dale Wood, vice president for system performance and chief quality officer, Charleston Area Medical Center, at dale.wood@camc.org or 304.388.7168.

References
[1] A. K. Jha et al. (2005) Care in U.S. Hospitals—The Hospital Quality Alliance Program. New England Journal of Medicine 353, 265–274; A. K. Jha et al. (2007) The Inverse Relationship Between Mortality Rates and Performance in the Hospital Quality Alliance Measures. Health Affairs 26, 1104–1110.
[2] A previous version of this case study included a table listing the "Top 10" hospitals across all 21 HQA measures. The methodology used to create that list is currently under examination, and therefore we have removed the list at this time.
[3] E. M. Antman et al. (2004) ACC/AHA Guidelines for the Management of Patients with ST-Elevation Myocardial Infarction: A Report of the American College of Cardiology/American Heart Association Task Force on Practice Guidelines. Circulation 110, e82–292.
[4] J. Butler et al. (2002) Outpatient Adherence to Beta-Blocker Therapy After Acute Myocardial Infarction. Journal of the American College of Cardiology 40, 1589–1595; J. Butler et al. (2004) Outpatient Utilization of Angiotensin-Converting Enzyme Inhibitors Among Heart Failure Patients After Hospital Discharge. Journal of the American College of Cardiology 43, 2036–2043.
November 2007


This study was based on publicly available information and self-reported data provided by the case study institution(s). The aim of Fund-sponsored case studies of this type is to identify institutions that have achieved results indicating high performance in a particular area, have undertaken innovations designed to reach higher performance, or exemplify attributes that can foster high performance. The studies are intended to enable other institutions to draw lessons from the studied organizations' experiences in ways that may aid their own efforts to become high performers. The Commonwealth Fund is not an accreditor of health care organizations or systems, and the inclusion of an institution in the Fund's case studies series is not an endorsement by the Fund for receipt of health care from the institution.