By Christina Frangou

img-button

A key measure used by the Centers for Medicare & Medicaid Services (CMS) to assess a hospital’s quality performance may not reflect quality at all, according to a new report.

Instead, a hospital’s rate of venous thromboembolism (VTE) events directly reflects how frequently the hospital conducts imaging tests to diagnose VTEs. In other words, the more a hospital looks, the more it finds.

“Hospitals may be unfairly deemed a poor performer for the outcome VTE measure if they have increased vigilance for VTE by performing more VTE imaging studies,” said study author Mila H. Ju, MD, MS, a clinical scholar with the American College of Surgeons (ACS) and a vascular surgery resident at Northwestern University, Chicago, after presenting the study at the 134th Annual Meeting of the American Surgical Association.

VTE is a controversial quality-of-care measure included in many quality improvement programs and public reporting initiatives. The CMS deemed VTE a “never event” that will not be additionally reimbursed after certain operations. Beginning in 2015, VTE occurrence will be tied to financial penalties through the CMS Value-Based Purchasing Initiative.

But surgeons have criticized the VTE outcome measure, arguing that these events do not reflect poor hospital care. Some evidence backs up this argument. A previous study from the Surgical Outcomes and Quality Improvement Center at Northwestern University found that hospitals with higher rates of VTE perform more imaging studies than other hospitals (JAMA 2013;310:1482-1489). That report, however, was criticized for its use of Medicare administrative data, which provide limited information about preoperative risk factors and exclude VTE events that occur in the post-discharge period. Administrative claims data also have higher false-positive and false-negative rates.

Dr. Ju and her colleagues designed the new study to overcome problems in previous studies.

They used ACS National Surgical Quality Improvement Program (NSQIP) data reported by 208 hospitals in 2009 and 2010. Patients in the study underwent 11 major surgical procedures associated with higher VTE rates. The data set included more than 30 patient factors for risk adjustment.

The analysis confirmed the previous study’s finding: VTE events in the first 30 days after surgery occurred more often at hospitals that performed more imaging studies.

Hospitals in the lowest quartile for imaging reported a VTE rate of 1.13%, whereas those in the highest quartile reported a rate of 1.91% (P<0.001). This trend persisted for VTE events that occurred only during the hospital stay.

Samuel R.G. Finlayson, MD, MPH, chair of surgery at the University of Utah, said the study raises important questions about how quality is measured.

“While comparing hospitals based on their rates of adverse outcomes appears, on the surface, to be a simple and fair approach, it turns out that quality in health care is much more complicated than that,” said Dr. Finlayson, who was the official discussant of the paper. “The problems that the authors have uncovered related to VTE rates strengthen the argument of those who believe that outcome-based, pay-for-performance programs are not ready for prime time.”

He noted that the research leaves some questions unanswered, such as why these hospitals conduct additional imaging tests.

VTEs occur too infrequently to be used as a quality measure, said Hiram C. Polk, MD, professor of surgery at the University of Louisville, in Kentucky, in a telephone interview.

He pointed to research from his department and the University Health Consortium, showing that deep vein thrombosis and pulmonary embolism occur rarely after major elective surgery, around or below 1%. Moreover, VTEs rates failed to improve after widespread implementation of the 2004 guidelines calling for more pharmacologic prophylaxis of VTE (Ann Surg 2011;253:215-220).

“Industry has funded studies that overestimate the true VTE rate, and that has put the focus on potential hazardous prophylaxis and less on the clinical significance of VTE… So we’re not focusing on the right thing, and these types of quality measures aren’t going improve quality in hospitals. This is the wrong target.”

This study suggested that certain hospital characteristics are associated with higher rates of VTE imaging. The characteristics included teaching intensity at a hospital and accreditations such as that of the Joint Commission. Other significant drivers included a hospital’s medicolegal risk environment according to CMS malpractice geographic practice cost index, and greater hospital market competition.

Hospitals that used VTE imaging more often were also those most likely to adhere to Surgical Care Improvement Project’s VTE prophylaxis guidelines. This may reflect an assumption that increased imaging intensity is associated with better quality, Dr. Ju said.

In this study, VTE imaging use rates were calculated from Medicare claims, and VTE prophylaxis adherence data were obtained from CMS’ Hospital Compare.

Hospitals showed considerable variation in VTE imaging, with a threefold increase from the lowest quartile at 5.3%, to 16% in the highest quartile.

The authors acknowledged that the study had limitations, because it only examined ACS NSQIP hospitals and inpatient VTE imaging use rates were for Medicare patients.

This year marked the 134th annual meeting of the American Surgical Association. Established in 1880, the American Surgical Association is the oldest surgical association in the United States.

Key Points

A previous study from the Surgical Outcomes and Quality Improvement Center at Northwestern University found that hospitals with higher rates of VTE perform more imaging studies than other hospitals (JAMA 2013;310:1482-1489).

In the current study, VTE events in the first 30 days after surgery occurred more often at hospitals that performed more imaging studies.

Hospitals that used VTE imaging more often were also those most likely to adhere to Surgical Care Improvement Project’s VTE prophylaxis guidelines. This may reflect an assumption that increased imaging intensity is associated with better quality.