Miriam E. Tucker

May 21, 2013

NATIONAL HARBOR, Maryland — The rate of venous thromboembolism (VTE) was nearly identical in hospitals with high rates of pharmacologic prophylaxis and those with low rates, a new study has found.

This is "counter to what we expected," Scott Flanders, MD, clinical professor of medicine and director of the hospitalist program at the University of Michigan, Ann Arbor, told Medscape Medical News. He explained that the clinical implication of this is that "we need to really refine our VTE risk-assessment strategies and focus on a much narrower population."

Dr. Flanders presented the results of the study, which was 1 of 3 "best of" oral abstracts, here at Hospital Medicine 2013.

The finding that VTE prophylaxis isn't having its intended impact in real-world hospital patients is particularly unsettling, given that the prevention of VTE — the most common preventable cause of in-hospital death — is considered a quality-improvement priority in hospitals.

Currently, pharmacologic VTE prophylaxis is recommended for all hospitalized medical patients deemed to be at high risk for clotting, and guidelines from many organizations suggest that nearly all hospitalized patients meet at least 1 of the at-risk criteria.

It shakes us up enough to re-ask the question.

Numerous efforts have been made to increase the rate of pharmacologic VTE prophylaxis in medical inpatients. For example, the Agency for Healthcare Research and Quality offers a VTE prophylaxis toolkit for hospitals, and the Centers for Medicare and Medicaid Services and the Joint Commission now require hospitals to report VTE prophylaxis rates as a performance measure.

These efforts have continued despite recently updated guidelines from the American College of Chest Physicians advising greater risk stratification (Chest. 2012;141[2 suppl]:1S).

One of the clinical implications of our study "is that the use of these broadly inclusive risk tools that categorize most medical patients as at-risk...probably isn't the best approach," Dr. Flanders told Medscape Medical News.

Session comoderator Kelley Skeff, MD, from Stanford University in California, pointed to the magnitude of these findings for hospitalists. "We now have a study that's asking us to question whether the goals we're using to measure ourselves by are the goals we should be using," he said.

"I look at this as a hypothesis-generating study," Dr. Skeff told Medscape Medical News. "I don't think it shows that all previous research is out, by any means, but it shakes us up enough to re-ask the question."

No Impact on VTE

Dr. Flanders and colleagues reviewed chart data on patients admitted to medical services for more than 2 days from January 2011 to September 2012 at 1 of the 35 hospitals that belong to the Michigan Hospital Medicine Safety Consortium.

Of the 20,794 patients eligible for study inclusion, 70% received pharmacologic prophylaxis on day 1 or 2 of hospitalization and 30% received no prophylaxis. In individual hospitals, rates of prophylaxis use ranged from 27% to 90%. For the highest-performing tertile of hospitals, the mean rate of prophylaxis use was 85.8%; for moderate performers, it was 72.6%; and for low performers, it was 55.5%.

The overall median length of stay was 4 days, and mean Caprini score was 5.6.

Overall, the VTE rate was 1.13%. Of the 236 VTE events, 15% occurred in the hospital; the rest occurred in the 90 days after discharge.

The rates of VTE did not differ by hospital performance. For the highest-performing tertile, the rate was 1.09%; for the moderate performers, it was 1.30%; and for the low performers, it was 0.99%.

The use of mechanical VTE prophylaxis was not part of the analysis, but among the tertiles of pharmacologic prophylaxis performance, there was no difference in mechanical prophylaxis use, Dr. Flanders noted.

There are several explanations for these counterintuitive findings, Dr. Flanders told Medscape Medical News. All point to the need for better risk assessment, he noted.

First, most previous studies that have shown the effectiveness of pharmacologic prophylaxis have looked at high-risk patients, such as those in the intensive care unit (ICU), surgical patients, and those with cancer, whereas this study excluded surgical and ICU patients.

"We were looking at a relatively lower-risk population, which is the largest group of patients we care for in the hospital," Dr. Flanders said.

In addition, many of the previous studies screened patients for VTE with ultrasound, which is not typically done in clinical practice. In our study, "we weren't looking for asymptomatic blood clots, only those leading to symptoms.... If you don't screen, the rate of VTE will be lower," he explained.

Also, hospital stays were shorter in this study than in many of the previous studies demonstrating the effectiveness of pharmacologic prophylaxis (4 days vs 7 or more days).

Our study looked at "a very representative population of patients we're seeing in hospitals across the state," he noted. "Many of the lessons from higher-risk patients who are in the hospital a long time are being translated to a lower-risk patient population.... But, as we've shown, the benefit of prophylaxis in the lower-risk group is not as significant as many would have thought."

However, improving risk stratification won't be easy. As this study suggests, the widely used Caprini score is too broadly inclusive and other tools are "complex, have multiple variables, assign points, and are generally unwieldy difficult to use," Dr. Flanders explained.

"It's much easier to simply ask if the patient is at risk for bleeding and, if the answer is no, give them prophylaxis," he told Medscape Medical News.

The Michigan consortium is currently developing a simpler tool using their data and is testing some of the existing tools to see how well they perform in their patient population.

Dr. Skeff urged caution. "I would say there are still data to support the current recommendations.... Our goal is to try to prevent disastrous consequences so our tendency is to be as conservative as possible."

However, he told Medscape Medical News that these data highlight the need for continuous re-examination of all benchmarks. "The wonderful thing about research in medicine is that it continues to re-examine itself and say, let's be sure about what we're doing."

Dr. Flanders has disclosed no relevant financial relationships. One of his coauthors reports receiving research funding from and/or being a speaker for Boehringer-Ingelheim, Bristol Myers Squibb, Bayer, Janssen, Johnson & Johnson, GlaxoSmithKline, Daiichi Sankyo, Elsai, and Iverson Genetics Diagnostics. Dr. Skeff reports being a consultant for Wolters Kluwer Health.

Hospital Medicine 2013: Society of Hospital Medicine (SHM) Annual Meeting. Presented May 18, 2013.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.