Is Venous Thromboembolism Prophylaxis Worth It?

Nancy A. Melville

April 03, 2014

LAS VEGAS — Amid stepped-up efforts to prevent hospital-associated venous thromboembolism, 2 new studies raise questions about the appropriate use and effectiveness of pharmacologic prophylaxis.

Given the low rates of venous thromboembolism, the usefulness of prophylaxis, which has its own risks, should be questioned. "It appears that we are overprescribing prophylaxis; this is particularly true in low-risk patients," said Paul Grant, MD, from the University of Michigan Health System in Ann Arbor.

"If this is the case, then we are exposing some patients to an increased risk of bleeding," he told Medscape Medical News.

Both studies, presented here at Hospital Medicine 2014, evaluated data on patients from 40 hospitals in the Michigan Hospital Medicine Safety Consortium, a statewide quality initiative supported by Blue Cross Blue Shield of Michigan.

Screening for Pharmacologic Prophylaxis

The first study evaluated the use of pharmacologic prophylaxis, on hospital admission, in the 2 years after the implementation of quality-improvement measures to prevent venous thromboembolism.

Results were presented by Scott S. Kaatz, DO, from the Hurley Medical Center in Flint, Michigan.

All 40,877 patients had a Caprini risk score above 2, indicating moderate to high risk for venous thromboembolism.

For at-risk patients without contraindications, pharmacologic prophylaxis increased from 59.2% in 2011 to 77.9% in 2013. However, there was also an increase in prophylaxis for patients with contraindications to the drugs, from 8.7% to 25.8% (P < .001 for trend).

During the same period, major bleeding events — a key concern with pharmacologic prophylaxis — increased from 0.2% to 0.6% in all patients receiving pharmacologic prophylaxis (P = .052 for trend).

"The good news is that the statewide quality-improvement process increased rates of pharmaceutical prophylaxis, which is a good thing," said Dr. Kaatz.

"The bad news is that we had about a 3-fold increase in the use of pharmacologic prophylaxis in patients with contraindications," he noted. "The consequence is that bleeding approximately tripled, although that did not reach statistical significance."

Among the conditions considered to be contraindications are high-risk metastasis to the brain and/or the presence of an intracranial monitoring device, severe head or spinal cord trauma within the previous 24 hours, gastrointestinal or genitourinary hemorrhage in the previous 6 months, intracranial hemorrhage in the previous 12 months, other hemorrhage within the previous 6 months, or a platelet count below 50,000/mm³.

Patients younger than 18 years, pregnant women, patients admitted for surgery or palliative care, and patients admitted directly to the intensive care unit (ICU) were excluded from the analysis, as were patients with a low risk for VTE and those who were receiving therapeutic doses of an anticoagulant.

Even when prevention of adverse outcomes is the goal, appropriate patients and processes must be identified with caution, said Dr. Kaatz.

"You have to be careful with quality-improvement processes because you might wind up having outcomes you didn't intend," he explained. "We need more research to identify which patients really are at risk."

Effectiveness of Pharmacologic Prophylaxis

The second study evaluated the effectiveness of the Caprini risk assessment model, which has been validated in some surgical patient populations, for identifying risk for venous thromboembolism in a medical population.

Of the 52,989 patients evaluated by Dr. Grant and his colleagues, 299 (0.56%) had a venous thromboembolism event in the 90 days after discharge.

The odds of an event was more than 3 times greater for patients with a Caprini score of at least 5, indicating high risk, than for those with a Caprini score of 0 to 4, indicating low risk (odds ratio, 3.3; P = .039).

The odds of having an event was not significantly different in patients with low-risk and moderate-risk scores.

Importantly, there was no significant difference in the rates of venous thromboembolism for patients receiving and not receiving pharmacologic prophylaxis (0.72% vs 0.86%; P = .26).

In addition, there are downsides to prophylaxis, including "patient discomfort from injections, cost, nursing time, and the possibility of drug reactions, including heparin-induced thrombocytopenia," said Dr. Grant.

"We are one of the first groups to look at this specific patient population of hospitalized medical patients — not ICU patients and not surgical patients — with a very large sample size and in great detail, and have noted that the venous thromboembolism event rate is very small," he told Medscape Medical News.

As a next step, Dr. Grant and his team will look at subgroups in the nonsurgical population to try to identify the specific risk factors that place patients at high risk — where prophylaxis shows a consistent benefit.

Understanding the Role of Prophylaxis

These findings underscore the need to better understand the role of venous thromboembolism prophylaxis in a low-risk population, said session comoderator Eduard E. Vasilevskis, MD, MPH, from the Center for Health Services Research at Vanderbilt University in Memphis, Tennessee.

"With a low prevalence of clinically significant venous thromboembolism events, discriminating between those with low and moderate risk was not possible, even with a robust risk prediction system that used more than 30 nurse-abstracted risk factors," Dr. Vasilevskis explained.

"The fact that venous thromboembolism events did not vary according to receipt of prophylaxis, even in high-risk groups," raises questions, he told Medscape Medical News.

The findings add evidence to previous research from the team, which was presented at the Hospital Medicine 2013 and reported at that time by Medscape Medical News.

That study "was the first to show, in this population, the lack of association between prophylaxis rates and venous thromboembolism rates," Dr. Vasilevskis noted.

"This was quite surprising at the time. This study confirms that even among those with a high-risk Caprini score, differences in venous thromboembolism rates do not exist," he said.

Dr. Kaatz, Dr. Grant, and Dr. Vasilevskis have disclosed no relevant financial relationships.

Hospital Medicine 2014. Presented March 26, 2014.


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.