Data From 1.23 Million Patients Confirms Warfarin Increases Mortality in Trauma Patients

Reed Miller

October 20, 2009

October 20, 2009 (Chicago, Illinois) — Surgeons' suspicion that warfarin (Coumadin, Bristol-Myers Squibb) significantly increases the mortality risk of traumatic injuries in both elderly and younger patients has been confirmed by a large retrospective cohort study [1].

Dr Lesly Dossett (Vanderbilt University, Nashville, TN) and colleagues from Vanderbilt and the University of Texas Health Science Center, in Houston, reviewed about two million records taken between 2002 and 2007 at 770 US centers in the National Trauma Databank maintained by the American College of Surgeons. The researchers found 1.23 million records from 402 centers, including about 36 000 from warfarin users, with sufficient information on comorbidities to analyze the link between mortality rates and warfarin therapy.

Dossett presented the findings October 15 at the American College of Surgeons 2009 Clinical Congress.

The study confirmed the researchers' anecdotal impression that the prevalence of patients on warfarin among trauma patients is increasing, especially among older patients. In 2002, 2.3% of all patients in the database were on warfarin vs 4% in 2007. The fraction of trauma patients over 65 taking warfarin grew from 7.3% to 12.8% during that period.

Warfarin Adds More Risk in Younger Patients

Adjusted for gender, age, race, injury severity, treatment center, and other comorbidities, the mortality risk ratio for warfarin vs no warfarin was 1.72 overall and 1.38 for patients over 65.

Dossett highlighted the impact of warfarin on traumatic-brain-injury (TBI) patients. The unadjusted mortality rates for TBI patients were 64% for patients on warfarin and 43% for patients not taking warfarin. The adjusted risk ratio associated with warfarin in TBI patients was 1.26 overall and 1.46 in patients under 65.

Warfarin therapy carries more risk for patients under 65. "It seems counterintuitive, but if you're older and have an accident such as a head injury, then you have a very high risk of dying regardless of whether or not you're on warfarin. If you're younger and don't have other [comorbidities], you generally can survive trauma, so that's where warfarin can have even a bigger effect," Dossett told heartwire .

The overall unadjusted mortality rate for traumatic injuries in the study was 9.8% for patients on warfarin and 4.8% for patients not on warfarin. The overall mortality risk ratio associated with warfarin was 1.51 for patients under 65 and 1.41 for older patients.

Dossett said that the overall unadjusted risk ratio is an important finding. "As a trauma surgeon, that's a useful number to know, because instead of having to calculate in my head and combine all of these risk factors, I can know that if they're on Coumadin, that's in and of itself a marker of a number of risk factors that doubles somebody's mortality.

"For people who prescribe warfarin, this is important information to add to the overall risk profile of that drug. [For example,] we see people who had a [deep vein thrombosis] three or four years ago, and there's probably no evidence to guide therapy for that long of a period," Dossett said.

What's the Best Way to Reverse Anticoagulation?

There has yet to be a study rigorously comparing the most common therapies for reversing the effects of anticoagulants in trauma patients, such as vitamin K, factor VII, or fresh-frozen plasma infusions, Dossett said. There is some literature showing that patients have better outcomes if the treating center follows a standard protocol for managing coagulopathy, but there is little information available to show which protocols work best, she said. So this research group's next major project will be to survey trauma centers and providers to find out how they are currently managing patients on anticoagulants, Dossett said.

"I think most centers probably don't have protocols for correcting coagulopathy. It's usually just up to the individual surgeon on call," she said. "We're planning on surveying trauma centers and providers to figure out, number one, do they have a protocol or not, and number two, what are the various methods that they use and what are the biases about those methods?"