TRAUMA/BRIEF COMMENTARY

Validation of the DimeJames Li, MD

From Miles Memorial Hospital, Damariscotta, ME.

0196-0644/$-see front matterCopyright © 2012 by the American College of Emergency Physicians.http://dx.doi.org/10.1016/j.annemergmed.2012.04.017

nif(nhiidniao

Oiitoghfo

dsoff

A podcast for this article is available at www.annemergmed.com.

SEE RELATED ARTICLE, P. 460.

[Ann Emerg Med. 2012;59:469-470.]

If you haven’t read the related commentary onanticoagulated head-injured patients, start there.1 This is anecessary addendum to that piece. Since that commentarywas put into the Annals article queue, a large high-qualitystudy of head-injured patients2 completed the review process.It adds substantial evidence on the risk of both immediateand delayed intracranial hemorrhage for anticoagulated head-injured patients. Indeed, the profusion of data added in thisissue of Annals provides some conclusiveness to the questionof how we should be managing such patients, particularlythose with normal initial computed tomography (CT) scanresults. To top it off, we are provided with the first rigorousevidence that we should scan head-injured patients who arereceiving clopidogrel.

The work by Nishijima et al2 roughly doubles the numberof previously studied CT results for head-injured patientsreceiving warfarin. It quadruples the number of such patientsinvestigated for delayed hemorrhages. Furthermore, thestudy was conducted in a mix of community and academiccenters and was thoughtfully designed so its conclusionscould apply to actual practice. When analyzed in the contextof previous reports (Table), it slightly reduces the aggregatedrate of intracranial hemorrhage found by initial CT scan andfurther tightens the confidence interval for the sum ofprevious reports. With the addition of these data, the

Table. Summary of intracranial hemorrhage risk in head-injured

Drug and ICH Type Source ICH/

Warfarin and immediate ICH* Table 11 158/Nishijima2 37/Subtotal 195/

Warfarin and delayed ICH†

Table 21 9/Nishijima2 6/Subtotal 15/

Clopidogrel and immediate ICH Nishijima2 33/Clopidogrel and delayed ICH† Nishijima2 3/

ICH, Intracranial hemorrhage; NNT, number needed to treat; CI, confidence interv*Analysis limited by variation in study settings and designs.

†Nishijima’s numbers reflect 4 patients lost to follow-up and 1 out-of-hospital death.

Volume , . : June

umber of anticoagulated patients needed to scan to detectntracranial hemorrhage is 8, which validates observationsrom the National Emergency X-Ray Utilization StudyNEXUS)-II group3 that found an identical rate (62/493;umber needed to treat 8) in patients clinically suspected ofaving coagulopathy. (The NEXUS-II data are not included

n the Table because coagulopathy was broadly defined andncluded patients receiving aspirin or other nonwarfarinrugs.) According to these considerations, this numbereeded to treat likely reflects the actual rate of traumatic

ntracranial hemorrhage if the entire population ofnticoagulated head-injured patients could be studied atnce.

But the real news is found in the other arms of the study.ne measured the risk of intracranial hemorrhage in head-

njured patients receiving clopidogrel. In examining the datan the context of this and previous reports, the risk forraumatic intracranial hemorrhage is at least as great as thatf head-injured patients receiving warfarin and may bereater (Table). These data should compel us to scan ouread-injured clopidogrel patients. They also provide impetusor a similar trial of head-injured patients receiving thatther ubiquitous antiplatelet agent, aspirin.

In addition, Nishijima’s group investigated the rate ofelayed intracranial hemorrhage after a normal initial CTcan result. Instead of admitting and rescanning their samplef anticoagulated patients, the authors chose telephoneollow-up, reflecting what we decided to do at my institutionor anticoagulated head-injured patients with normal initial

nts receiving warfarin or clopidogrel.

le Percentage NNT 95% CI of NNT

18 65 20

0 12 8 7–94 251 1152 61 37–100

12 8 6–121 81 28–238

patie

Samp

8967241,62224687911276243

al.

Annals of Emergency Medicine 469

Fwdbhs

S

Ftrao

Ar

R1

2

3

Validation of the Dime Li

CT scan results. Most likely, they chose this method for thesame reasons we did. The cost of the alternative is simplyuntenable.

In an understated testament to a tremendous effort,Nishijima’s group was able to obtain 2-week telephonefollow-up for 926 of 930 patients with head trauma whowere receiving warfarin or clopidogrel and were dischargedafter a normal CT scan result. Planned or not, the group’ssuccessful demonstration of an inexpensive telephoneprogram provides hope to institutions that are searching for areliable follow-up system for their own patients.

Despite the size of their sample and that they did not limit it tohead injuries that were only mild, Nishijima’s group found fewcases attributable to delayed hemorrhage. In my earliercommentary, I surmised that for every 25 consecutiveanticoagulated head-injured patients undergoing repeated CTscans, 1 would be found to have a delayed hemorrhage. Afteradding the data from Nishijima’s group, the number needed torescan increased to 61. In contrast to European guidelines, thesedata support the conclusion that admission and repeated scanningare unnecessary for the anticoagulated head-injured patient with anormal initial CT scan result. Instead, given the considerations ofrisk and cost, telephone follow-up with selective rescanning is areasonable alternative.

So here’s a synthesis of both articles and commentaries. Scan

anyone with head trauma who is receiving warfarin or clopidogrel.

470 Annals of Emergency Medicine

or patients with normal initial CT scan results, ensure follow-upith at least a telephone call and rescan those with symptoms ofelayed intracranial hemorrhage because such hemorrhage is rareut real. Finally, at least in the United States, the practice of routineospitalization for anticoagulated patients with normal initial CTcan results is too costly to be worthwhile.

upervising editor: Michael L. Callaham, MD

unding and support: By Annals policy, all authors are requiredo disclose any and all commercial, financial, and otherelationships in any way related to the subject of this articles per ICMJE conflict of interest guidelines (see www.icmje.rg). The author has stated that no such relationships exist.

ddress for correspondence: James Li, MD, E-mail jamesli@emotemedicine.org.

EFERENCES. Li J. Admit all anticoagulated head-injured patients? a million

dollars versus your dime. You make the call. Ann Emerg Med.2012;59:457-459.

. Nishijima DK, Offerman SR, Ballard DW, et al. Immediate anddelayed traumatic intracranial hemorrhage in patients with headtrauma and pre-injury warfarin or clopidogrel use. Ann Emerg Med.2012;59:460-468.

. Medzon R, Bracken M, Rathlev NK, et al. Clinically suspectedcoagulopathy in blunt head trauma. J Emerg Med. 2010;39:399-

405.

Did you know?

You can save your online searches and get the results by e-mail.

Visit www.annemergmed.com today to see what else is new online!

Volume , . : June