EM Journal Club Eastern Virginia Medical School CITATION: Shakur H, et al. Effects of tranexamic acid on death, vascular occlusive events, and blood transfusion in trauma patients with significant haemorrhage (CRASH-2): a randomized, placebo-controlled trial. Lancet. 2010 Jul 3; 376: 23-32.
I. WHAT IS BEING STUDIED?
Does tranexanic acid vs NaCl control in trauma patients with shock or high risk of hemorrhage decrease death without increased adverse outcomes such as vasooclusive events and decrease blood transfusion requirements
Multi-centered international trial 274 hospitals in 40 countrie ‘andomized double blind controlled’ Tranexanic acid vs NaCl control randomized, 1 mg over 10 min over 8 hrs, 274 hosp 40 countries, physician and pts blinded.
Adults trauma pts. At risk for significant hemorrhage (tachycardia >110 SBP < 90) within 8 hrs from injury, responsible doctor had to be ‘uncertain’ regarding the use an antifibrinolytic agent in bleeding patient (very unclear how this determination was made)
Children, those with contraindication to TXA, or doctor who is ‘certain’ regarding appropriate indication for using tranexamic agent in specific patient
Primary -Death at 28 days -> causes –
bleeding, vascular occlusion, multigrain failure, Secondary - vascular occlusive events (MI. CVA, PE, DVT) surgical intervention, blood transfusions, functional status at 28 days (dependency measured by Modified Oxford Scale) Weakness – no stratification based on injury severity score “Outcomes were recorded of they occurred while the patient was still in the hospital for up to 28 days after randomization”
II. Are the results of the study valid?
Unclear– patients apparently underwent a
physicians before being placed in a randomization sequence. Th\reating physicians first determined if patients had definitive indication for TXA or a contradiction. All of these subjects were excluded. Those with indeterminate indications for TXA were randomized. No guidance regarding standardized criteria regarding indications for TXA across study sites. Hugh potential for selection bias.
2. Were all patients who entered the trial
Unclear. very little information is provided
properly accounted for and attributed at its
regarding. It is unclear if patients that were
discharged prior to 28 days were followed-up as out patients or if their f/u concluded at the time of D/C. No description regarding the specifics of put-patient f/u were included
Unclear. As stated above, authors claim
99.6 follow-up but fail to describe exactly how they followed there patients and if they followed them post discharge and if so how did they do it?
4. Were patients, health workers and study
inclusion (for randomization) by treating
physicians. TXA and NS were undistinguishable.
5. Were study groups similar at the start of
Uncertain. Appears that patients were well
however some interventions such as amount of fluid resuscitation between groups was not described.
III. What were the results?
(difference between treatment and control
ARR 14.5% vs 16% = 1.5% or RR 0.91 (CI 95% .85-97) Risk Reduction Death from bleeding TXA vs NaCl ARR 4.9% vs. 5.7%= 0.8% RR 0.85 (CI 95% 0.76-0.96) Functional Status: Normal at 28 days TXA vs. NaCL ARR 14.7% vs. 13.3% = 1.4% RR 1.10 (CI 95% 1.04-1.19) 149 pts No Statistically sig differences in: Blood transfusions; Surgery;Vascular occlusive events;
2. What was the estimated treatment effect
Number needed to treat 1/ARR = 1/0.015 =66
IV. Will the results help me in caring for
1. Were all clinically important outcomes
Yes – authors appear to have considered
most important clinical outcone measures
No mention of following seizures as a complication which is commonly described side effect of TXA use. No cost analysis was discussed.
No statistically significant differences in
TXA though seizure frequency was not reported which could have been an important measure considering its description in the literature.
Additional Comments: TXA appears to demonstrate some improvement in death (ARR 1.5%) and bleeding (ARR 0.8%) though marginally so. Statistical significance and narrow CI’s are likely a reflection of the sample sizes more than anything else. The authors do not report an NNT, which is curious and unclear as to why it was not included. The NNT is 66, meaning one would have to treat 66 bleeding trauma patients with TXA in order to achieve positive effect in one. In lieu of either a cost analysis or a reporting on the most commonly described side effect of TXA (seizures) it may be difficult to justify routing use of TXA in trauma patients. The lack of an increase in Vascular Occlusive Events with TXA is favorable as these agents have also been associated with stroke and MI as complications. The issues of selection bias persists as the treating physicians needed to decide if the patients were ‘eligible for TXA’ or had a contraindication either of which have excluded subject from enrollment. The practice of using TXA might be more common in some of the countries that participated however no discussion regarding how those subjects were identified was offered.
Attracting bumblebees and making artificial bumblebee nest sites in the garden Seven species you might meet in the garden Workers hatched early in the year are much smaller than later siblings. White-tailed Red-tailed Bombus terrestris Buff-tailed bumblebee. Bombus lapidarius Red-tailed bumblebee. Bombus lapidarius Bombus terrestris Bombus horortum '