Effects of Tranexamic Acid on Death, Vascular Occlusive Events, and Blood Transfusion in Trauma Patients with Significant Haemorrhage (CRASH-2)
2010 · paper · Clinical trials
A randomised, placebo-controlled trial of tranexamic acid given early to 20,211 adult trauma patients with or at risk of significant bleeding. It reports lower all-cause death (14.5% vs 16.0%) and lower death from bleeding with the drug.
2 findings
Substance Figure 3, systolic blood pressure 76–89, tranexamic acid: "280/1609 (17·5%)". 280/1609 is 17·4%.
- p. 7 Figure 3, systolic blood pressure 76–89, tranexamic acid column
280/1609 (17·5%)
Judge's note: The page 7 crop shows '280/1609 (17·5%)'. 280/1609 = 17·40%, which cannot round to 17·5%. The adjacent placebo cell (313/1689 = 18·5%) and the RR 0·94 both fit 280/1609, so the printed percentage is the value at fault.
Second check: another model agreed, code check
Report a problem with this finding
Reference or label Reference 20 (MRC CRASH Trial Collaborators, prognostic models) is given as "BMJ 2008; 12: 12.", an implausible volume and page.
- p. 10 Reference 20
BMJ 2008; 12: 12.
Only the authors can say what is right here.
Judge's note: Rests on outside knowledge: that the BMJ 2008 volumes were in the 330s and that this article is BMJ 2008; 336: 425–29. The page 10 crop confirms the printed text 'BMJ 2008; 12: 12.' Within the reference list itself, reference 17 is 'BMJ 1995; 310', which makes volume 12 for 2008 implausible.
Second check: another model agreed
Report a problem with this finding
3 minor (typos and the like; most checked by the judge only)
Wording Introduction: "Tranexamic acid reduced the need for blood transfusion by a third (relative risk [RR] 0·61, 95% CI 0·54–0·70)": an RR of 0·61 is a reduction of about 39%, not one third.
- p. 1 Introduction, paragraph 3
blood transfusion by a third (relative risk [RR] 0·61, 95% CI
- p. 1 Introduction, paragraph 3
reduced the need for blood transfusion by a third (relative risk [RR] 0·61
Judge's note: Checked the crop of page 1, right column. The sentence prints RR 0·61 (0·54–0·70) and calls it a reduction 'by a third'. The point estimate gives 1-0·61 = 0·39. A one-third reduction (RR about 0·67) lies inside the CI, so the wording is a loose approximation, not a wrong number. This is a minor wording point.
Second check: another model agreed
Report a problem with this finding
Typo Table 1, systolic blood pressure 76–89, tranexamic acid column: "1615 (16.0%)" uses a full-stop decimal where the rest of the table uses the mid-dot (e.g. "1697 (16·8%)").
- p. 4 Table 1, systolic blood pressure 76–89, tranexamic acid column
1615 (16.0%)
Judge's note: The page 4 crop shows a full stop in '16.0%', while the neighbouring cells use mid-dots. This is a cosmetic typesetting inconsistency.
Second check: code check
Report a problem with this finding
Typo Reference 2 lists the author as "Zirkled LG"; the surname appears to be Zirkle.
- p. 10 Reference list, entry 2
Coughlin R, Zirkled LG
- p. 10 Reference 2
Zirkled LG
Only the authors can say what is right here.
Judge's note: Rests on outside knowledge: that the co-author of 'Injuries: the neglected burden in developing countries' (Bull World Health Organ 2009) is Lewis G Zirkle. The page 10 crop shows 'Coughlin R, Zirkled LG', so the spelling is in the PDF and is not an extraction artefact.
Second check: none (judge only)