Erratum to: Intensive Care Med (2008) 34:1935–1947 DOI 10.1007/s00134-008-1266-6

The article unfortunately contained errors:

Table 1, last line: the planned sample size for the RESOLVE study was 600, not 6000 as stated. 477/600 = 79.5% of planned recruitment, not 8% as currently stated.

Evidence of harm section on the same page as Table 1, second sentence: the ENHANCE bleeding rate should read 6.5%, not 5.5% as currently stated. The correct bleeding rate of 6.5% is reported in Table 2.

In Table 3 and in Appendix 2 the correct units for AST or ALT are IU/L not IU/dl, and for bilirubin the units should be mg/dL rather than g/dL.