In the correction (BMJ 2006;333:30, 1 Jul) to the article “Randomised, controlled trial of alternating pressure mattresses compared with alternating pressure overlays for the prevention of pressure ulcers: PRESSURE (pressure relieving support surfaces) trial” (BMJ 2006;3321413-5, 17 Jun), we incorrectly referred to haemoglobin levels rather than odds ratios. We should have said: “In table 4 of the full version on bmj.com (table 2 of the abridged version), the odds ratio for haemoglobin levels on admission or preoperatively should be 0.89 (0.82 to 0.97), and the corresponding P value should be 0.01.”