Transfusion linked to higher mortality, againBMJ 2012; 345 doi: http://dx.doi.org/10.1136/bmj.e8677 (Published 28 December 2012) Cite this as: BMJ 2012;345:e8677
Experts disagree about the safety of blood transfusion for adults with acute coronary syndromes. This is partly because observational studies often report a link between transfusion (or more liberal transfusion policies) and worse outcomes, including death. The latest study, a meta-analysis of 10 studies, reported one extra death for every eight patients transfused after a myocardial infarction (95% CI 6 to 17). In addition, harms were still significant after adjustments for baseline haemoglobin concentration, lowest haemoglobin concentration, and change in haemoglobin concentration during hospital stay. Transfusion was also associated with further myocardial infarctions, and the authors urge providers to rethink policies that encourage routine or liberal use of blood transfusions in these patients.
Nine of the studies were observational, and the only trial was small and underpowered. Transfused patients may have been older and sicker than others, and the authors didn’t have the fine detail they needed to rule out confounding in their observational analyses. Bigger and better trials are long overdue, says a linked comment (doi:10.1001/jamainternmed.2013.2855). The new findings aren’t secure enough to change practice, however, and we urgently need reliable unbiased evidence to guide treatment decisions for people with myocardial infarction and other acute coronary syndromes.
Profound anaemia is life threatening, and a doctor’s instinct is to replace lost red blood cells and restore oxygen carrying capacity. Only trials can tell us the safest and most effective threshold for transfusion, says the comment. The question is not whether to transfuse anaemic patients with myocardial infarction, but when.
Cite this as: BMJ 2012;345:e8677