a systematic review and meta-analysis of clinical trials of thyroid hormone administration to brain dead potential organ donors*

Objectives: To review all published clinical studies of thyroid hormone administration to brain-dead potential organ donors. Methods: A search of PubMed using multiple search terms re- trieved 401 publications including 35 original reports describing administration of thyroid hormone to brain-dead potential organ donors. Detailed review of the 35 original reports led to identifi- cation of two additional publications not retrieved in the original search. The 37 original publications reported findings from 16 separate case series or retrospective audits and seven random- ized controlled trials, four of which were placebo-controlled. Meta- analysis was restricted to the four placebo-controlled randomized controlled trials. Results: Whereas all case series and retrospective audits re- ported a beneficial effect of thyroid hormone administration, all seven randomized controlled trials reported no benefit of thyroid hormone administration either alone or in combination with other hormonal therapies. In four placebo-controlled trials including 209 donors, administration of thyroid hormone (n 108) com- pared with placebo (n 5 101) had no significant effect on donor cardiac index (pooled mean difference, 0.15 L/min/m 2 ; 95% con- fidence interval -0.18 to 0.48). The major limitation of the case series and retrospective audits was the lack of consideration of uncontrolled variables that confound interpretation of the results. A limitation of the randomized controlled trials was that the pro- portion of donors who were hemodynamically unstable or mar - ginal in other ways was too small to exclude a benefit of thyroid hormone in this subgroup. Conclusions: The findings of this systematic review do not support a role for routine administration of thyroid hormone in the brain-dead potential organ donor. Existing recommen- dations regarding the use of thyroid hormone in marginal donors are based on low-level evidence. (Crit Care Med 2012; 40: 1635-1644)