Evaluating the quality of trials of hypothermic circulatory arrest aortic surgery

The quality of level 1 evidence in reports on deep hypothermic circulatory arrest was assessed, and the confounding factors in surgical management and study design that can prevent meta-analysis formulation were determined. A systematic search of the literature was conducted using categorized nomenclature for randomized controlled trials in adult patients undergoing deep hypothermic circulatory arrest in the last 40 years. Twelve randomized controlled trials (2.3%) were found among 504 publications on deep hypothermic circulatory arrest listed on Medline from 1960; only 4 of them related to adults. One adequately powered study demonstrated reduced blood loss in deep hypothermic circulatory arrest using aprotinin. Three studies comparing retrograde and antegrade perfusion were underpowered. The median CONSORT score was 14 (range, 13-15). There were no consistent measures of similar outcomes (neuropsychometric, neurocognitive). No explanation was provided for the difference in reported ranges of neurological deficits in nonrandomized (5%-70%) and randomized (3%-9%) studies. Existing studies of deep hypothermic circulatory arrest are insufficient and inconsistent in the outcome measured, which explains the lack of a meta-analysis. Neurological injury remains high, and an appropriately powered study of interventions that can optimize cerebral perfusion is necessary.