Fifty-four studies (n=1,845) were included in the review: five randomised controlled trials (RCTs) (n=178); three were retrospective studies (n=439); one pre-post design (n=8); and 45 case series (n=1,220). Only two studies had planned post-hospital follow up (at three months).
Intraoperative (38 studies, four RCTs): Only three studies reported the relationship between near-infrared spectroscopy and clinical outcomes. One retrospective study (n=34) found that patients who died after a single ventricle first-stage palliation had lower regional oxygen saturation at the end of the operation (p<0.01), but there was no correlation to clinical neurologic abnormalities. Two case series found associations between decreased regional oxygen saturation and abnormalities in postoperative magnetic resonance imaging (MRI) scans; one of these showed a significant association between low regional oxygen saturation and new or worsening lesions (n=22, p=0.029).
Intensive care (11 studies, one crossover RCT): None of the studies reported the relationship between near-infrared spectroscopy and clinical outcomes or neurological follow-up. The common findings in the studies which examined correlations with other outcomes related to oxygenation were that there was a large amount of variability both between and within individuals.
Cardiac catheterisation laboratory (seven studies): None of the studies reported the relationship between near-infrared spectroscopy and clinical outcomes or neurological follow-up. The common findings in the studies that examined correlations with other outcomes related to oxygenation were that there was a large amount of variability both between and within individuals.