Simon Hyttel‐Sorensen, Gorm Greisen, Bodil Als‐Nielsen, Christian Gluud

Cerebral injury and long-term neurodevelopmental impairment is common in extremely preterm infants. Cerebral near-infrared spectroscopy (NIRS) enables continuous estimation of cerebral oxygenation. This diagnostic method coupled with appropriate interventions if NIRS is out of normal range (that is cerebral oxygenation within the 55% to 85% range) may offer benefits without causing more harms. Therefore, NIRS coupled with appropriate responses to abnormal findings on NIRS needs assessment in a systematic review of randomised clinical trials and quasi-randomised studies. To evaluate the benefits and harms of interventions that attempt to alter cerebral oxygenation guided by cerebral NIRS monitoring in order to prevent cerebral injury, improve neurological outcome, and increase survival in preterm infants born more than 8 weeks preterm. We used the standard search strategy of the Cochrane Neonatal Review Group to search the Cochrane Central Register of Controlled Trials (CENTRAL 2016, Issue 8), MEDLINE via PubMed (1966 to 10 September 2016), Embase (1980 to 10 September 2016), and CINAHL (1982 to 10 September 2016). We also searched clinical trial databases, conference proceedings, and the reference lists of retrieved articles for randomised clinical trials and quasi-randomised studies. Randomised clinical trials and quasi-randomised clinical studies comparing continuous cerebral NIRS monitoring for at least 24 hours versus blinded NIRS or versus no NIRS monitoring. Two review authors independently selected, assessed the quality of, and extracted data from the included trials and studies. If necessary, we contacted authors for further information. We conducted assessments of risks of bias; risks of design errors; and controlled the risks of random errors with Trial Sequential Analysis. We assessed the quality of the evidence with GRADE. One randomised clinical trial met inclusion criteria, including infants born more than 12 weeks preterm. The trial employed adequate methodologies and was assessed at low risk of bias. One hundred and sixty-six infants were randomised to start continuous cerebral NIRS monitoring less than 3 hours after birth until 72 hours after birth plus appropriate interventions if NIRS was out of normal range according to a guideline versus conventional monitoring with blinded NIRS. There was no effect of NIRS plus guideline of mortality until term-equivalent age (RR 0.50, 95% CI 0.29 to 1.00; one trial; 166 participants). There were no effects of NIRS plus guideline on intraventricular haemorrhages: all grades (RR 0.93, 95% CI 0.65 to 1.34; one trial; 166 participants); grade III/IV (RR 0.57, 95% CI 0.25 to 1.31; one trial; 166 participants); and cystic periventricular leukomalacia (which did not occur in either group). Likewise, there was no effect of NIRS plus guideline on the occurrence of a patent ductus arteriosus (RR 1.96, 95% CI 0.94 to 4.08; one trial; 166 participants); chronic lung disease (RR 1.27, 95% CI 0.94 to 1.50; one trial; 166 participants); necrotising enterocolitis (RR 0.83, 95% CI 0.33 to 1.94; one trial; 166 participants); and retinopathy of prematurity (RR 1.64, 95% CI 0.75 to 3.00; one trial; 166 participants). There were no serious adverse events in any of the intervention groups. NIRS plus guideline caused more skin marks from the NIRS sensor in the control group than in the experimental group (unadjusted RR 0.31, 95% CI 0.10 to 0.92; one trial; 166 participants). There are no data regarding neurodevelopmental outcome, renal impairment or air leaks.The quality of evidence for all comparisons discussed above was assessed as very low apart from all-cause mortality and adverse events: these were assessed as low and moderate, respectively. The validity of all comparisons is hampered by a small sample of randomised infants, risk of bias due to lack of blinding, and indirectness of outcomes. The only eligible randomised clinical trial did not demonstrate any consistent effects of NIRS plus a guideline on the assessed clinical outcomes. The trial was, however, only powered to detect difference in cerebral oxygenation, not morbidities or mortality. Our systematic review did not reach sufficient power to prove or disprove effects on clinical outcomes. Further randomised clinical trials with low risks of bias and low risks of random errors are needed.