LO041: Predicting the return of spontaneous circulation using near-infrared spectroscopy monitoring: a systematic review and meta-analysis

Introduction: Tissue oximetry using near-infrared spectroscopy (NIRS) is a non-invasive monitor of cerebral oxygenation. This new technology has been used during cardiac arrest because of its ability to give measures in low blood flow situations. The aim of this systematic review was to assess the evidence regarding the association between NIRS values and resuscitation outcomes in patients undergoing cardiopulmonary resuscitation. We hypothesized that higher NIRS values would be associated with better outcomes and that the strength of that association would differ depending on the timing of the NIRS measurements. Methods: This review was registered (Prospero CRD42015017380) and is reported as per the PRISMA guidelines. Medline, Embase and CENTRAL were searched from their inception to September 18th, 2015 using a specifically designed search strategy. Grey literature was also searched using Web of Science and Google Scholar. NIRS manufacturers and authors of included citations were contacted to inquire on unpublished results. Finally, the references of all retained articles were reviewed in search of additional relevant studies. Studies reporting NIRS monitoring in adults during cardiac arrest were eligible for inclusion. Case reports and case series of fewer than five patients were automatically excluded. Two reviewers assessed the quality of included articles and extracted the data. Results: Out of 3275 unique citations, 19 non-randomized observational studies (15 articles and four conference abstracts) were included in this review, for a total of 2436 patients. Six studies were evaluated at low risk of bias, nine at intermediate risk and four at high risk. We found a stronger association between the return of spontaneous circulation (ROSC) and the highest NIRS value measured during resuscitation (standard mean deviation (SMD) 3.46 (95%CI 2.31-4.62)) than between ROSC and the mean NIRS measures (SMD 1.33 (95%CI 0.92-1.74)) which was superior to the one between ROSC and initial measures (SMD 0.45 (95%CI 0.02-0.88)). Conclusion: Patients with good outcomes have significantly higher NIRS value during resuscitation than their counterparts. The association between ROSC and NIRS measurements was influenced the timing of measurements during resuscitation.