The Relationship Between Perioperative Regional Cerebral Oxygen Saturation and Postoperative Cognitive Dysfunction: A Systematic Review and Meta-Analysis

Qing Feng,Song Jiang,Songyun Deng,Yuhang Ai,Qianyi Peng,Yunan Mo,Lina Zhang
DOI: https://doi.org/10.4103/jtccm.jtccm_2_21
2020-01-01
Journal of Translational Critical Care Medicine
Abstract:Objective: Postoperative cognitive dysfunction (POCD) and neurologic deficit continues to be an important neuropsychological adverse affecting patients' outcome. We conducted this systematic review to investigate the relationship between regional cerebral oxygen saturation (rSO2) and early POCD in postoperative patients. Materials and Methods: Our search included MEDLINE (PubMed) and Cochrane library, from inception to October 31, 2018. We included studies reporting values of rSO2 at the beginning of and/or during surgery, and the primary outcome was POCD, and excluded articles who do not put postoperative cognitive function as the main observation. Two reviewers assessed the quality of the included articles using the Newcastle-Ottawa Scale and extracted the data. Results: Seven prospective cohort studies that included 532 patients incorporate into this systematic review. About 37.78% (201/532) patients achieved POCD. POCD was associated with significantly lower overall values of rSO2 during surgery compared to the population that did not achieve POCD (mean difference [MD] −1.98; 95% confidence interval [CI] −3.30 to −0.66; P = 0.003). In the subgroup, the lowest rSO2 values intraoperative period were a better predictor of POCD (MD −2.91; 95% CI −4.37 to − 1.46; P < 0.0001) than the mean rSO2 values (MD − 2.94; 95% CI −5.71 to − 0.17; P = 0.04). However, all of two were proved superior to baseline rSO2 values (MD −0.13; 95% CI -1.33 to 1.07; P = 0.83); patients with intraoperative cerebral oxygen <50% are nearly four times more likely to have early POCD (odds ratio = 3.65; 95% CI 1.62–8.23, P = 0.002). Conclusions: Patients with POCD have significantly lower cerebral oxygenation during operation than their counterparts. The lowest rSO2 values intraoperative period and patients with intraoperative cerebral oxygen <50% were a better predictor of POCD.
What problem does this paper attempt to address?