Near-infrared spectroscopy and processed electroencephalogram monitoring for predicting peri-operative stroke risk in cardiothoracic surgery

BACKGROUND Stroke is a feared complication after cardiothoracic surgery, with an incidence of around 2 to 3%. Anaesthesia and postoperative sedation may obscure clinical symptoms of stroke and thus delay diagnosis and timely intervention. OBJECTIVES The objective was to assess the value of intra-operative neuromonitoring and blood pressure monitoring for predicting the occurrence of peri-operative stroke within 3 days after cardiothoracic surgery. DESIGN Single-centre retrospective observational cohort study. SETTING Academic tertiary care medical centre. PATIENTS All consecutive patients with cardiothoracic surgery and intra-operative neuromonitoring admitted postoperatively to the Intensive Care Unit (ICU) between 2008 and 2017. MAIN OUTCOME MEASURES The primary endpoint was the occurrence of any stroke confirmed by brain imaging within 3 days postcardiothoracic surgery. Areas under the curve (AUC) of intra-operative mean arterial pressure (MAP), cerebral oxygen saturation (ScO 2 ) and bispectral index (BIS) below predefined thresholds were calculated, and the association with early stroke was tested using logistic regression analyses. RESULTS A total of 2454 patients admitted to the ICU after cardiothoracic surgery had complete intra-operative data for ScO 2 , BIS and MAP and were included in the analysis. In 58 patients (2.4%), a stroke was confirmed. In univariate analysis, a larger AUC MAP<60 mmHg [odds ratio (OR) 1.43; 95% confidence interval (CI), 1.21 to 1.68) and larger AUC BIS<25 (OR 1.51; 95% CI, 1.24 to 1.83) were associated with the occurrence of postoperative stroke while ScO 2 less than 50% or greater than 20% reduction from individual baseline was not (OR 0.91; 95% CI, 0.50 to 1.67). After multivariable analysis, AUC BIS<25 (OR 1.45; 95% CI, 1.12 to 1.87) and longer duration of MAP less than 60 mmHg (OR 1.52; 95% CI, 1.02 to 2.27) remained independently associated with stroke occurrence. CONCLUSION Cumulative intra-operative BIS values below 25 and longer duration of MAP below 60 mmHg were associated with the occurrence of peri-operative stroke within 3 days after cardiothoracic surgery. Prospective studies are warranted to evaluate a causal relationship between low BIS and stroke to establish whether avoiding intra-operative BIS values below 25 might reduce the incidence of peri-operative stroke.