Relationship between preoperative cerebrospinal fluid/serum albumin ratio and postoperative delirium in patients undergoing neuraxial anesthesia / 中华麻醉学杂志

ABSTRACT

Objective:To evaluate the relationship between preoperative cerebrospinal fluid/serum albumin ratio (Q-alb) and postoperative delirium (POD) in patients undergoing neuraxial anesthesia.Methods:The patients, aged 40-90 yr, of American Society of Anesthesiologists physical status Ⅰ or Ⅱ, underwent total knee/hip replacement under combined spinal-epidural block in our hospital from January 2018 to December 2020, were collected.After admission to the operating room, venous blood and cerebrospinal fluid samples were collected for determination of cerebrospinal fluid albumin, β-amyloid (Aβ) 1-42, Aβ 1-40, total tau protein (t-Tau), phosphorylated tau protein (p-Tau) and serum albumin levels (by enzyme-linked immunosorbent assay) and for calculation of Q-alb.When Q-alb was more than 10.2, the patient was considered to have blood-brain barrier disruption.Mini-Mental State Examination scale was used to evaluate the cognitive level on 1 day before surgery. The development of POD was evaluated using Confusion Assessment Method Chinese Reversion and Memorial Delirium Assessment Scale at 1-7 days after surgery.The patients were divided into POD group (P group) and non-POD (NP group) according to whether POD occurred.The receiver operating characteristic (ROC) curve was used to analyze the accuracy of Q-alb in predicting POD. Results:There were 49 cases in each group.Compared with group NP, concentrations of Aβ 1-42 and Aβ 1-40 were significantly decreased, concentrations of t-Tau and p-Tau albumin were increased, the ratio of Q-alb and blood-brain barrier disruption was increased in group P ( P<0.05). Before and after adjusting for confounding factors, Q-alb, cerebrospinal fluid Aβ 1-42, Aβ 1-40, t-Tau and p-Tau levels were risk factors for POD ( P<0.05). There was a positive linear regression relationship between Q-alb and levels of t-Tau and p-Tauin cerebrospinal fluid (t-Tau β=0.587, P<0.001; p-Tau β=0.427, P<0.001), and there was a negative linear regression relationship between Q-alb and levels of Aβ 1-42 and Aβ 1-40 in cerebrospinal fluid (Aβ 1-42 β=-0.762, P<0.001; Aβ 1-40 β=-0.531, P<0.001). There was no linear regression relationship between Q-alb and level of p-Tau in group P ( P=0.121). There was no linear regression relationship between Q-alb and level of Aβ 1-40 in group NP ( P=0.467). The results of ROC curve analysis showed that the area under the curve for Q-alb in predicting POD (95% confidence interval) was 0.827 (0.738-0.896). Conclusion:Preoperative higher Q-alb is the risk factor for POD in patients undergoing neuraxial anesthesia, and is more accurate in predicting POD.