[The value of intraoperative cerebral oxygen saturation in predicting postoperative neurocognitive dysfunction in...

Objective: To evaluate the value of intraoperative cerebral oxygen saturation in predicting postoperative neurocognitive dysfunction (PND) in elderly patients with mild cognitive impairment. Methods: A total of 210 cases of lumbar decompression, bone grafting and fusion surgery under general anesthesia were collected in the Third Central Hospital of Tianjin from June, 2019 to January, 2020, either sex, aged 65-75 year, BMI 19.5-32.5 kg/m(2), ASA physical status Ⅱ or Ⅲ, preoperative comorbidities with mild cognitive impairment. MoCA and MMSE were used to evaluate the cognitive function of patients 1 day before the operation, 7 days and 3 monthes after operation. PND group (n=38) and non-PND group (n=172) were selected according to postoperative MMSE and MoCA scale scores and the diagnostic criteria of PND. Heart rate (HR) , mean arterial pressure (MAP), pulse oxygen saturation (SpO(2)), bispectral index (BIS), cerebral tissue oxygen saturation (SctO(2), average left and right brain SctO(2) were recorded) were recorded pre-anesthetic (T(0)), ten minutes of anesthesia(T(1)), twenty minutes of anesthesia (T(2)), thirty minutes into the operation (T(3)), one hour into the operation (T(4)), end of the surgery (T(5)), and leave the PACU (T(6)). SctO(2) at time point T(0) was the base value of SctO(2), and the maximum percentage drop in SctO(2) from the base value was calculated (SctO(2max)%). Results: The incidence of PND was 18% (38/210) in 210 elderly patients undergoing surgery. The age of PND group and non-PND group was (71.0±2.1) and (67.8±2.0) years old, and the PACU time was (57±5) and (46±8) min, respectively. Compared with the non-PND group, the age of the PND group was higher (t=2.600, P<0.05) and the PACU time was longer (t=3.039, P<0.05). At the time points T(3), T(4), T(5) and T(6), SctO(2) in the PND group was (62±10) %, (60±11) %, (64±12) % and (66±10)%, respectively, lower than that in the non-PND group (67±60) %, (68±6) %, (69±5) % and (70±7)%, respectively, and the difference was statistically significant (t=3.369, 4.906, 3.787, 2.516, all P<0.05).The MoCA and MMSE scores of the PND group were (22.9±1.2) and (24.1±1.2) points, respectively, 1 day before surgery; and the MoCA and MMSE scores of the PND group were reduced to (20.8±1.2) and (21.3±0.7) points, respectively, 7 days after surgery, with statistically significant differences (t=3.523, 5.675, all P<0.05). MoCA and MMSE scores 7 days after surgery in the non-PND group were (22.4±1.3) and (23.1±1.6) points, respectively. Compared with the non-PND group, MoCA and MMSE scores 7 days after surgery in the PND group were reduced (t=2.630, 3.108, all P<0.05). The critical value of intraoperative SctO(2max)% was 13.74%, the area under the curve of PND was predicted to be 0.907 (95%CI: 0.819-0.995), sensitivity and specificity were 88.9% and 88.5%, respectively. Conclusion: SctO(2max)%>13.74% can be used as an indicator to predict PND occurrence in elderly patients with mild cognitive impairment during lumbar surgery.