A total of 150 patients scheduled for Abdominal surgery gave written informed consent and were enrolled in this study.
Common demographic factors were collected when the patients were admitted to the hospital, for example, age, gender, body mass index (BMI), and education level. Moreover, the hemocyte indexes, hepatic and renal functions were assessed by blood routine examination 1 day before and 3 days after the surgery.
To evaluate the patients'neurocognitive levels, a well-trained investigator performed a questionnaire to roundly assess their neurocognitive functions 1 day before and 3 days after the surgery. The questionnaire included: (1) Mini-Mental State Examination (MMSE); (2) Hopkins Verbal Learning Test-Revised, testing ability of Verbal learning and memory skills; (3) Brief Visuospatial Memory Test-Revised, testing ability of Visuospatial memory; (4) Trail Making Test; (5) Digital Scope Test; (6) Word Delayed Recall Test, testing ability of long-memory; (7) Word Delayed Interference Test;(8) Image Delayed Recall Test, testing ability of delay recall; (9) Verbal Fluency Test, testing fluency and executive function. At the very beginning, patients with severe cognitive dysfunction were excluded by the lower MMSE score than we required. Each qualified patient gets scores of these tests preoperatively and postoperatively. The investigators collected all of these scores and then obtained the standard deviation (SD) of every preoperative test. The investigators defined a worse test performance as a negative changed score, with an absolute value larger than 1.5 times SD. Patients with no less than two nega- tive changed scores, except for MMSE, were diagnosed as PND . According to this standard, the investigators divided the participants into the PND group and the non-PND group. Finally, the investigators defined two decreased scores as mild PND and more than two decreased scores as severe PND.
Patients got the same preoperative preparation, including ECG monitoring, blood oxygen saturation monitoring, arterial blood pressure, and central venous pressure monitoring. As for the anesthesia method, the investigators took combined anesthesia of inhalation and intravenous. The main anesthetic included sevoflurane, propofol, cisatracurium besylate or rocuronium bromide , and lidocaine for topical anesthesia. During the surgery, the patients were given mechanical ventilation at 12 times/min, with a tidal volume of 6-8ml/kg. Systemic arterial blood pressure was measured via radial artery catheterization. After induction of anesthesia, a central venous catheter was inserted for monitoring central venous pressure (CVP) and fluid management.
At the end of the operation peripheral venous blood was sampled to measure IL-33 expression in peripheral blood monocytes.After the surgery, the surgery time and the anesthesia time were recorded.
PND was diagnosed using the same definition as used in the ISPOCD1-study and previous other studies. To quantify practice effect, baseline scores were compared with subsequent test results three days later in control subjects. For patients, preoperative scores were compared with postoperative test results, subtracted the average practice effect from these changes, and then divided the result by the control-subject SD to obtain a Z score for each test. The test results were adjusted so that a positive Z score indicated deterioration from the baseline test. The Z scores of all tests in an individual patient were then summarized and divided by the SD for this sum of Z scores in the control subjects, creating a combined Z score. A patient was defined as having POCD when two Z scores in individual tests or the combined Z score were 2 or more.
Statistical analysis A continuous outcome was expressed as the mean ± SE, while a categorical outcome was expressed as percentages or frequencies. Differences in continuous and categorical outcomes were tested using the one-way analysis of variance and Chi square analysis, respectively. The unpaired Student t test, paired T test, analysis of variance and Pearson correlation analysis were performed on normally distributed data. Case-control differences in nominal data were evaluated with the Chi square test. Multivariate Cox regression analysis was done for predictors of adverse events. The IL-33 and neurobehavioral scores were compared using ROC curves. Kaplan-Meier curves were used for correlation with POCD. All statistical analyses were performed with SPSS 21.0 (IBM Inc.) and P \< 0.05 was considered signifcant.
