*2.2. Anesthesia*

Sedative premedication was prescribed in a dosage of 0.01–0.02 mg/kg midazolam, which was adjusted to the patient's condition. Before induction of anesthesia, entropy electrodes were applied to the patient's forehead as recommended by the manufacturer. Anesthesia was induced using propofol or etomidate (depending on the indications) in combination with fentanyl 2–3 μg/kg, followed by a neuromuscular block to facilitate tracheal intubation with rocuronium 0.6–1 mg/kg. Anesthesia was maintained using a volatile anesthetic (sevoflurane). The anesthesiologists were unrestricted in using conventional regimens of opioid analgesics and neuromuscular blocking agents as required. In order to maintain an anesthetic state in the SM Group, anesthesia was adjusted according to somatic response and hemodynamic events, while in the ESM Group, the anesthesiologists tailored anesthesia to achieve state entropy between 40–60 and an SPI value ≤ 50.

#### *2.3. Data Collection, Assesment of Postoperative of POCD, and Delirium*

For each patient included in the ESM Group, we recorded the state entropy (SE) and response entropy (RE) in the awakening state, every 15 min after the beginning of surgery and at extubation time. Because negligible differences exist between state and responsive entropy in curarized patients, we decided to acquire only state entropy data. Other data collected included patient characteristics, surgical procedure, anesthetic data, and the intraoperative hemodynamics.

Between the groups, we recorded and compared the incidence of hypotension, bradycardia, tachycardia (variation of more than 20% from preinduction values of mean arterial blood pressure and heart rate). Patients were discharge from the post-anesthesia care unit based on the modified Aldrete score criteria. Postoperative analgesia was guided according to patient demands and consisted of 1 g paracetamol every 6 h, 20 mg nefopam every 12 h, or morphine (0.1 mg/kg) every 8 h, as well as 50–100 mg ketoprofen every12 h in selected cases.

Postoperative cognitive dysfunctions were assessed 24 h, 48 h, and 72 h after surgery using the NEECHAM Confusion Scale. Patients' cognitive status could not be further evaluated because the majority of patients were discharged from ICU after 3 days. Another reason for taking into account only the first 3 postoperative days was other postoperative events that could have interfered with our findings. Screening was performed by trained medical personnel.

#### *2.4. Statistical Analysis*

The objective of this study was to demonstrate that the use of entropy and SPI monitoring in assessing anesthetic depth in emergency surgery is associated with a reduction in postoperative cognitive dysfunctions events. GraphPad 8Prism and MedCalc14.1 were used for statistical analysis.

Given the objective of this study, the correlations between the doses of anesthetics used and the NEECHAM score imposed a sufficient sample size to meet this goal. For this calculation, we used the MedCalc program 14.1 (Sampling-Correlation coefficient). We considered it appropriate to use a significance level of 0.05 to avoid the occurrence of a type 1 error (alpha level 2-sided) and 0.1 to avoid the occurrence of a type 2 error (beta) using an input of the correlation coefficient of 0.5 (the hypothesized or anticipated

correlation coefficient). At least 29 patients were required for each group and patients were randomized according to the permuted block technique.

The Anderson–Darling test was used to test the data distribution. Data with normal distribution were compared using the student's t-test and presented as mean with SD, and data that did not follow the normal distribution were analyzed using nonparametric tests (Mann–Whitney). Different methods for correlation analyses available from MedCalc14.1 were performed, namely Pearson correlation (r) for Gaussian distribution and Spearman rho for nonparametric data. Nominal data were compared using the chi-square test or Fisher's exact test. A *p*-value < 0.05 was considered statistically significant.

In order identify how we could avoid postoperative cognitive dysfunction, we developed a logistic regression model that used a NEECHAM score at 24 h higher than 24 (indicating the absence of cognitive dysfunction) as a dependent variable. The logistic regression model included the use of entropy monitoring and doses of fentanyl, sevoflurane, and norepinephrine.
