Machine learning for prediction of postoperative nausea and vomiting in patients with intravenous patient-controlled analgesia.

BACKGROUND: Postoperative nausea and vomiting (PONV) is a still highly relevant problem and is known to be a distressing side effect in patients. The aim of this study was to develop a machine learning model to predict PONV up to 24 h with fentanyl-based intravenous patient-controlled analgesia (IV-PCA). METHODS: From July 2019 and July 2020, data from 2,149 patients who received fentanyl-based IV-PCA for analgesia after non-cardiac surgery under general anesthesia were applied to develop predictive models. The rates of PONV at 1 day after surgery were measured according to patient characteristics as well as anesthetic, surgical, or PCA-related factors. All statistical analyses and computations were performed using the R software. RESULTS: A total of 2,149 patients were enrolled in this study, 337 of whom (15.7%) experienced PONV. After applying the machine-learning algorithm and Apfel model to the test dataset to predict PONV, we found that the area under the receiver operating characteristic curve using logistic regression was 0.576 (95% confidence interval [CI], 0.520-0.633), k-nearest neighbor was 0.597 (95% CI, 0.537-0.656), decision tree was 0.561 (95% CI, 0.498-0.625), random forest was 0.610 (95% CI, 0.552-0.668), gradient boosting machine was 0.580 (95% CI, 0.520-0.639), support vector machine was 0.649 (95% CI, 0.592-0.707), artificial neural network was 0.686 (95% CI, 0.630-0.742), and Apfel model was 0.643 (95% CI, 0.596-0.690). CONCLUSIONS: We developed and validated machine learning models for predicting PONV in the first 24 h. The machine learning model showed better performance than the Apfel model in predicting PONV.

Gastric Ultrasound Assessing Gastric Emptying of Preoperative Carbohydrate Drinks: A Randomized Controlled Noninferiority Study.

BACKGROUND: Tools for the evaluation of gastric emptying have evolved over time. The purpose of this study was to show that the risk of pulmonary aspiration is not increased with carbohydrate drink, by demonstrating that the gastric antral cross-sectional area (CSA) of the NO-NPO group is either equivalent to or less than that of the NPO (nil per os) group. METHODS: Sixty-four patients scheduled for elective laparoscopic benign gynecologic surgery were enrolled and randomly assigned to the NPO group (n = 32) or the NO-NPO group (n = 32). After having a regular meal until midnight before surgery, the NPO group fasted until surgery, while the NO-NPO group ingested 400 mL of a carbohydrate drink at midnight and freely up to 2 hours before anesthesia. The primary outcome was the gastric antral CSA by gastric ultrasound in right lateral decubitus position (RLDP). Noninferiority was defined as a mean difference of CSA <2.8 cm2. Secondary outcomes included CSA in supine position, gastric volume (GV), GV per weight (GV/kg), GV/kg >1.5 mL/kg, and Perlas grade. RESULTS: CSA in RLDP was not different between the NPO group (6.25 ± 3.79 cm2) and the NO-NPO group (6.21 ± 2.48 cm2; P = .959). The mean difference of CSA in RLDP (NO-NPO group - NPO group) was 0.04 (95% confidence interval [CI], -1.56 to 1.64), which was within the noninferiority margin of 2.8 cm2. CSA was not different between the 2 groups (4.17 ± 2.34 cm2 in NPO group versus 4.28 ± 1.23 cm2 in NO-NPO group; P = .828). GV in NPO group (70 ± 56 mL) was not different from NO-NPO group (66 ± 36 mL; mean difference, 3.66; 95% CI, -20 to 27; P = .756). GV/kg in the NPO group (1.25 ± 1.00 mL/kg) was not different from the NO-NPO group (1.17 ± 0.67 mL/kg; P = .694). The incidence of GV/kg > 1.5 mL/kg was not different between NPO (31.3%) and NO-NPO group (21.9%; P = .768). The median (interquartile range) of the Perlas grade was 1 (0-1) in NPO group and 0.5 (0-1) in NO-NPO group (P = .871). CONCLUSIONS: Preoperative carbohydrates ingested up to 2 hours before anesthesia do not delay gastric emptying compared to midnight fasting, as evaluated with gastric ultrasound.

Comparison of two different shapes of stylets for intubation with the McGrath MAC® video laryngoscope: a randomized controlled trial.

OBJECTIVE: This study was performed to compare two different shapes of stylets, 60° and J-shaped stylets, for intubation using the McGrath MAC® video laryngoscope (MVL). METHODS: Two hundred twenty-two patients undergoing surgery under general anesthesia were randomly allocated to Group J (n = 111) or Group 60° (n = 111) and intubated using the MVL with the stylet bent into the allocated shape. The time to intubation (TTI) and other intubating profiles were compared between the groups. Multivariate regression analysis was used to determine the relationship between factors related to difficult intubation and TTI. RESULTS: The TTI was not different between the two groups. There were also no differences in the intubating profiles between the two groups. In both groups, the TTI was longer with a modified Mallampati score (mMS) of ≥3 and percentage of glottic opening (POGO) score of <50. In Group J, the TTI was longer with a body mass index (BMI) of ≥30 kg/m2. CONCLUSION: The TTI during tracheal intubation with the MVL was not different between the two groups. The TTI was longer with an mMS of ≥3 and POGO score of <50. In Group J, the TTI was longer with a BMI of ≥30 kg/m2.

Comparison of single minimum dose administration of dexmedetomidine and midazolam for prevention of emergence delirium in children: a randomized controlled trial.

PURPOSE: Emergence delirium (ED) is common in children after sevoflurane anesthesia and should be prevented for patient safety. A prospective, double-blind, randomized, controlled study was performed to compare the efficacy of minimal dosage of midazolam versus dexmedetomidine to prevent ED in children undergoing tonsillectomy. METHODS: Seventy children aged 24 months to 12 years were allocated to receive midazolam (0.03 mg/kg) or dexmedetomidine (0.3 µg/kg) 5 min before the end of surgery. The incidence and severity of ED were assessed using a four-point scale and the pediatric anesthesia emergence delirium scale, respectively. The emergence time and postoperative pain scores were also evaluated. RESULTS: The incidence of ED was 31.3% in the midazolam group and 26.5% in the dexmedetomidine group (P = 0.668). The severity of ED was similar in both groups (9.6 ± 5.8 in the midazolam group, vs. 8.1 ± 5.9 in the dexmedetomidine group, P = 0.299). The emergence time was comparable in the two groups [11.0 (8.3-13.8) min in midazolam group vs. 12.0 (10.0-13.5) min in dexmedetomidine group (P = 0.218)]. Postoperative pain score was higher in the midazolam group [0 (0-1)] than in the dexmedetomidine group [0 (0-0)] (P = 0.011). CONCLUSION: Dexmedetomidine and midazolam at single minimum dosages had equal effectiveness to prevent ED in children without delaying emergence time, when administered at the end of surgery. With regards to postoperative analgesic efficacy, although dexmedetomidine showed statistically significant higher analgesic effect than midazolam, further clinical investigations are needed to validate our findings.

Evaluation of Gastric Emptying Time of a Rice-Based Meal Using Serial Sonography.

The aim of this prospective study was to evaluate the gastric emptying time of a rice-based meal by serial ultrasonography of the stomach. After baseline ultrasonographic assessment of ten fasted healthy volunteers, volunteers ingested standardized 420 g, 536 kcal rice-based meal (bibimbap), and serial evaluations were performed every hour until the stomach became empty. At baseline, all the participants had an empty stomach. The average time of complete gastric emptying of the rice-based meal was 5.8 ± 0.8 h (95% confidence interval (CI), 5.0 h to 6.5 h). Since the first postintake cross-sectional area (CSA) measurement, a decrease was observed, and CSA was maintained until postprandial 3-4 h (P > 0.05). It declined rapidly 4 h after meal intake (P=0.031), reaching the nadir at approximately 6 h after meal intake. The gastric CSA and hunger score showed a positive correlation (correlation r = 0.616, P < 0). The rice-based meal is emptied after 5.8 ± 0.8 h on average in healthy volunteers. Based on our results, 6.5 h (upper limit of CI) of fasting after the ingestion of a rice-based meal would be a safe preoperative fasting time, and this is in accordance with the current guidelines for preoperative fasting.