Good clinical research practice (GCRP) in pharmacodynamic studies of neuromuscular blocking agents III: The 2023 Geneva revision

The set of guidelines for good clinical research practice in pharmacodynamic studies of neuromuscular blocking agents was developed following an international consensus conference in Copenhagen in 1996 (Viby-Mogensen et al., Acta Anaesthesiol Scand 1996, 40 , 59–74); the guidelines were later revised and updated following the second consensus conference in Stockholm in 2005 (Fuchs-Buder et al., Acta Anaesthesiol Scand 2007, 51 , 789–808). In view of new devices and further development of monitoring technologies that emerged since then, (e.g., electromyography, three-dimensional acceleromyography, kinemyography) as well as novel compounds (e.g., sugammadex) a review and update of these recommendations became necessary. The intent of these revised guidelines is to continue to help clinical researchers to conduct high-quality work and advance the field by enhancing the standards, consistency, and comparability of clinical studies. There is growing awareness of the importance of consensus-based reporting standards in clinical trials and observational studies. Such global initiatives are necessary in order to minimize heterogeneous and inadequate data reporting and to improve clarity and comparability between different studies and study cohorts. Variations in definitions of endpoints or outcome variables can introduce confusion and difficulties in interpretation of data, but more importantly, it may preclude building of an adequate body of evidence to achieve reliable conclusions and recommendations. Clinical research in neuromuscular pharmacology and physiology is no exception.     

Frame Invariance and Scalability of Neural Operators for Partial Differential Equations

Partial differential equations (PDEs) play a dominant role in the mathematical modeling of many complex dynamical processes. Solving these PDEs often requires prohibitively high computational costs, especially when multiple evaluations must be made for different parameters or conditions. After training, neural operators can provide PDEs solutions significantly faster than traditional PDE solvers. In this work, invariance properties and computational complexity of two neural operators are examined for transport PDE of a scalar quantity. Neural operator based on graph kernel network (GKN) operates on graph-structured data to incorporate nonlocal dependencies. Here we propose a modified formulation of GKN to achieve frame invariance. Vector cloud neural network (VCNN) is an alternate neural operator with embedded frame invariance which operates on point cloud data. GKN-based neural operator demonstrates slightly better predictive performance compared to VCNN. However, GKN requires an excessively high computational cost that increases quadratically with the increasing number of discretized objects as compared to a linear increase for VCNN.     

Patient characteristics,anaesthetic workload and techniques in the UK: an analysis from the 7th National Audit Project (NAP7) activity survey

Detailed contemporary knowledge of the characteristics of the surgical population, national anaesthetic workload, anaesthetic techniques and behaviours are essential to monitor productivity, inform policy and direct research themes. Every 3–4 years, the Royal College of Anaesthetists, as part of its National Audit Projects (NAP), performs a snapshot activity survey in all UK hospitals delivering anaesthesia, collecting patient-level encounter data from all cases under the care of an anaesthetist. During November 2021, as part of NAP7, anaesthetists recorded details of all cases undertaken over 4 days at their site through an online survey capturing anonymous patient characteristics and anaesthetic details. Of 416 hospital sites invited to participate, 352 (85%) completed the activity survey. From these, 24,177 reports were returned, of which 24,172 (99%) were included in the final dataset. The work patterns by day of the week, time of day and surgical specialty were similar to previous NAP activity surveys. However, in non-obstetric patients, between NAP5 (2013) and NAP7 (2021) activity surveys, the estimated median age of patients increased by 2.3 years from median (IQR) of 50.5 (28.4–69.1) to 52.8 (32.1–69.2) years. The median (IQR) BMI increased from 24.9 (21.5–29.5) to 26.7 (22.3–31.7) kg.m^–2. The proportion of patients who scored as ASA physical status 1 decreased from 37% in NAP5 to 24% in NAP7. The use of total intravenous anaesthesia increased from 8% of general anaesthesia cases to 26% between NAP5 and NAP7. Some changes may reflect the impact of the COVID-19 pandemic on the anaesthetic population, though patients with confirmed COVID-19 accounted for only 149 (1%) cases. These data show a rising burden of age, obesity and comorbidity in patients requiring anaesthesia care, likely to impact UK peri-operative services significantly.     

Machine learning to predict myocardial injury and death after non-cardiac surgery

Myocardial injury due to ischaemia within 30 days of non-cardiac surgery is prognostically relevant. We aimed to determine the discrimination, calibration, accuracy, sensitivity and specificity of single-layer and multiple-layer neural networks for myocardial injury and death within 30 postoperative days. We analysed data from 24,589 participants in the Vascular Events in Non-cardiac Surgery Patients Cohort Evaluation study. Validation was performed on a randomly selected subset of the study population. Discrimination for myocardial injury by single-layer vs. multiple-layer models generated areas (95%CI) under the receiver operating characteristic curve of: 0.70 (0.69–0.72) vs. 0.71 (0.70–0.73) with variables available before surgical referral, p < 0.001; 0.73 (0.72–0.75) vs. 0.75 (0.74–0.76) with additional variables available on admission, but before surgery, p < 0.001; and 0.76 (0.75–0.77) vs. 0.77 (0.76–0.78) with the addition of subsequent variables, p < 0.001. Discrimination for death by single-layer vs. multiple-layer models generated areas (95%CI) under the receiver operating characteristic curve of: 0.71 (0.66–0.76) vs. 0.74 (0.71–0.77) with variables available before surgical referral, p = 0.04; 0.78 (0.73–0.82) vs. 0.83 (0.79–0.86) with additional variables available on admission but before surgery, p = 0.01; and 0.87 (0.83–0.89) vs. 0.87 (0.85–0.90) with the addition of subsequent variables, p = 0.52. The accuracy of the multiple-layer model for myocardial injury and death with all variables was 70% and 89%, respectively.