Anaesthetist prediction of postoperative opioid use: a multicentre prospective cohort study.

Background: The Apfel simplified risk score includes four risk factors: female sex, non-smoking status, postoperative nausea and vomiting or motion sickness history, and postoperative opioid use. The score is calculated preoperatively, so postoperative opioid use must be predicted. We aimed to determine whether anaesthetists can predict patients' postoperative opioid use and dose. Methods: Specialist anaesthetists from eight hospitals preoperatively predicted opioid use and dose in the post-anaesthesia care unit (PACU) and for the first 24 h postoperatively, which was compared with actual opioid use and dose. Opioid doses were converted to oral morphine equivalents (MEQ). Correlations between predicted and actual opioid use and dose were analysed with Spearman's rho and linear regression. Results: A total of 487 anaesthetist-patient pairs were included. Anaesthetists overpredicted opioid use (398 [82%] predicted vs 251 [52%] actual patients requiring opioids in the PACU; 396 [81%] predicted vs 291 [60%] actual in the first 24 h) (Spearman's rho [95% confidence interval] 0.24 [0.16-0.33], P<0.001 in the PACU; 0.36 [0.28-0.44], P<0.001 in the first 24 h). Anaesthetists also overpredicted opioid dose (median [inter-quartile range] 12 [8-20] mg predicted MEQ vs 4 [0-18] mg actual MEQ in the PACU; 32 [18-60] mg vs 24 [0-65] mg MEQ in the first 24 h) (Spearman's rho 0.21 [0.13-0.29], P<0.001 in the PACU; 0.53 [0.40-0.60], P<0.001 in the first 24 h). Conclusions: Specialist anaesthetists cannot accurately predict opioid use or dose in the PACU or the first 24 postoperative hours. The Apfel risk criterion for postoperative opioid use may be inaccurate in clinical practice.

Artificial intelligence and machine learning in prehospital emergency care: A scoping review.

Our scoping review provides a comprehensive analysis of the landscape of artificial intelligence (AI) applications in prehospital emergency care (PEC). It contributes to the field by highlighting the most studied AI applications and identifying the most common methodological approaches across 106 included studies. The findings indicate a promising future for AI in PEC, with many unique use cases, such as prognostication, demand prediction, resource optimization, and the Internet of Things continuous monitoring systems. Comparisons with other approaches showed AI outperforming clinicians and non-AI algorithms in most cases. However, most studies were internally validated and retrospective, highlighting the need for rigorous prospective validation of AI applications before implementation in clinical settings. We identified knowledge and methodological gaps using an evidence map, offering a roadmap for future investigators. We also discussed the significance of explainable AI for establishing trust in AI systems among clinicians and facilitating real-world validation of AI models.

Smaller Return Cannula in Venoarterial Extracorporeal Membrane Oxygenation Does Not Increase Hemolysis: A Single-Center, Cohort Study.

The aim of this study was to explore the association between arterial return cannula diameter and hemolysis during peripheral VA ECMO. We identified 158 adult patients who received peripheral VA ECMO at our institution from the national ECMO database (EXCEL) between January 2019 and July 2021. We classified patients into a small cannula group (15 Fr diameter, n = 45) and a large cannula group (≥17 Fr diameter, n = 113), comparing incidences of clinical hemolysis and plasma free hemoglobin ( pf Hb). Moderate hemolysis is defined as having pf Hb 0.05-0.10 g/L and severe hemolysis as having pf Hb >0.10 g/L sustained for at least two consecutive readings or leading to a circuit change. There were no significant differences in rates of moderate hemolysis between small and large cannula groups (1 vs . 6; p = 0.39) and severe hemolysis (0 vs . 3; p = 0.27), nor was the pf Hb level significantly different at 4 hours (0.086 ± 0.096 vs . 0.112 ± 0.145 g/L; p = 0.58) and at 24 hours (0.042 ± 0.033 vs . 0.051 ± 0.069 g/L; p = 0.99). There were no increased rates of hemolysis when comparing small versus large arterial return cannula diameter in peripheral VA ECMO.

Hospital medication errors: a cross-sectional study.

BACKGROUND: Medication errors (MEs) are among the most common types of incidents reported in Australian and international hospitals. There is no uniform method of reporting and reducing these errors. This study aims to identify the incidence, time trends, types and factors associated with MEs in a large regional hospital in Australia. METHODS: A 5-year cross-sectional study. RESULTS: The incidence of MEs was 1.05 per 100 admitted patients. The highest frequency of errors was observed during the colder months of May-August. When distributed by day of the week, Mondays and Tuesdays had the highest frequency of errors. When distributed by hour of the day, time intervals from 7 am to 8 am and from 7 pm to 8 pm showed a sharp increase in the frequency of errors. One thousand and eighty-eight (57.8%) MEs belonged to incidence severity rating (ISR) level 4 and 787 (41.8%) belonged to ISR level 3. There were six incidents of ISR level 2 and only one incident of ISR level 1 reported during the five-year period 2014-2018. Administration-only errors were the most common accounting for 1070 (56.8%) followed by prescribing-only errors (433, 23%). High-risk medications were associated with half the number of errors, the most common of which were narcotics (17.9%) and antimicrobials (13.2%). CONCLUSIONS: MEs continue to be a problem faced by international hospitals. Inexperience of health professionals and nurse-patient ratios might be the fundamental challenges to overcome. Specific training of junior staff in prescribing and administering medication and nurse workload management could be possible solutions to reducing MEs in hospitals.