Research Appraisals
Evidence-based critical appraisals of the latest medical research, systematically evaluated using Oxford CEBM methodology.
Showing 59 appraisals
Scandinavian journal of pain
Multimodal deep learning fusion for automatic pain detection in cancer patients
INTRODUCTION: Since pain is a multidimensional and subjective experience, pain assessment remains challenging. With advances in artificial intelligence (AI), automatic pain assessment (APA) systems offer a valuable opportunity for objective pain evaluation. However, most approaches focus on a single modality. In this proof-of-concept study, exploring multimodal fusion strategies in a controlled experimental setting, we present a deep learning framework for multimodal fusion that combines facial, acoustic, and textual information to improve APA in cancer patients. METHODS: A multimodal dataset was created from video-recorded interviews with oncologic patients. In Phase I, audio, video, and transcripts were segmented at the sentence level and temporally aligned using the Eudico Linguistic Annotator (ELAN) to ensure frame-level correspondence across modalities. In Phase II, modality-specific features were extracted: Facial Action Units from OpenFace, acoustic descriptors (MFCCs, chroma, spectral contrast, and Mel-spectrogram) from a dedicated speech-processing pipeline, and sentence-level textual embeddings from ITA-BERT. During training, the most effective analytical strategy was chosen through knowledge transfer approaches. The ELAN-assisted annotation pipeline streamlined expert labeling. Two architectures were implemented and compared: bimodal autoencoder fusion models and a transformer-based model with pairwise cross-modal attention. These models were trained and evaluated using subject-independent and stratified 5-fold cross-validation. To address the lack of independence between segments, a strictly subject-independent cross-validation strategy was adopted. RESULTS: Knowledge transfer using pretrained large-scale models outperformed traditional feature-based approaches and was applied to multimodal pain detection. Multimodal models achieved performance comparable to the strongest unimodal modality (text), while showing improved balance across modalities, suggesting potential complementary effects. Both multimodal architectures demonstrated high accuracy in distinguishing between pain and non-pain classes. The bimodal autoencoder achieved stable results across folds, with a mean accuracy of about 80 % and balanced error distribution. The pairwise transformer with cross-modal attention achieved similar performance, with smooth training and validation loss curves. No evident divergence between training and validation loss curves was observed across folds, suggesting stable behavior within the cross-validation setting. However, subject-level overfitting cannot be excluded given the limited sample size. CONCLUSION: Multimodal fusion enhances system robustness by integrating complementary signals. Despite limitations and the need for improvement, multimodal deep learning strategies can support the detection of observable pain-related expressions.
26 July 2026
Read appraisal →BMJ open
Shared decision making in pain care: a systematic review study protocol of decision aids
INTRODUCTION: Shared decision making (SDM) in healthcare is an ethical imperative and essential to patient-centred care. SDM is particularly useful when several preference-sensitive treatment options exist and in the setting of chronic conditions or longitudinal management. Chronic pain management embodies these characteristics, yet SDM often remains insufficient in this population. Decision aids are designed to facilitate SDM by helping patients and clinicians understand treatment options, clarify patient values, and guide collaborative decision processes. Despite their importance, the use of decision aids in pain management is inconsistent and their reported effectiveness has been variable. However, the current landscape of decision aids for pain management has not been described. This lack of understanding of what tools exist, how they are structured, what decisions they address and how they were developed makes it difficult to advance implementation efforts or identify meaningful gaps in available resources. As such, the purpose of this systematic review is to identify and characterise decision aids for pain management across the lifespan. Specifically, this review will describe the clinical decisions addressed, decision aid structures and delivery formats, development processes and outcomes used to evaluate the impact of existing decision aids. METHODS AND ANALYSIS: Electronic searches were performed in PubMed, CINAHL and Ovid Embase from inception through January 2026. Studies will be assessed for quality using the Mixed Methods Appraisal Tool. Data will be extracted and presented with the aim of describing the: (a) content and structures of pain-related decision aids, (b) development processes used to create existing pain-related decision aids and (c) outcome measures used to evaluate the impact of decision aids in clinical care. ETHICS AND DISSEMINATION: This review does not require ethics approval. Findings will be disseminated to clinicians, researchers and patients through journal publications, conference presentations and in collaboration with patient partners. PROSPERO REGISTRATION NUMBER: CRD420251085288.
11 July 2026
Read appraisal →Journal of the American Medical Informatics Association : JAMIA
Engineering biomarker representations of vital signs data enhances deep learning mortality prediction
OBJECTIVES: We evaluated bidirectional long short-term memory models for predicting inpatient mortality using different approaches to processing vital signs data collected during the initial 24 h of intensive care unit (ICU) admissions. MATERIALS AND METHODS: We compared 3 vital-sign representations: (1) raw data recorded every 5 min, (2) preprocessed data averaged hourly, and (3) preprocessed data using biomarker representations that extends a digital oximetry biomarker toolbox of PhysioZoo software, applied to blood pressure, heart rate, temperature, respiratory rate, and SpO2. RESULTS: Across 2 large ICU datasets, HiRID and eICU, models trained on the frequency-normalized representation achieved higher discrimination and lower Brier scores than those trained on raw 5-min and hourly averaged data. DISCUSSION: The use of biomarker representations of vital signs yielded the largest improvements in discrimination and overall probabilistic performance reflected by lower Brier scores for predicting inpatient mortality by deep learning. CONCLUSION: Thus, we recommend using a similar approach to vital signs preprocessing for time-series predictive models.
3 July 2026
Read appraisal →Medical science monitor : international medical journal of experimental and clinical research
Advances in Remote Monitoring Technology Applications in Anesthesia: A Narrative Review
The swift progress of digital and sensor technologies is hastening the incorporation of remote monitoring into anesthesiology. While several reviews have explored telemedicine and artificial intelligence in anesthesia, most existing summaries either focus on conceptual outlook or lack systematic comparison of technical platforms and original clinical validation data. The present review provides a comprehensive, clinically oriented synthesis of remote monitoring in anesthesia, with clear focuses on technical principles, perioperative applications, platform comparisons, and evidence-based clinical outcomes. We critically assess clinical advantages and implementation hurdles, covering the full perioperative pathway - preoperative evaluation, intraoperative observation, and postoperative recovery. A head-to-head comparison of fifth generation of cellular network technology (5G), the Internet of Things (IoT), and cloud computing platforms is presented to clarify their infrastructure demands and suitability for real-world clinical scenarios. Notably, this review summarizes available clinical evidence, including evidence from the Trial of Remote Continuous versus Intermittent National Early Warning Score Monitoring after major surgery (TRaCINg) - a feasibility randomized controlled trial whose exploratory findings suggest that continuous remote monitoring may reduce unplanned intensive care unit admissions, shorten hospital stays, and facilitate earlier detection of postoperative complications. We further propose innovative solutions including multimodal data fusion and federated learning-driven predictive analytics to overcome current limitations in data interoperability, security, and clinical effectiveness. Accordingly, this paper synthesizes the latest advances in remote monitoring technology in anesthesia, clarifies its unique value relative to existing reviews, and provides practical guidance for clinical translation and future research.
2 July 2026
Read appraisal →The Orthopedic clinics of North America
Wearable Devices and Artificial Intelligence to Enhance Perioperative Care in Total Knee Arthroplasty
Rapidly evolving wearable devices have the potential to transform health care delivery and enhance outcomes. Applications of wearable smart devices range from activity and motion tracking to real-time monitoring of vital signs and biochemical markers. Recent evidence provides support for the use of wearable devices in chronic disease management, rehabilitation, and perioperative care of patients with total knee arthroplasty. Data derived from these devices may enhance treatment personalization and early identification of complications. Challenges remain regarding accuracy, standardization, privacy, and patient adherence. Future advances must overcome these limitations.
2 July 2026
Read appraisal →Current opinion in anaesthesiology
Preoperative considerations for nonoperating room anesthesia
PURPOSE OF THE REVIEW: This review aims to address the unique challenges in nonoperating room anesthesia (NORA) locations, emphasizing the importance of patient selection, risk stratification, and comprehensive preoperative evaluation to ensure safe anesthetic care for increasingly complex patients. RECENT FINDINGS: The volume of NORA procedures has risen significantly, with patients often presenting higher comorbidity burdens and advanced age. Standardized protocols and validated assessment tools, such as the STOP-Bang questionnaire and a simple frailty questionnaire (e.g., FRAIL), can aid anesthesiologists in effectively stratifying risk and tailoring anesthesia plans. Challenges in NORA include logistical constraints, personnel dynamics, and environmental factors that can compromise patient safety. The integration of telemedicine and artificial intelligence into preoperative assessments shows promise in improving efficiency and safety by allowing for remote evaluations and tailored care. SUMMARY: Adequate patient selection and preoperative optimization are essential for enhancing outcomes in NORA locations. Adherence to national safety guidelines and multidisciplinary collaboration is crucial for navigating the complexities of remote anesthesia care. Future research should focus on refining preoperative screening methods and utilizing artificial intelligence to better address the unique needs of patients undergoing procedures outside traditional operating rooms.
23 June 2026
Read appraisal →Current opinion in anaesthesiology
Traditional learning meets adaptive learning: the present and future of education in nonoperating room anesthesia.
PURPOSE OF REVIEW: Nonoperating room anesthesia (NORA) and continuing education in this setting are both in a state of change. Procedures performed outside the operating room, especially those that require nuanced anesthetic care, are becoming increasingly prevalent. Numerous new innovations are transforming care across NORA settings, especially within gastroenterology, interventional and neurointerventional radiology, and electrophysiology. Concurrently, education for NORA providers is rapidly evolving, influenced by technological advancements and artificial intelligence. RECENT FINDINGS: Traditional learning, a mode of education in which each learner receives essentially the same content, is now blending with adaptive learning, a field of precision education in which dynamic educational content is curated for individual learners, timed and paced to their preferences and progress, and coupled with artificial intelligence-generated feedback. The mainstays of traditional education for NORA providers remain relevant: content created by professional societies, podcasts, textbooks, and primary literature. Yet new, artificial intelligence-driven approaches for adaptive learning have entered the scene: clinical decision-making tools, literature review and knowledge mapping assistants, online platforms for precision education, virtual reality, and synthetic patients or avatars. SUMMARY: This article explores the blending of traditional and adaptive educational modalities in NORA continuing education and discusses the implications and obstacles for the future.
23 June 2026
Read appraisal →Current pain and headache reports
Artificial Intelligence in Neuropathic Pain: From Mechanisms to Neuromodulation and Regenerative Strategies
PURPOSE OF REVIEW: Neuropathic pain remains challenging due to its heterogeneous mechanisms and variable treatment response. This narrative review evaluates recent advances in artificial intelligence (AI) and machine learning (ML) for patient phenotyping, treatment selection, outcome prediction, neuromodulation, and regenerative therapies. RECENT FINDINGS: ML-based phenotyping integrates genomic, neuroimaging, sensory, and behavioral data. Multimodal automatic pain assessment is approaching clinical deployment, supported by recent expert consensus. Large language models have been benchmarked against multidisciplinary teams for spinal cord stimulation (SCS) candidate selection. Real-world cohort data confirm sustained 24-month SCS benefit. AI is increasingly applied to regenerative interventions, including pulsed radiofrequency, which is now linked to epigenetic remodeling of dorsal root ganglion pathways. Most studies remain single-center and retrospective. AI may enable mechanism-based stratification, individualized neuromodulation programming, and rational integration of regenerative therapies in the treatment of neuropathic pain. Robust prospective validation, transparent reporting, and ethical safeguards are required before approaching translational maturity in selected settings.
13 June 2026
Read appraisal →JMIR medical informatics
Machine Learning for Intraoperative Bleeding Prediction in Patients Undergoing Surgery: Scoping Review
BACKGROUND: Intraoperative bleeding is a critical event that impacts surgical safety and patient outcomes. Machine learning (ML) has demonstrated potential in prediction tasks, yet its methodological rigor and clinical translation face challenges. OBJECTIVE: This scoping review aims to systematically synthesize the current state of development, performance, and validation of ML models for predicting intraoperative bleeding, and to identify key barriers to their clinical implementation. METHODS: Following the PRISMA-ScR (Preferred Reporting Items for Systematic Reviews and Meta-Analyses extension for Scoping Reviews) guidelines, we systematically searched 7 databases (PubMed, Web of Science, Embase, CINAHL, CNKI [China National Knowledge Infrastructure], Wanfang, and VIP [China Science and Technology Journal Database]) from their inception to April 2025. Moreover, 2 reviewers (SY and PZ) independently screened studies, extracted data using the Checklist for Critical Appraisal and Data Extraction for Systematic Reviews of Prediction Modeling Studies (CHARMS), and assessed the risk of bias using the Prediction Model Risk Of Bias Assessment Tool (PROBAST). A narrative synthesis was used for data analysis. RESULTS: Out of 2651 screened records, 23 studies were included (sample sizes ranging from 48 to 48,543). Tree-based ensemble models (eg, random forests and extreme gradient boosting) were the most frequently used (16/23, 70%), followed by logistic regression (13/23, 57%), and deep learning (11/23, 48%). Model discrimination varied widely (mean area under the curve [AUC] 0.82, SD 0.08, range 0.63-0.93). Integration of multimodal data (electronic health records+imaging) was associated with higher performance. However, model validation was often inadequate; only 6 studies (6/23, 26%) performed external validation, and performance often declined (eg, AUC decreased from 0.85 to 0.63 in 1 study). Reporting exhibited selective bias; AUC was commonly reported (19/23, 83%), whereas key classification metrics, such as calibration (10/23, 43%) and precision (4/23, 17%), were often omitted. PROBAST assessment indicated a high risk of bias in all included studies (23/23, 100%). CONCLUSIONS: While ML models demonstrate technical promise for predicting intraoperative bleeding, our PROBAST assessment revealed a universally high risk of bias across all included studies. This fundamental methodological limitation, coupled with a severe lack of external validation and poor transparency in reporting, severely constrains the current clinical reliability of these models. Future research must prioritize prospective multicenter validation, adherence to Transparent Reporting of a Multivariable Prediction Model for Individual Prognosis or Diagnosis (TRIPOD) reporting guidelines, and enhanced model interpretability to bridge the gap toward clinical utility.
13 June 2026
Read appraisal →Annals of medicine
Analgesic effect of premixed nitrous oxide in postoperative rehabilitation for ankle fractures: a randomized controlled trial
INTRODUCTION: The global incidence of ankle fractures is on the rise, effective postoperative rehabilitation is a crucial aspect of surgical management. However, the occurrence of severe pain during the rehabilitation continues to pose a substantial clinical challenge. METHODS: The study utilizes a dual-arm, single-center, double-blind, randomized controlled trial design. A total of 100 participants were enrolled. Participants included patients experiencing acute pain (self-reported pain score ≥4) who underwent postoperative rehabilitation for ankle fractures. Participants undergoing rehabilitation training were randomized to receive either 65% nitrous oxide or 100% oxygen. The primary outcome measured was the pain score, while secondary outcomes encompassed anxiety scores, physiological indices, side effects, patient and therapist satisfaction, acceptance, and residual pain. RESULTS: Pain scores were found to be significantly lower in the nitrous oxide group than in the oxygen group (T1: median difference -3.0, [95% CI -3.4 to -2.6]; p < 0.001; T2: median difference -2.0, [95% CI -2.8 to -1.2]; p < 0.001). Compared with the oxygen group, both therapists (p < 0.001) and participants (p < 0.001) in the nitrous oxide group reported significantly higher satisfaction levels. Acceptance rate showed significant intergroup differences (p < 0.001). No severe adverse effects were observed in either group. Anxiety scores between the two groups did not show a significant difference (p = 0.31, η2 = 0.027). CONCLUSION: The safety and analgesic profile of nitrous oxide renders it a suitable option for pain management in postoperative ankle fractures rehabilitation, thus providing a safe and easily manageable alternative to the current available strategies. CLINICAL TRIAL REGISTRATION: We have registered at https://www.chictr.org.cn and the registration number is: ChiCTR2400089379. In comparison with oxygen gas, nitrous oxide has been demonstrated to reduce pain during rehabilitation training following ankle fracture surgery, offering a practical and effective alternative to oral analgesics.The use of nitrous oxide in postoperative rehabilitation training for ankle fractures is relatively safe, with few and mild adverse reactions.Patients and rehabilitation therapists reported high satisfaction with nitrous oxide intervention, and residual pain was significantly reduced.
12 June 2026
Read appraisal →Medicine
Effects of oxycodone and sufentanil on the quality of recovery after laparoscopic total hysterectomy: A randomized controlled trial
BACKGROUND: To compare the effects of oxycodone and sufentanil on the quality of recovery after laparoscopic total hysterectomy. METHODS: Ninety patients scheduled for laparoscopic total hysterectomy between January and December 2023 were randomly divided into 2 groups: oxycodone group (Group O) and sufentanil group (Group S), with 45 patients in each group. The primary outcome measure was postoperative recovery status recorded using the Quality of Recovery-40 questionnaire. RESULTS: At 24 hours postoperatively, the Quality of Recovery-40 scores of group O and group S were comparable (170 [156-179], 160 [146-178], P = .111). There were no statistically significant differences between the 2 groups in terms of visual analogue scale pain scores at any postoperative time points, postoperative cumulative sufentanil consumption, length of hospital stay, patient satisfaction, and incidence of postoperative nausea and vomiting, and incidence of postoperative dizziness (P > .05). CONCLUSIONS: Oxycodone and sufentanil provide comparable postoperative recovery after total laparoscopic hysterectomy.
1 June 2026
Read appraisal →Medicine
Common biomarkers of Alzheimer disease and postoperative delirium
Alzheimer disease (AD) and Postoperative delirium (POD) may share a common mechanism, but their shared genes and potential novel therapeutic targets remains unknown. We selected datasets GSE48350, GSE4226, and GSE16759 containing AD and control samples from the GEO database for AD model development, and use the GSE63061 dataset for AD model validation. The GSE163943 dataset containing PODs and controls was used to identify common genes. Differentially expressed genes (DEGs) associated with POD and AD respectively, as well as common DEGs between 2 diseases were analyzed. Shared DEGs were assessed using gene ontology and KEGG enrichment and PPI interaction networks. Machine learning algorithms, nomogram and decision tree analyses were used to interpret top 5 genes and visualize the models. 259 DEGs in POD with 148 being up-regulated and 111 down-regulated, and 199 DEGs in AD with 25 up-regulated and 174 down-regulated were found, of which 23 were common to POD and AD. Gene ontology analysis indicated that DEGs were primarily concentrated in 9 distinct biological processes, while KEGG results revealed 3 major metabolic pathways. Five hub genes emerged as potential diagnostic biomarkers for both diseases, Bruton tyrosine kinase (BTK), NCF2, CRH, FCGR3A, and SERPINA3. The LASSO model determined that the 23 common genes could serve as biomarkers for both POD and AD patients, and the Random Forest model is the the most suitable model. POD or AD decision trees showed BTK and NCF2 were sufficient to distinguish between POD or AD patients and healthy controls. BTK and NCF2 were identified common biomarkers of POD and AD. Further investigations are required to verify correlations between gene expression levels and pathological characteristics.
31 May 2026
Read appraisal →The journal of ECT
Comparison of the Effects of Dexmedetomidine and Remifentanil on Seizure Duration, Hemodynamics, and Recovery Time in Electroconvulsive Therapy
OBJECTIVES: Electroconvulsive therapy (ECT) is an important treatment method that is widely used in psychiatric conditions resistant to pharmacological treatments. METHODS: This prospective, randomized study included 100 patients undergoing ECT. Group D received dexmedetomidine and propofol, while group R received remifentanil and propofol. In this study, key parameters such as seizure duration, hemodynamic changes, and recovery time were analyzed. RESULTS: The investigation revealed no statistically significant differences between the groups with regard to mean age, gender distribution, or seizure duration ( P > 0.05). Subsequent to induction, group R exhibited statistically significantly lower systolic blood pressure, diastolic blood pressure, and mean arterial pressures in comparison to group D ( P < 0.05). Conversely, heart rate, systolic blood pressure, diastolic blood pressure, and mean arterial pressure levels in group R were found to be statistically significantly higher than in group D ( P < 0.05). The recovery time in group D was found to be statistically significantly longer than in group R ( P < 0.05). CONCLUSIONS: Dexmedetomidine provides superior hemodynamic stability, while remifentanil ensures faster recovery. Dexmedetomidine has been shown to be more advantageous in terms of hemodynamic stability, while remifentanil has been demonstrated to result in shorter recovery times. Clinicians should tailor anesthetic choices based on patient profiles and therapeutic goals.
30 May 2026
Read appraisal →BMJ open
Does a protocolised oxygenation target impact clinical outcomes in mechanically ventilated patients in the intensive care unit across the 10 academic and community hospitals in the UCHealth system? Protocol and statistical analysis plan for an EHR-embedded pragmatic stepped-wedge randomised controlled trial.
INTRODUCTION: We designed the Optimising Care in Critically Ill at the UCHealth by Liberalising the Target O₂ in Mechanically ventilated intensive care unit (ICU) Patients to determine the effectiveness of a multimodal educational and electronic health record order panel intervention in limiting occult hypoxaemia and hyperoxaemia in patients receiving invasive mechanical ventilation by targeting a prespecified oxygen saturation (SpO₂ 90%-96%) range. METHODS AND ANALYSIS: This trial is a pragmatic electronic health record-embedded, multisite, cluster-randomised, stepped-wedge implementation of a multimodal educational and electronic health record order panel intervention aimed at achieving a standardised intermediate target range of SpO₂ (90%-96%) or partial pressure of oxygen (PaO₂ 60-100 mm Hg) in mechanically ventilated adult patients admitted to the ICU across a 10 hospital health system that includes a large academic centre and smaller community hospitals.The primary endpoint is ventilator-free days to day 30, defined as the number of days alive and not receiving support through invasive mechanical ventilation following the first initiation of invasive mechanical ventilation in a participating ICU during the hospitalisation. Secondary outcomes include a variety of clinical endpoints: hospital-free days to day 30, all-cause mortality to day 90, need for supplemental oxygen at discharge and incidence of occult hypoxaemia and hypoxaemia. We will analyse primary and secondary endpoints using a mixed effects modelling framework, with specific distributions chosen depending on the type of outcome (eg, binary, count, ordinal, time-to-event). ETHICS AND DISSEMINATION: Research is performed under a waiver of informed consent for minimal-risk research, as approved by Colorado Multiple Institutional Review Board (24-0065). Results will be disseminated in peer-reviewed publications and at national and international conferences. TRIAL REGISTRATION NUMBER: NCT06501118.
28 May 2026
Read appraisal →Scientific reports
Impact of a structured case report on self-reported responses to simulated emergency scenarios: a randomized survey-based study
Case reports are a long-standing component of medical research, but their association with self-reported responses to simulated clinical scenarios remains underexplored. This study investigated whether exposure to a structured case report influences simulated emergency medical decisions. This randomized survey-based study was conducted among licensed physicians between March and May 2025. Participants were randomly assigned to an intervention group, which received a structured case report on blunt neck trauma, or a control group, which did not. Both groups subsequently provided self-reported Likert-scale responses to identical simulated emergency scenarios, covering five domains: local trauma consequences, accompanying injuries, diagnostic methods, airway management strategies, and hospital selection. Responses were collected on Likert scales. Group differences were assessed using Mann-Whitney U tests, t-tests, and chi-squared tests; prespecified exploratory subgroup analyses examined the moderating role of clinical experience. Sixty-three participants completed the study. Baseline demographics and qualifications did not differ significantly between groups. Exposure to the case report was associated with statistically significant differences in self-reported responses to simulated scenarios, with the most pronounced associations in airway management strategies (all five items statistically significant, median effect size r = 0.44). Intervention participants assigned higher ratings to maintaining spontaneous breathing and supraglottic airway use, while assigning lower appropriateness to endotracheal intubation and cricothyrotomy. Moderate associations were observed in diagnostic choices (e.g., pathological breath sounds, ultrasound use) and hospital selection (e.g., ECMO availability). No consistent effects were found for local or accompanying injuries. Exploratory stratified analyses revealed that junior physicians (≤ 5 years of practice) appeared more responsive, with significant differences across multiple domains and larger effect sizes (median r > 0.6). In contrast, no statistically significant between-group differences were detected in the senior (> 15 years) subgroup, although this exploratory analysis was limited by small subgroup size. Structured case reports may influence self-reported responses to simulated emergency scenarios, particularly among less experienced clinicians. These findings suggest that the educational impact of case-based learning formats may vary depending on the level of clinical experience, but do not allow conclusions about real-world clinical behavior or patient outcomes. Future work should further investigate adaptive case report formats and their potential integration into emergency medicine training.Trial registration: German Clinical Trials Register (DRKS00038961), registered von 13 January 2026.
27 May 2026
Read appraisal →JMIR research protocols
AI-Supported, Integrative Prediction of Postoperative Delirium: Protocol for the CONFUSED Study
BACKGROUND: Postoperative delirium (POD) is a frequent and serious complication in older surgical patients, characterized by acute cognitive dysfunction and fluctuating levels of consciousness. POD is associated with prolonged hospitalization, long-term cognitive decline, reduced quality of life, and increased mortality. Despite its clinical relevance, the underlying pathophysiological mechanisms remain poorly understood, and reliable biomarkers for early prediction and prevention are lacking. OBJECTIVE: The CONFUSED study aims to identify molecular and clinical predictors of POD by integrating clinical data with proteomic, transcriptomic, and epigenetic analyses. The primary objective is to develop predictive models for POD using multimodal data. Secondary objectives include the identification of delirium-associated genes, proteins, and epigenetic signatures, as well as the exploration of patient subgroups at increased risk for POD. METHODS: CONFUSED is a prospective observational cohort study conducted at a German university hospital. Adult patients undergoing major surgery under general anesthesia will be enrolled until 100 cases of POD have been observed, which is expected to require a total sample size of approximately 200 to 300 patients. Blood samples are collected at 4 predefined time points: before premedication, immediately after surgery, and on postoperative days 2 and 5. Samples undergo comprehensive proteomic profiling, transcriptomic analysis using RNA microarrays, DNA methylation analysis, and genotyping of selected polymorphisms. Clinical data, including demographics, comorbidities, perioperative variables, medications, and delirium assessments using the Confusion Assessment Method (CAM) and CAM for the intensive care unit, are systematically recorded. Statistical analyses include univariate and multivariate methods, as well as machine learning approaches such as random forests and support vector machines, to identify relevant biomarkers and develop predictive models. The study protocol follows STROBE (Strengthening the Reporting of Observational Studies in Epidemiology) and TRIPOD (Transparent Reporting of a Multivariable Prediction Model for Individual Prognosis or Diagnosis) guidelines and was approved by the responsible ethics committees. RESULTS: The study was registered in the German Clinical Trials Register (DRKS00033854) on March 18, 2024. Recruitment started in January 2024 and is ongoing at the time of manuscript submission. As of now, 135 patients have been enrolled. Sample collection and laboratory analyses are ongoing. Data analysis began in January 2026, with first results anticipated in July 2026. Final data lock is anticipated after the completion of recruitment. CONCLUSIONS: By integrating multimodal molecular data with clinical parameters and applying advanced machine learning techniques, the CONFUSED study aims to improve the prediction and understanding of POD. The results are expected to support the development of personalized preventive strategies and contribute to improved perioperative care for patients at risk of POD.
26 May 2026
Read appraisal →Otolaryngologia polska = The Polish otolaryngology
Leveraging large language models to audit perioperative total laryngectomy care: anesthetic, nutritional, and complication profiles in a real‑world cohort
<p><strong>Introduction: </strong>Total laryngectomy (TL) remains the standard treatment for advanced la‑ ryngeal cancer, and its perioperative management requires close collaboration within a multidisciplinary team, including anaesthesiologists and clinical nutritionists.</p><p><strong>Aim:</strong> The aim of this study was to analyse the perioperative course of patients un‑ dergoing TL from both an anaesthesiologist's and a nutritionist's perspective, with a particular focus on the feasibility and performance of structured data extraction from electronic medical records using large language models (LLMs).</p><p><strong>Material and methods:</strong> We performed a single‑centre retrospective analysis of 43 consecutive patients undergoing TL (ICD‑9 codes 30.32, 30.39, 30.41, 30.49) be‑ tween January 2024 and April 2026. Clinical data, including physician notes (n = 845), nursing notes (n = 2578), discharge summaries (n = 173), and consultations (n = 123), were subjected to structured extraction using large language models (gemma-3-31Bit/gemma-4-26B-it). Demographics, tumour stage, length of stay (LOS), periopera‑ tive nutrition, pharmacotherapy, postoperative complications, temporal laboratory trends, and rehabilitation were analysed.</p><p><strong>Results: </strong>The cohort comprised 35 men (81%) and 8 women (19%), with a mean age of 66.4 9.8 years; pT4a (49%) and pN0 (82%) predominated. Median LOS was 15.0 days (IQR 13.1-16.5). Enteral nutrition via nasogastric tube was used in 95% of patients, and oral intake was initiated at a median of postoperative day 13. Multimodal anal‑ gesia based on intravenous paracetamol (60%), ketoprofen (47%), and metamizole (40%) was routinely employed, while strong opioids were used in only 5% of pa‑ tients. Postoperative complications occurred in 22.5% of cases, with pharyngocuta‑ neous fistula (PCF) being the most frequent (12.5%). In‑hospital mortality was 0%, and overall mortality during a median 12‑month follow‑up was 2.3%. CRP peaked on postoperative day 3 (mean 114 mg/L). Pharmacologic VTE prophylaxis was admi‑ nistered in 65% of patients.</p><p><strong>Conclusions:</strong> In TL patients, enteral feeding forms the backbone of nutritional sup‑port, and multimodal analgesia provides effective pain control, whereas VTE pro‑phylaxis and other modifiable elements of perioperative care warrant optimization within prehabilitation and ERAS-based protocols.</p><p>&nbsp;</p>.
26 May 2026
Read appraisal →The journal of trauma and acute care surgery
Chilling the nerve, easing the pain?: A randomized clinical trial evaluating surgeon-administered bedside percutaneous cryoneurolysis for rib fracture pain.
BACKGROUND: A cornerstone of rib fracture management is multimodal pain control, which includes scheduled nonopioid analgesics, as-needed opioids, regional or neuraxial blockade, and surgical stabilization of rib fractures. However, adverse effect profiles and individual factors limit their use in all patients. Surgeon-administered, ultrasound-guided percutaneous cryoneurolysis performed at the patient bedside is a promising analgesic adjunct. METHODS: We performed a prospective, randomized clinical trial assessing benefit and safety of surgeon-administered percutaneous cryoneurolysis to our existing multimodal rib fracture pain control bundle (standard of care [SoC]) for injured adults aged 18 to 64 years. Patients with fractured ribs 3 to 9 were randomized within 72 hours of admission to receive either surgeon-administered, ultrasound-guided percutaneous cryoneurolysis at the bedside and our multimodal pain control bundle, or SoC alone. Patients undergoing surgical stabilization of rib fractures were excluded. The primary outcome was pain score at discharge. Secondary outcomes included hospital length of stay, intervention-associated adverse events, morphine milligram equivalent (MME) use, Short Form (SF)-12, and McGill Pain Score (Patient Reported Outcomes), which were assessed at discharge and 1-, 3-, and 12-month intervals after discharge. RESULTS: Forty-three patients were randomized: 24 in the intervention arm and 19 in the SoC arm, with 3 screen failures. The median age was 53 years (interquartile range [IQR], 45-60 years), 9 patients (23%) were female, the median Injury Severity Score was 13 (IQR, 10-17), and the median number of rib fractures was 6 (IQR, 4-8). Patients were well matched with no differences between groups. No intervention-associated adverse events were identified. Pain scores, MME use, and Patient Reported Outcomes were not different between groups at discharge, 1 month, 3 months, or 12 months. CONCLUSION: Application of surgeon-administered, ultrasound-guided percutaneous cryoneurolysis at the bedside is safe but was not associated with reduced pain or MME use when compared with SoC. ( J Trauma Acute Care Surg . 2026;100: 863-870. LEVEL OF EVIDENCE: Therapeutic/Care Management; Level II.
23 May 2026
Read appraisal →Neurosurgical review
The safety and efficacy of perioperative nefopam for analgesia in spine surgery: A systematic review and meta-analysis of randomised controlled trials
OBJECTIVE: Moderate-to-severe pain is common after spine surgery. While opioids remain standard analgesia, their adverse effects have prompted opioid-sparing strategies. This systematic review evaluates the safety and efficacy of nefopam, a non-opioid and non-NSAID, as adjuvant perioperative analgesia in spine surgery. METHODS: Medline, Embase, Scopus and Cochrane databases were searched from inception to May 9, 2025. RevMan V5.4 was used to analyse mean differences (MD) or standardised mean differences (SMD) and 95% confidence intervals (CIs) for continuous outcomes and risk ratios (RR) with 95% CIs for dichotomous outcomes. The GRADE framework was used to assess the certainty of evidence. RESULTS: Seven randomised controlled trials (n = 553) were included. Nefopam did not reduce 24-hour morphine use (MD -1.47 mg; 95% CI: -3.86 to 0.92 mg; GRADE moderate), 24-hour incisional pain (SMD, -0.01; 95% CI -0.22 to 0.20; GRADE moderate), 48-hour incisional pain (SMD, -0.24; 95% CI -0.50 to 0.02; GRADE low), 24-hour neuropathic pain (SMD, -0.19; 95% CI -0.49 to 0.10; GRADE low), 72-hour neuropathic pain (SMD, -0.33; 95% CI -1.02 to 0.36; GRADE very low), or length of stay (MD, 0.08days; 95% CI -0.73 to 0.88days; GRADE Low). There were no significant differences in drug-related adverse effects (sedation (RR 1.30), nausea/vomiting (RR 1.03), dizziness (RR 1.14), and urinary retention (RR 2.36)), (GRADE: very low to low). CONCLUSION: Perioperative nefopam did not reduce acute morphine use, pain scores, hospital stay, or drug-related adverse events in spine surgery, though most studies used doses below the median effective dose, limiting conclusions on its efficacy and safety.
23 May 2026
Read appraisal →BMJ evidence-based medicine
Tramadol versus placebo for chronic pain: a systematic review with meta-analysis and trial sequential analysis
OBJECTIVES: The objective of our study was to assess the benefits and harms of tramadol vs placebo in adults with chronic pain. DESIGN: The research method was a systematic review of randomised clinical trials with meta-analysis. The review followed the Trial Sequential Analysis and the Grading of Recommendations Assessment, Development and Evaluation (GRADE) approach. DATA SOURCES: The Cochrane Library, MEDLINE, Embase, Science Citation Index and BIOSIS were searched for trials published from inception to 6 February 2025. ELIGIBILITY CRITERIA FOR SELECTING STUDIES: Studies were eligible for inclusion if they were published and unpublished randomised clinical trials comparing tramadol vs placebo in adults with any type of chronic pain. Risk of bias was assessed according to the Cochrane Handbook for Systematic Reviews of Interventions. MAIN OUTCOME MEASURES: The main outcome measures were pain level, adverse events, quality of life, dependence, abuse and depressive symptoms. RESULTS: We included 19 randomised placebo-controlled clinical trials enrolling 6506 participants. All outcome results were at high risk of bias. Meta-analysis and Trial Sequential Analysis showed evidence of a beneficial effect of tramadol on chronic pain (mean difference numerical rating scale (NRS) -0.93 points; 97.5% CI -1.26 to -0.60; p<0.0001; low certainty of evidence). However, the effect size was below our predefined minimal important difference of 1.0 point on NRS. Beta binomial regression showed evidence of a harmful effect of tramadol on serious adverse events (OR 2.13; 97.5% CI 1.29 to 3.51; p=0.001; moderate certainty of evidence), mainly driven by a higher proportion of cardiac events and neoplasms. It was not possible to conduct a meta-analysis of the quality of life due to a lack of data. Meta-analysis and Trial Sequential Analysis showed that tramadol increased the risk of several non-serious adverse events including nausea (number needed to harm (NNH) 7), dizziness (NNH 8), constipation (NNH 9), and somnolence (NNH 13) (all very low certainty of evidence). CONCLUSION: Tramadol may have a slight effect on reducing chronic pain levels (low certainty of evidence) while likely increasing the risk of both serious (moderate certainty of evidence) and non-serious adverse events (very low certainty of evidence). The potential harms associated with tramadol use for pain management likely outweigh its limited benefits.
23 May 2026
Read appraisal →Renal failure
Machine learning-based risk prediction model for postoperative acute kidney injury in surgical patients with chronic kidney disease (CKD): development, validation, and SHAP-based explainability
This study aimed to develop machine learning models to predict postoperative acute kidney injury (AKI) in surgical patients with pre-existing chronic kidney disease (CKD), a serious complication associated with high morbidity and accelerated disease progression. Using data from 3,851 surgical patients with CKD from the MIMIC-IV database, nine supervised machine learning models were developed to predict AKI, as defined by the KDIGO criteria. Postoperative AKI occurred in 24.2% (932/3,851) of the patients. The XGBoost model demonstrated the best predictive performance, achieving an AUC of 0.844, accuracy of 0.791, sensitivity of 0.79, and F1-score of 0.73, followed by Random Forest and LightGBM. The calibration curve showed a good agreement between the predicted and observed risks. For interpretability, SHAP analysis was used to identify SOFA score, estimated glomerular filtration rate (eGFR), systolic blood pressure, albumin, and phosphorus as the most influential predictors. In conclusion, machine learning models, particularly XGBoost, provide accurate and interpretable predictions of postoperative AKI in this high-risk population. The integration of explainable AI facilitates transparent risk stratification, which may support personalized perioperative care and the development of real-time clinical decision support systems.
21 May 2026
Read appraisal →Journal of investigative surgery : the official journal of the Academy of Surgical Research
Wrist-Ankle Acupuncture on Postoperative Nausea and Vomiting Prophylaxis in High-Risk Female Patients: A Pragmatic, Randomized, Single-Blind, Sham-Controlled Trial
BACKGROUND: Postoperative nausea and vomiting (PONV) affects up to 80% of high-risk patients, compromising Enhanced Recovery After Surgery (ERAS) outcomes and prolonging hospitalization. This randomized controlled trial evaluated wrist-ankle acupuncture (WAA) efficacy for PONV prevention in high-risk female surgical patients. METHODS: This single-center, randomized, patient- and assessor-blinded, sham-controlled trial enrolled female patients (aged 18-60 years, ASA I-II) scheduled for elective surgery under general anesthesia with documented motion sickness or previous PONV. Participants received either WAA (sterile needles at four standardized acupoints, 30-minute retention) or sham acupuncture postoperatively. The primary outcome was 24-h PONV incidence using the Rhodes Index of Nausea, Vomiting, and Retching (R-INVR). RESULTS: Among 132 patients (median age 41 years), WAA significantly reduced 24-h PONV incidence versus sham (42.4% vs. 71.2%; RR 0.60, 95% CI 0.42-0.82; p = 0.002). Median R-INVR scores were lower with WAA [0 (0-5) vs. 8 (0-11); p < 0.001]. WAA patients reported significantly reduced nausea/vomiting at 30 min, 6 h, and 24 h (all p ≤ 0.006) and higher satisfaction scores (p < 0.001). Antiemetic use was comparable (9.1% vs. 16.7%; p = 0.299). No adverse events occurred. CONCLUSIONS: WAA significantly reduced PONV incidence and severity with sustained effects and improved satisfaction, offering a promising non-pharmacological prevention strategy.
21 May 2026
Read appraisal →Physiological measurement
Perioperative hypoxaemia early warning based on high-frequency waveform fusion and deep learning
Objective.The accurate prediction of impending intraoperative hypoxaemia events is paramount for patient safety. Current models relying on structural parameters exhibit limitations, including low sampling frequency, weak temporal coherence, and high data attrition rates. Consequently, long-term sequence, high-frequency waveform data are utilised in this study to construct a deep learning-based hypoxaemia prediction model, by which conventional shortcomings are addressed whilst predictive accuracy and robustness are enhanced.Approach.Respiratory, blood oxygenation, and electrocardiogram waveforms were extracted from the VitalDB database. Two processing methods were employed: extraction of waveform-derived variables to construct numerical datasets; and utilisation of the Gramian Angular Summation Field (GASF) and the four-quadrant matrix method to construct image-based datasets. Through multiple dataset construction methods, five datasets were constructed. An LSTM algorithm was employed, using a 5 min learning window and prediction windows ranging from 1 to 13 min, to conduct a comprehensive performance evaluation.Main results.Optimal performance was achieved by all models within a one-minute prediction window. Highly competitive accuracy (0.932), AUC (0.953), and stability (maximum AUC variance 0.067) were demonstrated by the 'Multimodal image model', whilst the poorest performance was exhibited by the 'structured only model'. Moreover, compared to singlemodal models, multimodal models incorporating structured parameters yield only marginal improvements. Consequently, singlemodal data derived solely from raw waveforms are suggested as sufficient for prediction requirements, by which greater application potential is held due to enhanced computational efficiency.Significance.This study systematically demonstrated the significant advantages of high-frequency waveform data. Results indicate that waveform processing methods based on the GASF and the four-quadrant matrix approach can effectively enhance model stability and performance. Furthermore, the singlemodal model relying solely on waveform data, excluding structured parameters, can meet clinical prediction requirements while reducing computational costs. This study provides novel data structures and methodologies for perioperative hypoxaemia prediction.
20 May 2026
Read appraisal →PloS one
The carbon footprint of surgical operations: 2023-2025 systematic review update
Surgical care significantly contributes to greenhouse gas emissions, with operating rooms consuming three to six times more energy than other departments relying heavily on single-use materials and anaesthetic gases. We conducted a systematic review of literature from 2023 to 2025 following PRISMA guidelines. The PubMed, Scopus and Healthcare LCA databases were searched, and eligibility criteria were applied. A total of 26 studies met inclusion criteria. Data were extracted on each study's scope and system boundaries, reported emission values, identified carbon hotspots, and any mitigation strategies evaluated. A structured quality assessment was conducted to evaluate the reliability of the included studies and to identify potential sources of bias. The included studies demonstrated wide variability in reported carbon footprints of surgery, ranging from under 5 kgCO₂e for minor procedures to over over 400 kgCO₂e for complex single-stage operations and approximately 1,000 kgCO₂e for whole multistage patient pathways. This systematic review underscores that surgical operations could have a significant carbon footprint, with emissions hotspots concentrated in consumable materials and energy consumption. It also reveals substantial variability and methodological heterogeneity in how surgical carbon footprints are calculated, pointing to the urgent need for standardized life-cycle based frameworks in surgical settings. Establishing common standards will enable more reliable benchmarking of surgical emissions and better comparisons across studies.
19 May 2026
Read appraisal →Annals of medicine
Impact on analgesia, diaphragmatic function, and recovery between erector spinae plane block versus superior trunk block in arthroscopic shoulder surgery: a randomized controlled trial
BACKGROUND: Effective analgesia and preservation of diaphragmatic function are key considerations in analgesia for shoulder surgery. The superior trunk block provides analgesia with reduced phrenic nerve involvement, while the erector spinae plane block offers minimal impact on diaphragm motion. This randomized controlled trial compared the analgesic efficacy, impact on diaphragmatic motion, and postoperative recovery between the two blocks. METHODS: Sixty patients undergoing arthroscopic shoulder surgery were randomized to receive either erector spinae plane block or superior trunk block. Primary outcomes were postoperative VAS and changes in diaphragmatic excursion. Secondary outcomes included Quality of Recovery-15 (QoR-15) scores, morphine-equivalent consumption, and the handgrip strength motor blockade. RESULTS: The superior trunk block resulted in significantly lower dynamic VAS at 1-h postoperatively (0.1 [0.0, 0.2] vs. 5.7 [4.0, 7.6]; p < 0.001) and reduced 24-h morphine consumption (7.8 [2.5, 15.0] mg vs. 12.7 [7.5, 17.3] mg; p = 0.038) compared to the erector spinae plane block. However, diaphragmatic excursion was better preserved in the erector spinae plane block group (8.37% ± 20.7% vs. -20.09% ± 22.2%; p < 0.001), with a lower incidence of partial hemidiaphragm paresis (3.3% vs. 46.7%; p < 0.001). At 24 h postoperatively, QoR-15 scores were higher in the superior trunk block group (p = 0.047), and no patient in either group developed handgrip motor blockade. CONCLUSIONS: Superior trunk block offers superior early postoperative analgesia and better overall recovery, while erector spinae plane block minimizes diaphragmatic impairment. However, the erector spinae plane block may represent an option only in carefully selected patients at high respiratory risk, acknowledging its significantly poorer early analgesic profile.
15 May 2026
Read appraisal →Journal of investigative surgery : the official journal of the Academy of Surgical Research
Meta-Analysis of PENG+LFCN versus FICB on ERAS-Related Outcomes: Postoperative Pain, Motor Function and Recovery in Hip Surgery
OBJECTIVE: To compare pericapsular nerve group block combined with lateral femoral cutaneous nerve block (PENG+LFCN) with fascia iliaca compartment block (FICB) for postoperative analgesia and functional recovery after hip surgery within Enhanced Recovery After Surgery (ERAS) pathways. METHODS: PubMed, Embase, the Cochrane Library, CNKI, Wanfang Data, and CQVIP were searched through January 2026. Comparative studies of PENG+LFCN versus FICB in hip surgery were included. Data were pooled using a random-effects model, with assessment of bias, heterogeneity, and publication bias. RESULTS: Sixteen studies were included. The PENG+LFCN group had significantly lower resting VAS (MD = -0.54; p < 0.00001) and dynamic VAS (MD = -0.98; p < 0.00001) scores. It also showed a lower incidence of postoperative muscle weakness (OR = 0.10; p = 0.0001), better preservation of quadriceps muscle strength, shorter time to first ambulation (MD = -9.22 h; p < 0.0001), and reduced length of stay (MD = -0.75 days; p < 0.00001). Additionally, morphine consumption (MD = -24.22 mg; p = 0.001) and PCA pump activations (MD = -2.46; p < 0.00001) were lower in the PENG+LFCN group. CONCLUSION: PENG+LFCN may be a more favorable regional analgesic strategy than FICB for hip surgery within ERAS pathways. PROTOCOL REGISTRATION: This systematic review and meta-analysis was registered at PROSPERO https://www.crd.york.ac.uk/PROSPERO/view/CRD420261295309.
13 May 2026
Read appraisal →Psychological medicine
Cognitive effects of intravenous Ketamine in treatment-resistant depression: A systematic review
Intravenous (IV) low-dose ketamine has emerged as a promising treatment for patients with treatment-resistant depression (TRD). However, its impact on cognitive functioning remains unclear. This systematic review examines the cognitive and executive effects of IV ketamine in TRD, focusing on their relationship to depressive and suicidal outcomes. A systematic search of Cochrane, MEDLINE, Embase, and PsycINFO databases was conducted up to May 15, 2025, using the terms depression, cognition, and ketamine. This review was conducted in accordance with the PRISMA guidelines, and the protocol was registered in PROSPERO (ID: 1160487). Risk of bias was evaluated using the Cochrane RoB 2 tool for randomized trials and the ROBINS-I tool for non-randomized studies. Twenty-one studies, comprising approximately 900-1,180 participants with TRD, assessed cognitive domains of processing speed, working memory, attention, verbal and visual memory, cognitive flexibility, and executive control. Procognitive effects were frequently observed in processing speed and working memory, while attention results were preserved or modestly improved, and verbal and visual memory results were heterogeneous. Executive control, particularly inhibitory performance on Stroop paradigms, improved in several trials. Two studies directly examined cognition as it relates to suicidal behaviors. No cognitive deterioration was reported. Subanesthetic IV ketamine appears to preserve and enhance specific cognitive functions in TRD, notably across processing speed, working memory, and executive control. These procognitive effects, particularly in executive control, may mediate ketamine's antisuicidal action. Standardized longitudinal studies are warranted to clarify their durability and clinical significance.
12 May 2026
Read appraisal →Scandinavian journal of pain
Breathing therapy as a supplement to opioid management for complex chronic non-cancer pain conditions - a systematic review
OBJECTIVES: The purpose of this systematic review was to analyze whether breathing therapy is effective in reducing pain, improving health-related quality of life, improving physical functioning/activity, and improving sleep quality in patients suffering from complex chronic non-cancer pain conditions (CNCP). CONTENT: An electronic literature search was conducted in MEDLINE, CINAHL, EMBASE, Cochrane Library, PEDro, and PsycINFO from inception to October 2024. The inclusion criteria were randomized clinical trials (RCT) examining breathing therapy as a sole or central intervention component in adult patients with CNCP. From 2,369 abstracts, a total of 10 RCTs (n=638) met the inclusion criteria. The population in the selected studies were patients with chronic neck pain, chronic low back pain, fibromyalgia, or tension-typed headache. Breathing therapy significantly improved pain and/or health-related quality of life in six studies. However, the quality of the included RCTs ranged from high to critically low, and substantial heterogeneity in participants, intervention, and methods prevented synthesis of results across studies. SUMMARY: This systematic review highlights breathing therapy as a promising pain management strategy in patients suffering from CNCP. We evaluated two RCTs examining similar daily slow diaphragmatic breathing for 15 min to have a low risk of bias; both demonstrating statistically significant relevant pain reduction up to 37 %. However, the strength of recommendations for clinical practice depends on the level of evidence as indicated by a risk of bias assessment (internal validity), consistency of results between studies, and generalizability (external validity). Based on the findings of this systematic review evidence of breathing therapy invention in patients suffering from CNCP conditions remains sparse. OUTLOOK: Striving to eliminate or minimize opioid management for complex CNCP, high-quality research is needed to reinforce the evidence base for non-pharmacological interventions such as breathing therapy to support modern pain management rather than former traditional pharmacological treatment. Trial registry number: PROSPERO #CRD42023460181.
12 May 2026
Read appraisal →Molecular biology reports
Genetic modulation of pain pathways: toward a new era in pain therapy- a systematic review
Gene therapy represents a new strategy in pain management targeting the causes of pain rather than its symptoms. This review summarizes recent advances across RNA interference (RNAi), viral vector delivery systems, CRISPR/Cas9, gene replacement therapy, and endogenous opioid gene delivery, emphasizing efficacy, safety, and mechanisms of action in pain conditions. RNAi and gene replacement techniques remain powerful tools for reducing pain and improving the quality of life by modulating pain-associated genes. Long-term relief may also be achieved with CRISPR/Cas9 and site-directed delivery using nanoparticle systems. Safety profiles especially with CRISPR/Cas9 remains a concern. A systematic search of Web of Science, Medline, Scopus, and Google Scholar identified 512 records (January 2010-March 2023). Following PRISMA screening, 18 studies met inclusion criteria. These studies evaluated strategies of gene therapy in neuropathy, arthritis, fibromyalgia, and complex regional pain syndrome. The review outlines available options and highlights preclinical findings.
11 May 2026
Read appraisal →Artificial intelligence in medicine
Size doesn't matter: Assessing the trustworthiness of large language models in medical contexts: A focus on epidural information retrieval.
BACKGROUND: Since the release of ChatGPT, numerous LLMs have emerged, providing easy access to information without the need for technical expertise. However, relying on these systems can influence important life decisions, such as the choice to use epidural analgesia during childbirth. Epidural analgesia is widely regarded as the "gold standard" for pain relief during childbirth. However, limited access to anaesthesiologists and gaps in knowledge may prompt individuals to seek information from unverified sources, including AI systems. Misinformation in this area can discourage the use of effective analgesia, highlighting the need to assess the accuracy of LLM-generated content. OBJECTIVE: To evaluate the reliability of LLM-generated information regarding epidural analgesia. METHODS: We posed 10 standardized questions about epidural analgesia to 12 LLMs, each question reformulated 10 times in both Spanish and English, resulting in 2400 responses. Two anaesthesiologists were involved in the assessment process. One expert performed the initial ratings, while the second independently verified the evaluations assessing the outputs using an extended SERVQUAL framework. RESULTS: ChatGPT performed best, followed by Gemini 2. Medium-sized models, such as Phi-3 and OpenChat, outperformed several larger models like Llama-2 or Llama-3, challenging the notion that "bigger is better" and offering potential advantages in low-resource settings (e.g., Phi-3 outperformed Llama-2 with an average increase of 81% across all metrics). Specialized models did not show superior performance. Except for ChatGPT, English responses were generally more reliable than Spanish, with some Spanish outputs incoherent. ChatGPT also exhibited the least variability between responses. CONCLUSIONS: Despite promising performance, LLMs display limitations in medical contexts. Collaboration between national and international medical societies is crucial to develop evidence-based resources to guide LLM training and improve information trustworthiness.
10 May 2026
Read appraisal →Biomedical physics & engineering express
A novel machine learning approach for prediction of postoperative delirium using multi-domain features of high-density EEG
Background.Postoperative delirium (POD) poses a significant risk to patients, and accurate prediction of POD can provide guidance for positive interventions. Although many studies have applied machine learning (ML) to electronic health records to predict POD, there has been a lack of studies utilizing electroencephalogram (EEG) data to accurately predict POD.Methods.This retrospective study included 142 patients from two hospitals, among whom 33 developed POD. We extracted multi-domain features from high-density EEG (HD-EEG), screened 39 key features using recursive feature elimination, and analyzed the feature space structure using linear discriminant analysis (LDA). Four ML models, namely support vector machine (SVM), logistic regression (LR), random forest (RF), and gradient boosting machine, were compared. The dataset was divided using 5-fold cross-validation for model training and testing.Results.Compared to the other three models, SVM (linear kernel) performed best in predicting POD, achieving a classification accuracy of 95.71%. Alpha (8-12 Hz) and theta (4-8 Hz) powers, combined with nonlinear dynamics, critically contributed to the model. Furthermore, the performance of the linear models (SVM and LR) was significantly (p< 0.001) better than that of the nonlinear models, with LDA confirming strong feature space separability (Jensen-Shannon divergence = 0.69).Conclusions.HD-EEG data can be used to establish high-quality ML models for POD predictions. Multidomain features enable more comprehensive integration of EEG information related to delirium. The feature space structure significantly affects the performance of the prediction model, with the linear model offering distinct advantages in predicting POD.
2 May 2026
Read appraisal →The New England journal of medicine
Ketamine or Etomidate for Tracheal Intubation of Critically Ill Adults
BACKGROUND: For critically ill adults undergoing tracheal intubation, observational studies suggest that the use of etomidate to induce anesthesia may increase the risk of death. Whether the use of ketamine rather than etomidate decreases the risk of death is uncertain. METHODS: In a randomized trial conducted in 14 emergency departments and intensive care units in the United States, we randomly assigned critically ill adults who were undergoing tracheal intubation to receive ketamine or etomidate for the induction of anesthesia. The primary outcome was in-hospital death from any cause by day 28. The secondary outcome was cardiovascular collapse during intubation, defined by the occurrence of a systolic blood pressure below 65 mm Hg, receipt of a new or increased dose of vasopressors, or cardiac arrest. RESULTS: A total of 2365 patients underwent randomization and were included in the trial population; 1176 were assigned to the ketamine group and 1189 to the etomidate group. In-hospital death by day 28 occurred in 330 of 1173 patients (28.1%) in the ketamine group and in 345 of 1186 patients (29.1%) in the etomidate group (risk difference adjusted for trial site, -0.8 percentage points; 95% confidence interval [CI], -4.5 to 2.9; P = 0.65). Cardiovascular collapse during intubation occurred in 260 of 1176 patients (22.1%) in the ketamine group and in 202 of 1189 patients (17.0%) in the etomidate group (risk difference, 5.1 percentage points; 95% CI, 1.9 to 8.3). Prespecified safety outcomes were similar in the two groups. CONCLUSIONS: Among critically ill adults undergoing tracheal intubation, the use of ketamine to induce anesthesia did not result in a significantly lower incidence of in-hospital death by day 28 than etomidate. (Funded by the Patient-Centered Outcomes Research Institute and others; RSI ClinicalTrials.gov number, NCT05277896.).
24 Apr 2026
Read appraisal →Critical care explorations
The Effects of Conservative Fluid Management and Active Deresuscitation on Markers of Tissue Perfusion, Kidney Injury, and Vascular Injury in Critically Ill Adults: A Secondary Analysis of the Role of Active Deresuscitation After Resuscitation-2 (RADAR-2) Trial
OBJECTIVES: Test the hypothesis that conservative fluid management with active deresuscitation would not adversely affect tissue perfusion or kidney injury and would be associated with reduced vascular injury compared with usual care. DESIGN: Secondary analysis of the Role of Active Deresuscitation After Resuscitation-2 (RADAR-2) trial. SETTING: ICUs. PATIENTS OR SUBJECTS: Critically ill patients enrolled in the RADAR-2 trial. INTERVENTIONS: Conservative fluid management with active deresuscitation vs. usual care. MEASUREMENTS AND MAIN RESULTS: Measures of tissue hypoperfusion (whole blood lactate), acute kidney injury (AKIRisk score and urinary cystatin-C), and vascular injury (plasma hyaluronan, syndecan-1, and angiopoietin-2) were compared between groups. For each analyte, change from baseline was compared between groups and the median inter-group difference at each timepoint was estimated with bootstrapped CIs. Exploratory logistic regression examined associations between plasma biomarker levels (including N-terminal pro-B-type natriuretic peptide [NT-proBNP]), 28-day mortality, and treatment allocation. Whole blood lactate levels were similar between groups at all timepoints. Using change from baseline comparisons, no statistically detectable between-group differences were observed in AKIRisk scores or urinary cystatin-C levels. Plasma vascular injury biomarkers showed no statistically detectable between-group differences at any timepoint. High baseline hyaluronan (adjusted odds ratio [aOR], 5.75; 95% CI, 1.94-17.02; p = 0.002), syndecan-1 (aOR, 8.82; 95% CI, 2.67-29.15; p < 0.001), and NT-proBNP greater than 2500 pg/mL (aOR, 21.48; 95% CI, 3.57-129.41; p < 0.001) were independently associated with increased 28-day mortality. There was no evidence of differential treatment response based on these biomarker levels. CONCLUSIONS: Conservative fluid management and active deresuscitation were not associated with worsening tissue perfusion or acute kidney injury. A reduction in vascular injury markers was not observed. Given the modest sample size and resultant imprecision, clinically important effects cannot be excluded.
23 Apr 2026
Read appraisal →BMJ open
Early mobilisation after abdominal surgery: a concept analysis
OBJECTIVE: To clarify and define the clinical practice concept of early mobilisation after abdominal surgery. DESIGN: A concept analysis guided by Walker and Avant's method. DATA SOURCES: MEDLINE (Ovid), AMED-(Ovid), Embase (Elsevier) and CINAHL (EBSCO) were searched through 5 December 2024. ELIGIBILITY CRITERIA: Relevant studies that included combinations of the terms 'early mobilisation', 'early ambulation', 'early acceleration', 'abdominal surgery' and 'surgical procedures' were selected. We restricted the search to English full-text publications involving adult patients, limited to the year 2000 and onward. Inclusion criteria were original research articles describing the timing and/or type of mobilisation. DATA EXTRACTION AND SYNTHESIS: The study derives its defining attributes, antecedents and consequences through data analysis. To enhance understanding of the model, we constructed related and contrary cases of the concept and outlined relevant empirical referents. RESULTS: In total, 140 studies were included in the analysis. Early mobilisation is characterised by the key defining attributes of initiating active physical movement, including standing, sitting in a chair or walking, within the first 24 hours of surgery. Antecedents include haemodynamic and respiratory stability, adequate pain management, and the patient's cognitive and physical readiness. Contextual antecedents include competent and adequately staffed healthcare teams. Consequences include improved physiological recovery and enhanced postoperative outcomes. CONCLUSIONS: This analysis provides a clarified, practice-focused definition of early mobilisation after abdominal surgery. By delineating its key attributes and contextual prerequisites, the study offers a conceptual foundation that can support clinical guidelines, promote consistent implementation and inform future research aimed at optimising postoperative recovery.
20 Apr 2026
Read appraisal →Medicine
Effect of parecoxib on postoperative anxiety and cognitive function after pancreaticoduodenectomy: A clinical trial
BACKGROUND: After pancreaticoduodenectomy, perioperative neurocognitive disorders (PND) are a common complication. Parecoxib, a selective cyclooxygenase-2 inhibitor, is used for its analgesic and antiinflammatory effects, but its impact on postoperative anxiety and cognitive function is not clear. METHODS: The patients undergoing laparoscopic pancreaticoduodenectomy at the Second Hospital of Hebei Medical University were enrolled in the study from November 1, 2020 to September 30, 2022 in this randomized controlled trial. Patients were randomly divided into Group P who received 20 mg of parecoxib intravenously 10 minutes before anesthesia and at the end of surgery and Group C who received an equal volume of normal saline at the same times. Primary outcomes included the effect of parecoxib on anxiety, cognitive impairment, and postoperative intestinal function recovery, with the incidence of PND calculated using the Z-value method. Secondary outcomes included pain control, length of hospital stay, postoperative recovery, complication rates, and analgesic use. RESULTS: 80 patients were enrolled and randomly divided into 2 groups of 40 equally: there were no significant differences in postoperative flatus passage, defecation time, hospital stay, or complication rates (P > .05). In terms of postoperative anxiety, the experimental group had significantly lower State Anxiety Inventory scores at 24 hours postoperatively (T2) (P < .05). Cognitive function, measured by Montreal Cognitive Assessment scores, did not show significant differences; however, cognitive decline was less pronounced in the experimental group. PND rates were 5% (2/40) in the experimental group versus 12.5% (5/40) in the control group, suggesting a possible clinical tendency toward reduced incidence, although the difference did not reach statistical significance. The rates of PND at 24 hours (T2) and 7 days (T3) postoperatively were 10% (4/40) and 5% (2/40) for the experimental group, compared to 15% (6/40) and 12.5% (5/40) for the control group using the Z-value method, Although the between-group differences did not reach statistical significance, a lower incidence was observed in the parecoxib group, suggesting a possible clinical tendency. CONCLUSION: Parecoxib, when used with pancreaticoduodenectomy, alleviates postoperative anxiety and neurocognitive disorders without affecting postoperative intestinal function recovery or increasing the incidence of postoperative complications.
19 Apr 2026
Read appraisal →PloS one
High-versus conventional-volume pericapsular nerve group (PENG) block for total hip arthroplasty: A randomized, controlled trial
BACKGROUND: Currently, the optimal volume of pericapsular nerve group (PENG) block for analgesia after total hip arthroplasty (THA) has not been clarified. In this trial, we investigated whether a high-volume PENG block has a superior analgesic effect than a conventional PENG block for primary THA. METHODS: Forty patients receiving primary THA under spinal anesthesia were enrolled and randomly divided into the high-volume PENG group (40 mL of 0.375% ropivacaine) and the conventional-volume PENG group (20 mL of 0.375% ropivacaine). Dexamethasone (5 mg) was added to the local anesthetic in both groups. A blinded researcher performed pain scores and lower limb sensory and motor block assessments at 3, 6, and 24 h after surgery. The time of first walking, time of first opioid consumption, total opioid consumption within 48 h, nerve block, opioid-related complications, and length of hospital stay were recorded. The primary outcome was the dynamic pain scores at 6 h post-surgery. RESULTS: Thirty-seven patients were included in the final analysis, among whom 18 were in the high-volume group and 19 in the conventional-volume group. There were no significant differences between the two groups in dynamic pain scores at 6h after surgery [median and interquartile range (p25, p75) 5(4,6) vs 4 (4,5)] and other secondary outcomes. CONCLUSIONS: The high-volume PENG block is not superior to conventional-volume PENG block in improving post-operative analgesia in patients undergoing primary THA and does not increase the risk of quadriceps weakness. TRIAL REGISTRATION: Chinese Clinical Trial Registry (ChiCTR, https://www.chictr.org.cn). Clinical trial registration number: ChiCTR2300077281.
18 Apr 2026
Read appraisal →Medicine
Publication trends of dexmedetomidine and perioperative cognitive dysfunction: A bibliometric analysis from 1995 to 2024
INTRODUCTION: Dexmedetomidine (Dex) has shown potential in mitigating perioperative cognitive dysfunction (POCD). This study aimed to systematically evaluate the research hotspots and development trends in Dex and POCD using bibliometric analysis. METHODS: A systematic search was conducted on the Web of Science Core Collection database to identify publications related to Dex and POCD from 1995 to 2024. Bibliometric analysis was performed using VOSviewer (V 1.6.20), CiteSpace (V 6.3.R1), and R V4.3.3. RESULTS: A total of 585 articles on Dex and POCD between 1995 and 2024 were analyzed. These articles involved 1688 authors from 1857 institutions across 140 countries or regions, published in 240 journals, and citing 12,881 sources. The annual growth rate of publications in this field was 16.21%, with a peak of 91 articles in 2021. China led in the number of articles (n = 379), multiple country publications (MCP = 29), and total citations = 4989). The USA ranked second with 74 articles, 18 MCPs, and 2608 citations, followed by South Korea with 29 articles and 533 citations. Harvard University published the most articles (n = 51). Wang Dong-Xin was the most influential author, with an h-index of 7. Anesthesiology led the high-impact journals with an H-index of 11. The keyword analysis identified "neuroinflammation," "risk," "injury," and "efficacy" as the latest research themes. CONCLUSION: This bibliometric analysis provides a comprehensive overview of the research trends regarding Dex and POCD. Future research should continue to explore the mechanisms, efficacy, and safety of Dex in preventing POCD.
18 Apr 2026
Read appraisal →Thorax
Cardiopulmonary exercise testing before lung resection surgery: still indicated? Evaluating predictive utility using machine learning.
RATIONALE: Despite significant advances in patient care and outcomes, criteria for cardiopulmonary exercise testing (CPET) in risk stratification guidelines for lung resection have not been updated in over a decade. We hypothesised that CPET no longer holds additional predictive value for postoperative complications. METHODS: In this secondary analysis, we included lung resection candidates from two prospective, multicentre studies eligible for CPET and assessed with preoperative pulmonary function tests (PFTs) and arterial blood gas analysis. Postoperative pulmonary (PPCs) and cardiovascular complications (PCCs) were documented during hospitalisation. We trained five types of machine learning models applying nested cross-validation to predict complications and compared predictive performance based on four metrics, including area under the receiver operating characteristic curve (AUC-ROC). RESULTS: A total of 497 patients were included. PPCs developed in 71 (14%) patients. Adding CPET parameters to PFTs and baseline clinical data did not improve the ability of models to predict PPCs in unselected patients (AUC-ROC=0.72-0.78; p=0.47), nor in those meeting American College of Chest Physicians (ACCPs) (n=236; AUC-ROC=0.64-0.78; p=0.70) or European Respiratory Society/European Society of Thoracic Surgery (ERS/ESTS) criteria (n=168; AUC-ROC=0.59-0.76; p=0.92). PCCs developed in 90 (18%) patients. CPET parameters likewise did not improve model performance for the prediction of PCCs in unselected patients (AUC-ROC=0.65-0.73; p=0.96), nor in the ACCP (AUC-ROC=0.61-0.73; p=0.82) or ERS/ESTS subgroups (AUC-ROC=0.62-0.69; p=0.87). CONCLUSIONS: In contemporary surgical practice, CPET did not improve the predictive performance of machine learning models for PPCs or PCCs in patients with an indication based on established guidelines or in those without. The role of CPET in preoperative risk stratification for lung resection should be re-evaluated. TRIAL REGISTRATION NUMBER: NCT03498352, NCT04826575.
17 Apr 2026
Read appraisal →A&A practice
Automated Identification of Cardiopulmonary Disease Cases for Preoperative Risk Stratification Using Machine Learning: A Retrospective Analysis
BACKGROUND: Preoperative chart review is time-consuming and prone to errors, particularly for cardiopulmonary conditions that impact anesthetic planning. We developed a guideline-aligned "clinical insight bot" that mines free-text documentation to surface perioperative cardiovascular risk signals relevant to the 2024 Mult Society perioperative guideline for noncardiac surgery. METHODS: We analyzed 1000 de-identified medical cases from the PhysioNet MIMIC database. Medical terminology was extracted using regex-based NLP and categorized into 13 clinical specialties. Text features were encoded using TF-IDF vectorization and 1536-dimensional semantic embeddings stored in a PostgreSQL vector database (pgvector). Four machine learning models-Logistic Regression, Random Forest, Support Vector Machine (SVM), and Naive Bayes-were trained with stratified fivefold cross-validation to classify cases as "cardiopulmonary-only" versus "mixed/other." Performance was evaluated using accuracy, precision, recall, and F1 score, with statistical comparison via McNemar's test and bootstrap confidence intervals. RESULTS: In a held-out test set of 200 notes (28 positive; 172 negatives; ~14% prevalence), a linear support vector machine achieved the best overall balance (F1 ≈ 0.71), with high precision (positive predictive value 0.94) and very low false positive rate (FPR) (1/172 ≈ 0.6%). False negatives were the dominant residual error class. The pipeline processed documents near-instantaneously and, when scaled to 1000 notes, replaced on the order of tens of clinician review hours (≈100× efficiency gain) while maintaining performance across common preoperative document types. CONCLUSIONS: A lightweight, guideline-aligned insight bot can transform unstructured preoperative notes into concise, stepwise prompts that flag cardiovascular risk signals before the day of surgery. High precision with a very low FPR supports safe integration with anesthesiology workflows by minimizing paging noise, whereas time savings create operational and financial value. Future work should emphasize multicenter validation, structured data fusion (including labs, imaging, and vitals) to improve sensitivity, and prospective evaluation of downstream clinical and operational outcomes.
17 Apr 2026
Read appraisal →International journal of medical informatics
Clinician preferences for explainable AI in critical care: a comparative study of interpretable models and visualizations for intubation decision support
BACKGROUND: The complexity of many AI models hinders their clinical adoption because the clinicians using them do not regard them as transparent. This study addresses the lack of clinician-centered explainable AI (XAI) interfaces by designing and evaluating intuitive visual explanations for intubation prediction, testing the hypothesis that workflow-compatible designs enhance acceptance. OBJECTIVE: This study compares three, time-aware, visual explanations for XAI-based intubation prediction and evaluate their acceptance, comprehension, and perceived utility among clinicians. METHODS: We developed machine learning models to estimate the near-term risk of deterioration in the patient's condition which may lead to mechanical intubation using ICU time-series data. We generated global and local explanations using SHAP and designed three customized visual formats-a temporal force plot, a temporal bar chart, and a dual-encoded SHAP heatmap. Clinicians (n = 206) evaluated comprehension and usability using objective questions and a Likert-based survey. RESULTS: Based on 4608 critically ill patients with 10 medical variables over 7 hours of data for each patient, the Random Forest (RF) model achieved the highest area under the curve (AUC): 0.94. Furthermore, the local explanations were customized and evaluated by 206 clinicians through a survey conducted on the Prolific platform. A customized heatmap representation was selected as the visualization with the highest perceived clinical utility and alignment with clinical workflows. DISCUSSION: The reported findings support the need for explanation formats to be tailored to clinical reasoning and task context, supporting the concept of cognitive fit. The heatmap's close alignment with clinicians' mental models and its graphical integrity enhances interpretability and trust. This study demonstrates that explanation effectiveness depends on contextual relevance, rather than a universal standard, and that the presentation format itself significantly shapes clinicians' trust in XAI systems. CONCLUSION: This study advances clinical XAI by introducing a time-aware explanation framework for ICU intubation decisions. By integrating temporal trends with model reasoning, our visualizations closely align with clinicians' cognitive workflows. Rigorous clinician-centered evaluation identified the dual-encoded SHAP heatmap as the most useful and workflow-compatible visualization, highlighting the importance of explanation design alongside predictive accuracy for clinical adoption.
15 Apr 2026
Read appraisal →A&A practice
Hypotension Prediction Index-Guided Hemodynamic Management on Postoperative Organ Outcome: A Systematic Review and Meta-Analysis of Randomized Clinical Trials
Intraoperative hypotension (IOH) is a frequent occurrence during noncardiac surgery. Hypotensive episodes can compromise tissue perfusion, and cumulative exposure has been associated with adverse outcomes such as acute kidney injury (AKI), myocardial injury, and increased mortality. The Hypotension Prediction Index (HPI) is an artificial intelligence tool designed to predict hypotensive events before they occur, allowing clinicians to intervene proactively to prevent or mitigate hypotension. Although randomized trials and pooled analyses have shown its effectiveness in reducing both the incidence and duration of IOH, its impact on patient-centered outcomes remains unclear. From 1946 to August 8, 2025, we prospectively queried MEDLINE (PubMed), Embase, Cochrane Library, and Web of Science. We included randomized controlled trials in adult surgical patients with invasive arterial monitoring, comparing HPI-guided perioperative hemodynamic management to standard care. Primary outcome was postoperative AKI, and secondary outcomes were postoperative myocardial injury, pneumonia, arrhythmia, stroke, postoperative cognitive dysfunction, including delirium, surgical site infection, length of hospital stay, and mortality. Two authors independently evaluated the risk of bias according to the methods described in the Cochrane Handbook for Systematic Reviews of Interventions. A meta-analysis was performed using RevMan 5.4.1 software. Dichotomous data were presented as risk ratio, 95% confidence interval [CI]; continuous data were presented as mean difference, 95% CI. All results were pooled using random-effects models. Of 883 imported references, 14 studies (N = 2030) were included for final analysis. No significant difference in incidence of postoperative AKI was observed between HPI-guided management and standard care, RR = 0.87, 95% CI (0.71-1.07). Postoperative myocardial injury was significantly reduced in HPI-guided management group, RR = 0.61, 95% CI (0.39-0.96), although exclusion-based sensitivity analysis suggests that this finding may not be robust. No other significant differences were observed for hospital length of stay, arrhythmia, stroke, cognitive dysfunction (including delirium), surgical site infection, pneumonia, or mortality. Although HPI-guided management is associated with a reduction in postoperative myocardial injury, its significance does not withstand sensitivity analysis. All other postoperative complications, including AKI incidence, are not reduced in the intervention groups. This study has limitations inherent to the heterogeneity of included patient populations, surgical interventions performed, standard of care practices, and reporting of postoperative outcomes.
15 Apr 2026
Read appraisal →BMJ (Clinical research ed.)
Non-drug perioperative interventions to reduce postoperative pulmonary complications after abdominal surgery: systematic review and meta-analysis
OBJECTIVE: To evaluate the effectiveness of perioperative non-drug interventions in reducing postoperative pulmonary complications (PPCs) in adults undergoing abdominal surgery. DESIGN: Systematic review and meta-analysis. DATA SOURCES: Ovid MEDLINE, Embase, and Web of Science from database inception to January 2025 and updated in January 2026, with no language restrictions. STUDY SELECTION: Randomised controlled trials assessing the effectiveness of perioperative non-drug interventions for the prevention of PPCs in adults undergoing elective abdominal surgery under general anaesthesia, with clearly defined PPCs. MAIN OUTCOME MEASURES: The primary outcome was the proportion of patients developing PPCs. Secondary outcomes included the proportion of patients with PPC subtypes according to European Perioperative Clinical Outcome definitions (respiratory infection, respiratory failure, pleural effusion, atelectasis, or pneumothorax) and hospital length of stay. DATA EXTRACTION AND SYNTHESIS: Two reviewers independently screened studies, extracted data, and assessed risk of bias with the Cochrane RoB 2.0 tool. Data were synthesised using meta-analyses and trial sequential analyses, with the evidence certainty assessed using the Grading of Recommendations, Assessment, Development and Evaluation (GRADE) approach. RESULTS: 255 trials including 55 260 participants were included, evaluating 10 types of interventions with 39 subtypes for PPC prevention. PPCs occurred in 6467 (11.7%) participants across all included trials. High certainty evidence showed that low fraction of inspired oxygen (FiO2) significantly reduced PPCs (risk ratio 0.81, 95% confidence interval 0.71 to 0.92). Moderate certainty evidence showed benefit for four intervention types: lung protective ventilation (risk ratio 0.66, 0.57 to 0.76), physiotherapy (0.55, 0.46 to 0.65), analgesia (0.73, 0.64 to 0.84), and nutrition (0.74, 0.63 to 0.87), with individualised positive end expiratory pressure, composite lung protective ventilation, early mobilisation, and epidural analgesia also showing benefit at the subtype level. Trial sequential analysis confirmed sufficient cumulative evidence for all the above interventions except early mobilisation. By contrast, goal directed haemodynamic therapy, targeted blood pressure management, restrictive fluid therapy, and postoperative bi-level positive airway pressure showed no evidence of benefit, with moderate certainty. CONCLUSIONS: This synthesis establishes an evidence hierarchy for PPC prevention in abdominal surgery. Low FiO2 is the only intervention supported by high certainty evidence and should be prioritised in clinical practice. Other beneficial strategies include lung protective ventilation, physiotherapy, analgesic techniques, and nutrition interventions. Conversely, the role of goal directed haemodynamic therapy-despite its widespread use-warrants reconsideration for PPC prevention. These findings facilitate prioritisation of effective interventions and development of evidence based guidelines. STUDY REGISTRATION: PROSPERO CRD42025637449.
11 Apr 2026
Read appraisal →A&A practice
Efficacy and Safety of Ketamine or Esketamine versus Fentanyl-Class Opioids as Adjuncts to Propofol for Gastrointestinal Endoscopy: A Systematic Review and Meta-Analysis
OBJECTIVE: Propofol is widely used for gastrointestinal (GI) endoscopies but may cause respiratory depression and hemodynamic compromise when used alone. Adjuncts such as ketamine or esketamine and fentanyl-class opioids (fentanyl, alfentanil, and sufentanil) are used to optimize sedation, yet their comparative efficacy and safety remain uncertain. This meta-analysis evaluated ketamine or esketamine plus propofol (KP regimens) versus fentanyl/alfentanil/sufentanil plus propofol (FP regimens). METHODS: Following Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines, we systematically searched PubMed, Scopus, Cochrane Library, and Google Scholar from inception to February 2025 for randomized controlled trials (RCTs). Eligible studies included patients undergoing GI endoscopy (Population), receiving ketamine or esketamine with propofol (Intervention), compared with fentanyl-class opioids with propofol (Comparison), and reporting outcomes including propofol dose, comfort, adverse events, recovery time, or procedure time (Outcomes). Risk of bias was assessed with the Cochrane tool. Heterogeneity was evaluated using Cochran's Q test and the I2 statistic, and clinical heterogeneity was considered based on study design and population differences. Sensitivity analyses were conducted to test robustness of findings. RESULTS: Fifteen RCTs (1492 participants) were included. KP regimens significantly reduced total propofol use compared with FP (mean difference [MD] -0.42; 95% confidence interval [CI] -0.66 to -0.19; P = .0005), with greater benefit observed for esketamine (p for subgroup = 0.001). No significant differences were found in sedation (MD 0.01; 95% CI -0.40 to 0.43; P = .95) or pain scores (MD 0.24; 95% CI -0.18 to 0.65; P = .26). Safety outcomes were comparable across groups, with no significant differences in desaturation, nausea/vomiting, bradycardia, or tachycardia, except for hypotension, which was lower with KP (risk ratio [RR] 0.56; 95% CI 0.39 to 0.82; P = .003). Recovery and procedure times were similar between groups. CONCLUSIONS: Both KP and FP regimens are effective and safe for GI endoscopy. KP regimens reduce total propofol requirements and significantly lower the risk of hypotension, while showing comparable sedation quality, analgesia, and recovery profiles to FP. These findings suggest KP may offer safety and dosing advantages, though larger standardized trials are needed to confirm its clinical superiority.
10 Apr 2026
Read appraisal →PloS one
An interpretable deep learning framework for predictive modeling of postoperative infections in ICU patients
A significant proportion of intensive care unit (ICU) patients undergo surgical procedures, and some may develop postoperative infections. Accurately predicting postoperative infection risk and identifying key contributing factors is crucial for improving postoperative management and understanding infection mechanisms. However, this task is challenging due to the complex interplay of multiple risk factors. While machine learning models can model these intricate associations to predict postoperative infection risk, their lack of interpretability - failing to uncover each factor's impact-hinders their adoption in clinical settings. To address this difficulty, we introduced an interpretable deep neural network (DNN) model that integrates a permutation feature importance test (PermFIT). PermFIT rigorously evaluates the impact of each feature on postoperative infection risk through a rigorous statistical inference. By using only the identified important features as inputs, the DNN's predictive performance can be further enhanced. We conducted an extensive study using electronic health records (EHRs) from the Medical Information Mart for Intensive Care (MIMIC-III), a large-scale ICU EHR database. Under the PermFIT framework, our DNN model effectively identifies significant factors associated with postoperative infections while delivering the most accurate postoperative infection risk predictions. These findings highlight the clinical utility of our proposed DNN framework in managing postoperative care for ICU surgical patients, ultimately improving their health outcomes.
10 Apr 2026
Read appraisal →Annals of medicine
Comparison of VCV and PCV-VG modes on diaphragmatic function in diabetic patients undergoing laparoscopic colorectal surgery: a prospective randomized controlled study
BACKGROUND: Diabetic patients are prone to induce diaphragmatic weakness, which can lead to postoperative pulmonary complications (PPCs). The optimal mechanical ventilation mode may potentially improve postoperative diaphragmatic function. This study evaluates the effects of two ventilation modes under driving pressure-guided ventilation strategy on diaphragmatic function, as assessed by diaphragm thickening fraction (DTF) and diaphragm excursion (DE), in diabetic patients following laparoscopic colorectal surgery. METHODS: Eighty patients diagnosed with Type II diabetes scheduled for elective laparoscopic colorectal surgery, were randomly allocated to either the pressure-controlled volume-guaranteed ventilation (PCV-VG) group (Group P) or the volume-controlled ventilation (VCV) group (Group V) during surgery. The primary outcome was diaphragmatic function assessed during both tidal breathing and maximal inspiratory effort after surgery. Secondary outcomes included intraoperative mechanical power, PPCs, and other complications. RESULTS: A total of eighty patients were included in the final analysis. The averaged area under the curve (AUC) for mechanical power during ventilation was significantly lower in Group P than in Group V (p = 0.002). PCV-VG significantly improved both DE and DTF within the first two days post-surgery (AUCDEtidal: p = 0.088, AUCDTFtidal: p = 0.004, AUCDEmax: p = 0.029, AUCDTFmax: p = 0.017). Postoperative diaphragmatic weakness was less frequent in Group P than in Group V (p = 0.019). However, there was no difference in the incidence of PPCs between the two groups (p = 0.155). CONCLUSION: PCV-VG mode can reduce intraoperative mechanical power, better preserve postoperative diaphragmatic function. However, these improvements did not translate into clinical benefits, as evidenced by the lack of reduction in the incidence of PPCs.
10 Apr 2026
Read appraisal →European journal of orthopaedic surgery & traumatology : orthopedie traumatologie
Memantine as a pain management strategy post-total knee arthroplasty: a randomized controlled trial
INTRODUCTION: Total knee arthroplasty (TKA) provides significant pain relief and improved function for patients with knee osteoarthritis. Despite its effectiveness, postoperative pain remains a major challenge. Effective management of this acute postoperative pain is critical for enhancing recovery. The study investigates the effectiveness of memantine in reducing postoperative pain in patients undergoing TKA. METHODS: This double-blind, randomized controlled trial aimed to evaluate the efficacy of memantine for managing postoperative pain after TKA. Conducted from April to September 2022, eligible participants over 18 years with ASA grades I or II were recruited. Exclusions included patients with substance abuse histories, chronic opioid use, significant comorbidities, or contraindications to memantine. Participants were randomly assigned to receive either memantine (20 mg daily preoperatively for three days and 10 mg daily postoperatively for two weeks) or a placebo. Both groups received standard analgesics during hospitalization and upon discharge. Pain levels were assessed using Visual Analog Scale (VAS) scores and Short-Form McGill Pain Questionnaire (SF-MPQ) at various time points. Secondary outcomes measured included opioid usage and nocturnal awakening. RESULTS: Memantine did not significantly lower pain levels compared to placebo, as measured by VAS (mean difference: 0.138, CI 95%: - 0.468-1.103, P = 0.418) and SF-MPQ (mean difference: 1.962, CI 95%: - 1.013-4.936, P = 0.190). Patients receiving memantine needed opioid analgesia (2.85 ± 0.67 times) significantly (P < 0.001) less frequently than the Placebo patients (4.14 ± 1.06 times). Nocturnal awakening did not differ significantly between the two groups (Memantine: 3.59 ± 2.15, Placebo: 2.33 ± 2.12, P = 0.067). CONCLUSION: The comparison of pain levels between oral memantine and placebo indicates that memantine does not significantly reduce pain in patients undergoing TKA. However, it significantly lowers the need for opioids postoperatively.
10 Apr 2026
Read appraisal →Scientific reports
Effects of different inflatable heater temperature settings on early postoperative recovery after hysteroscopic surgery: a randomized clinical trial
To investigate the effect of perioperative temperature settings of an inflatable warming device on postoperative recovery quality in patients undergoing hysteroscopic myomectomy, this randomized controlled trial enrolled 258 patients receiving laryngeal mask general anesthesia at Xuzhou Central Hospital, China, between March 2022 and August 2024. Patients were allocated to perioperative temperature management with an inflatable warming device set to 38 °C (Group L, n = 129) or 43 °C (Group H, n = 129). A total of 211 patients were included in the final analysis after accounting for exclusions. At 24 h postoperatively, the QoR-40 score was significantly higher in Group H compared to Group L (P < 0.05). At baseline (T0), no significant differences in core body temperature, mean arterial pressure (MAP), or heart rate (HR) were observed between groups (P > 0.05). From T1 to T6, Group H exhibited higher and more stable core body temperature, MAP, and HR compared to Group L (P < 0.05). There were no statistically significant differences in QoR-40 scores at 48 h (P > 0.05) or in the incidence of postoperative restlessness, chills, and infection between the two groups. Compared to 38 °C, the perioperative application of a 43 °C inflatable warming device improved early postoperative recovery quality at 24 h in hysteroscopic myomectomy patients without increasing complication risks.
10 Apr 2026
Read appraisal →BMJ open
Facilitators of and barriers to teamwork between operating room nurses and other practitioners: a meta-synthesis of qualitative studies
OBJECTIVES: Seamless collaboration in the operating room is the critical foundation for patient safety and successful outcomes. This synergy enhances efficiency, minimises errors and is a key determinant of the final surgical quality. The objective of this study is to synthesise qualitative studies on facilitators of and barriers to teamwork between operating room nurses (ORNs) and other practitioners. DESIGN: Systematic review and meta-synthesis of qualitative studies. DATA SOURCES: A systematic search was conducted across nine electronic databases: PubMed, Embase, Scopus, Web of Science (Core Collection), CINAHL, PsycINFO, CNKI, WanFang and Vip. The search covered the period from 1 January 2014 to 30 September 2025. The search strategy used both medical subject headings and free-text terms, supplemented by the snowball method and grey literature. ELIGIBILITY CRITERIA FOR SELECTING STUDIES: Qualitative studies exploring teamwork involving ORNs were included. Of 4839 initially retrieved records, 29 studies finally met the inclusion criteria. DATA EXTRACTION AND SYNTHESIS: Two researchers independently extracted data, including the author, year of publication, country, study design, study objective, interview location, number of participants, age, methods of data collection and analysis and duration of interviews. The key findings were synthesised into two overarching themes: facilitators (eg, professional skills, role awareness, team familiarity, patient-centred culture, effective leadership) of and barriers (eg, individualistic behaviour, hierarchical culture, insufficient resources, punitive management, environmental constraints) to teamwork between ORNs and other practitioners. A meta-synthesis was conducted to synthesise study results from different groups and countries. NVivo 15 was used for data management. RESULTS: A total of 29 qualitative studies (published between 2014 and 2025) were included in this study. Key facilitators of teamwork included professional skills, role awareness, team familiarity, patient-centred culture and effective leadership. Major barriers encompassed individualistic behaviours, punitive management, hierarchical culture, environmental constraints and insufficient resources. These factors worked across individual, interpersonal and systemic levels to shape OR teamwork. CONCLUSION: Optimising OR teamwork requires a multifaceted approach simultaneously addressing these facilitators and barriers. Interventions should simultaneously cultivate individual competencies (eg, skills, role awareness), foster positive team dynamics (eg, familiarity, psychological safety, inclusive leadership), and rectify systemic issues (eg, hierarchical culture, resource constraints, punitive management). In this way, teamwork between ORNs and other practitioners can be enhanced, thereby contributing to improved perioperative safety and outcomes. PROSPERO REGISTRATION NUMBER: CRD42024506414.
9 Apr 2026
Read appraisal →Medical science monitor : international medical journal of experimental and clinical research
Ultrasound-Guided Modified Clavipectoral Fascial Plane Block With Superficial Cervical Plexus Block for Midshaft Clavicular Surgery: A Prospective Randomized Controlled Trial
BACKGROUND This study evaluated the efficacy and safety of modified clavipectoral fascial plane block (mCPB) plus superficial cervical plexus block (SCPB) compared with interscalene brachial plexus block (ISBP) in midshaft clavicular fracture surgery. MATERIAL AND METHODS Fifty-six patients scheduled for midshaft clavicular surgery under regional anesthesia were randomly allocated to a control (ultrasound-guided ISBP plus SCPB) or experimental (mCPB plus SCPB) group at a 1: 1 ratio. The primary outcome was the Numerical Rating Scale (NRS) score at 12 hours postoperatively. Secondary outcomes were NRS scores at 4, 8, and 24 hours; rescue analgesic administration frequency; block success rate; procedural and onset times; and hemidiaphragmatic paralysis incidence. RESULTS Block success was 100% in both groups, without differences in performance time or onset time. A significant group × time interaction was observed (P<0.001). Compared with the control group, the experimental group had significantly lower NRS scores at 12 hours (estimated mean difference -1.29; P<0.001) and 24 hours (estimated mean difference -2.36; P<0.001). No significant differences occurred at 4 or 8 hours. Rescue analgesic requirements within 24 hours were significantly reduced in the experimental group (P=0.002). Hemidiaphragmatic paralysis incidences were 71.4% in the control group and 0% in the experimental group (P<0.001). CONCLUSIONS In midshaft clavicular fracture surgery, mCPB plus SCPB provided effective anesthesia with procedural characteristics comparable to ISBP, while avoiding hemidiaphragmatic paralysis and demonstrating superior analgesia at 12 and 24 hours. These findings require validation in larger multicenter trials.
8 Apr 2026
Read appraisal →Biomedical physics & engineering express
Detecting early signs of patient deterioration at home using wearable sensors: a personalized anomaly detection approach
In the rapidly emerging field of transferring healthcare from hospitals to patients' homes, it is essential that signs of remote deterioration at home should be detected early to timely upscale care when needed. As patient deterioration is a rare event, datasets have insufficient event data for supervised machine learning. Personalized anomaly detection (AD) may be a good alternative. Therefore, this study aimed to obtain insight into the use of personalized AD models in detecting early signs of patient deterioration at home using wearable sensor data. To address this, Isolation Forest and Local Outlier Factor models were applied to detect signs of deterioration in terms of mortality or unplanned readmission in the 24 preceding hours in two datasets: one of heterogeneous patients, the other of postoperative patients. A pipeline was developed for continuously updating personalized AD models over time and applying them to detection windows twice per day. Results were compared with the Remote Early Warning Score. Isolation Forest (AUROC: 0.69) and Local Outlier Factor (AUROC: 0.67) models were able to find some early signs of deterioration in heterogeneous patients (n = 113). However, in postoperative patients (n = 193), Isolation Forest (AUROC: 0.41) and Local Outlier Factor (AUROC 0.44) performed badly. The Remote Early Warning Score was able to find some early signs of deterioration for both groups (AUROC: 0.63-0.76). Based on these findings, three requirements were formulated that should be fulfilled for a potentially successful application of personalized AD. First, the training set should be normal and representative of non-deteriorating patients. Second, signs of deterioration should exhibit abnormal characteristics. Third, non-deteriorating patients should not exhibit abnormal characteristics.
5 Apr 2026
Read appraisal →Medicine
Comparison of costoclavicular with infraclavicular block for upper limb surgeries: A meta-analysis
BACKGROUND: The influence of costoclavicular versus infraclavicular nerve block on the analgesia for upper limb surgeries remains elusive. This meta-analysis and systematic review aim to study the analgesic efficacy of costoclavicular versus infraclavicular nerve block for upper limb surgeries. METHODS: We have searched several databases, including PubMed, Embase, Web of Science, EBSCO, and Cochrane Library databases from inception to January 2024, and randomized controlled trials (RCTs) assessing the effect of costoclavicular versus infraclavicular nerve block for upper limb surgeries were included. The inclusion criteria were presented as follows: study design was RCT, patients underwent upper limb surgeries, and intervention treatments were costoclavicular nerve block versus infraclavicular nerve block. Studies without English abstracts were excluded. The methodological quality of the included studies was evaluated by the modified Jadad scale. Standard mean difference with 95% confidence interval (CI) was used to assess continuous outcomes, while odds ratio (OR) with 95% CI was applied to evaluate dichotomous outcomes. RESULTS: Six RCTs and 434 patients were included in this meta-analysis. Compared with infraclavicular nerve block for upper limb surgeries, costoclavicular nerve block was able to significantly decrease block onset time (SMD = -0.29; 95% CI = -0.53 to -0.05; P = .02), but demonstrated no influence on performance time (SMD = -0.26; 95% CI = -0.85 to 0.34; P = .40), number of passes (SMD = -0.08; 95% CI = -0.40 to 0.24; P = .61), vascular puncture (OR = 0.25; 95% CI = 0.04 to 1.56; P = .14) or Horner syndrome (OR = 1.02; 95% CI = 0.04 to 27.64; P = .99). CONCLUSION: Costoclavicular nerve block may need shorter block onset time for upper limb surgeries compared to infraclavicular nerve block, but more RCTs were needed to confirm these findings.
5 Apr 2026
Read appraisal →Regional anesthesia and pain medicine
Human versus artificial intelligence: evaluating ChatGPT's performance in conducting published systematic reviews with meta-analysis in chronic pain research
INTRODUCTION: Artificial intelligence (AI), particularly large-language models like Chat Generative Pre-Trained Transformer (ChatGPT), has demonstrated potential in streamlining research methodologies. Systematic reviews and meta-analyses, often considered the pinnacle of evidence-based medicine, are inherently time-intensive and demand meticulous planning, rigorous data extraction, thorough analysis, and careful synthesis. Despite promising applications of AI, its utility in conducting systematic reviews with meta-analysis remains unclear. This study evaluated ChatGPT's accuracy in conducting key tasks of a systematic review with meta-analysis. METHODS: This validation study used data from a published meta-analysis on emotional functioning after spinal cord stimulation. ChatGPT-4o performed title/abstract screening, full-text study selection, and data pooling for this systematic review with meta-analysis. Comparisons were made against human-executed steps, which were considered the gold standard. Outcomes of interest included accuracy, sensitivity, specificity, positive predictive value, and negative predictive value for screening and full-text review tasks. We also assessed for discrepancies in pooled effect estimates and forest plot generation. RESULTS: For title and abstract screening, ChatGPT achieved an accuracy of 70.4%, sensitivity of 54.9%, and specificity of 80.1%. In the full-text screening phase, accuracy was 68.4%, sensitivity 75.6%, and specificity 66.8%. ChatGPT successfully pooled data for five forest plots, achieving 100% accuracy in calculating pooled mean differences, 95% CIs, and heterogeneity estimates (I2 score and tau-squared values) for most outcomes, with minor discrepancies in tau-squared values (range 0.01-0.05). Forest plots showed no significant discrepancies. CONCLUSION: ChatGPT demonstrates modest to moderate accuracy in screening and study selection tasks, but performs well in data pooling and meta-analytic calculations. These findings underscore the potential of AI to augment systematic review methodologies, while also emphasizing the need for human oversight to ensure accuracy and integrity in research workflows.
4 Apr 2026
Read appraisal →Critical care explorations
Physician Gestalt Compared With Artificial Intelligence Model to Predict Intubation in Critically Ill Patients
IMPORTANCE: Accurate prediction of intubation in critically ill patients could enable interventions that improve patient outcomes. However, the performance of intensive care physicians compared with machine learning (ML) models remains unknown. OBJECTIVES: To investigate intensive care physicians' ability to predict the need for intubation within 24 hours and compare their performance against an established ML model, Vent.io. DESIGN, SETTING, AND PARTICIPANTS: This prospective observational study in two ICUs surveyed intensivists to test their ability to predict the need for intubation of adult patients under their care. Physician predictions of intubation were then compared with predictions from Vent.io. MAIN OUTCOMES AND MEASURES: Primary metrics include prediction sensitivity, specificity, and descriptive statistics for both physicians and ML model. Generalized linear mixed models investigated the fixed effect of the predictor (physician vs. Vent.io) on both sensitivity and specificity while accounting for the random effects from different physicians. Similar modeling was used to investigate the relationship between physician confidence and correctness. RESULTS: Overall, physicians were quite confident in their predictions of intubation with a median score of 8 (on a 0-10 point scale, with 0 being not at all confident and 10 being extremely confident) out of the 302 surveys administered. Sensitivity was 0.190 and 0.714 for physicians and Vent.io, respectively. Specificity was 0.960 and 0.673 for physicians and Vent.io, respectively. Generalized linear mixed modeling showed that physician confidence was associated with greater odds of correctly predicting intubation outcome (odds ratio [OR] 1.49; 95% CI, 1.22-1.84; p < 0.001). Vent.io had significantly greater odds of being correct when patients required intubation compared with physicians (OR 18.68; 95% CI, 1.87-186.31; p = 0.013). However, intensive care physicians outperformed Vent.io at correctly predicting when patients did not require intubation (OR 24.80; 95% CI, 13.22-46.52; p < 0.001). CONCLUSIONS AND RELEVANCE: Although predictive performance compared with human experts is promising, Vent.io needs real-time testing in a randomized clinical trial to determine if its deployment can improve clinical outcomes.
4 Apr 2026
Read appraisal →Current opinion in anaesthesiology
Artificial intelligence-driven predictive analytics for postoperative management and recovery in trauma patients
PURPOSE OF REVIEW: Post-traumatic care is evolving from a reactive, protocol-driven paradigm to a predictive, personalized approach. This review examines how artificial intelligence and machine learning are redefining postoperative management by predicting complications before they manifest. RECENT FINDINGS: Recent literature (2023-2025) highlights three major advances: (a) the validation of gradient boosting algorithms (e.g. Extreme Gradient Boosting) that significantly outperform traditional scoring systems for predicting trauma-induced coagulopathy; (b) the development of interpretable, phenotype-specific models for venous thromboembolism risk stratification, particularly in traumatic brain injury; and (c) the emergence of real-time sepsis prediction tools that account for the sterile inflammation inherent to trauma. However, a recurring limitation in current research is the reliance on retrospective datasets and single-center validations, underscoring the critical need for rigorous external validation across diverse patient populations before widespread clinical adoption. SUMMARY: Artificial intelligence is not merely a monitoring tool but a driver of precision medicine in trauma. By leveraging diverse modalities, from computer vision in radiology to natural language processing in electronic health records, clinicians can now anticipate adverse events. To bridge the gap between algorithm and bedside, future efforts must focus on overcoming significant implementation barriers, such as data interoperability, and ensuring model generalizability.
4 Apr 2026
Read appraisal →Yonsei medical journal
Predicting Post-Induction Hypotension in Diverse Surgical Populations: A Multiclass Classification Universal Model Using Machine Learning Techniques
PURPOSE: Our study aims to develop a machine learning model that not only predicts the occurrence of post-induction hypotension (PIH) but also assesses its severity, addressing a broader patient population than previous studies which mostly focused on a single population. MATERIALS AND METHODS: In our study, we extracted data from 71473 patients aged 18 years and older who underwent general anesthesia for non-cardiac surgery at a tertiary care hospital, using the electronic medical record system for modeling. We used patient demographics, baseline and pre-induction blood pressure, preoperative laboratory tests, surgical details, and anesthetics data, focusing on predicting PIH. The severity of hypotension was assessed by integrating and calculating the integral value of hypotensive periods. We employed several machine learning techniques and evaluated their performance using accuracy, precision, F1-scores, and macro-averaged area under the curve. Additionally, SHapley Additive exPlanation values were used to determine the key factors influencing the predictions. RESULTS: A multiclass classification model, which assesses hypotension severity through the area of hypotension, surpassed the binary model with an F1-score of 0.664. Among various machine learning techniques, the eXtreme Gradient Boosting (XGBoost) model exhibited superior prediction performance, achieving an accuracy of 0.755 and an F1-score of 0.664. Models using preoperative blood pressure and demographic data outperformed those using other datasets. CONCLUSION: We found that using the XGBoost ensemble machine learning technique aids in predicting PIH before surgery, and introducing a multiclass classification model that indicates the severity of hypotension to clinicians leads to an overall enhancement in prediction performance, thereby increasing its clinical utility for real-world applications.
4 Apr 2026
Read appraisal →Physiological measurement
Deep learning approach for stroke volume variation estimation: retrospective modeling and prospective deployment in real-world practice
Objective. We developed stroke volume variation (SVV) Net, a deep learning-based model for estimating SVV, and validated its performance and clinical applicability through both large-scale retrospective modeling and prospective real-time deployment within intraoperative settings.Approach. A SVVNet was trained on arterial blood pressure (ABP) waveforms from 2849 patients undergoing general anesthesia at Asan Medical Center (AMC). For retrospective validation, we utilized AMC cohorts and a publicly accessible database to assess agreement, trending ability, noise robustness (quantifies via signal quality index), and inter-individual deviation against ABP-based pulse wave analysis (ABP-PWA) and other DL models. For prospective validation, the optimized model was deployed onto a tablet and prospectively evaluated in real-time during 67 surgeries.Main results. In the retrospective external validation, SVVNet achieved higher agreement mean absolute error (MAE = 0.70, Pearsonρ= 0.92), compared to ABP-PWA algorithm (MAE = 1.30, Pearsonρ= 0.80) and lower inter-individual deviation. It was more robust to signal noise (MAE = 4.33 vs 30.87). In the prospective real-time deployment, our deployed model achieved an MAE of 1.33 and a Pearson correlation of 0.87 against the ground truth.Significance. This study emphasizes the advantages of the DL-based model compared to ABP-PWA-based algorithms and presents prospective results that support the clinical applicability of SVVNet in the operating room. Our model has the potential to replace pre-existing monitoring systems and to enhance the complex surgical environment.
4 Apr 2026
Read appraisal →Anaesthesia, critical care & pain medicine
Hemodynamic phenotyping 4.0.
The concept of integrating hemodynamic variables to define specific profiles or phenotypes has been established for decades. Describing hemodynamic phenotypes plays a key role in educating healthcare professionals about cardiovascular physiology, enhancing the understanding of shock mechanisms, and informing treatment strategies. Recently, two notable innovations have emerged to support bedside identification of hemodynamic phenotypes: machine learning (ML) algorithms and visual decision support tools. When it comes to "small data," such as a limited set of hemodynamic variables, ML algorithms may not be essential for data integration or interpretation. In addition, the hemodynamic phenotypes identified by ML techniques often mirror traditional textbook profiles, though occasionally with inconsistencies that may impact patient safety. This raises valid questions about the need to integrate complex and proprietary ML algorithms for bedside hemodynamic assessment. By contrast, visual tools leverage clinicians' innate ability to process graphical information rapidly, improving the understanding of cardiovascular physiology and enabling recognition of hemodynamic profiles at a glance. As such, they may offer a practical, accessible, and cost-effective alternative to ML-based solutions. Future studies comparing the clinical impact of visual versus ML-driven phenotyping are now needed to guide further development and implementation.
2 Apr 2026
Read appraisal →Adv Anesth
The Current Use of Artificial Intelligence in Anesthesiology
Artificial intelligence (AI) applications are increasingly being used in areas of medicine such as radiology. However, within anesthesiology, there is not widespread use of AI applications at this time, though it has the potential to have a profound impact. Many applications of AI, including machine learning, deep learning, and natural language processing, are being used to create applications to enhance patient monitoring, optimize drug administration, improve decision-making, and predict postoperative complications.
1 Dec 2025
Read appraisal →Intensive Care Med
Critical care management of the patient with pharmaceutical poisoning
Poisonings account for up to 10% of admissions to intensive care units and represent a common yet complex challenge in critical care. While most patients recover with supportive care alone, a significant proportion require targeted interventions and advanced organ support. This state-of-the-art review provides a practical, evidence-based guide to manage pharmaceutical poisoning in the intensive care unit, with emphasis on real-world applicability for intensivists. It highlights toxicant-specific considerations for early resuscitation, structured by the ABCDE approach. Essential principles of the diagnostic workup are also summarized. In addition, the review offers actionable guidance on when and how to escalate care, moving from supportive measures to specific interventions including gastrointestinal decontamination, extracorporeal toxicant elimination, antidote therapy, and rescue strategies such as intravenous lipid emulsion and extracorporeal membrane oxygenation. Key algorithms and decision pathways-such as the diagnostic evaluation, ABCDE-based stabilization, gastrointestinal decontamination, and extracorporeal elimination approaches-are presented with visually structured figures. A comprehensive table provides an overview of the most relevant antidotes used in critical care, including indications and dosing.
1 Dec 2025
Read appraisal →