Research Appraisals
Evidence-based critical appraisals of the latest medical research, systematically evaluated using Oxford CEBM methodology.
Showing 5 appraisals
Minimally invasive therapy & allied technologies : MITAT : official journal of the Society for Minimally Invasive Therapy
Ventral hernia repair in emergency settings. A machine learning model to predict post-operative complications.
BACKGROUND: Emergency ventral hernia repair remains a challenging procedure due to patient instability, contaminated surgical fields, and heterogeneity in hernia types and operative techniques. Predicting postoperative complications in this setting is difficult using traditional statistical methods. Machine learning (ML) may offer improved predictive accuracy by recognizing nonlinear patterns among multiple perioperative factors. METHODS: A retrospective multicenter analysis was performed using data from the ACTIVE (Acute Treatment for Incisional Ventral Hernias) study, including 557 adult patients undergoing emergent ventral hernia repair between 2018 and 2021 in 31 Italian surgical centers. Demographic, preoperative, intraoperative, and postoperative variables were analyzed. Three ML algorithms-Decision Tree, Random Forest, and Deep Learning Neural Network-were trained and validated using five-fold cross-validation after class balancing with SMOTE. Model performance was compared with traditional logistic regression using accuracy, area under the ROC curve (AUC), and F1 score. RESULTS: Postoperative complications occurred in 181 patients (32.5%), while major complications (Clavien-Dindo ≥ II) occurred in 10%. Random Forest achieved the best performance (AUC 0.95, accuracy 0.88, F1 score 0.86), outperforming logistic regression (AUC 0.82, accuracy 0.78). The most influential predictors were operative duration, ASA score, and sepsis for overall complications, while bowel obstruction and BMI were key factors for major complications. Surgical approach (open vs. laparoscopic) did not independently correlate with adverse outcomes, highlighting the complexity of patient- and case-specific interactions. CONCLUSIONS: Machine learning models can accurately predict postoperative complications following emergent ventral hernia repair, surpassing traditional regression methods. These findings suggest that ML-based decision tools could support risk stratification and optimize surgical planning in high-risk emergency settings. Prospective validation is warranted to integrate AI-assisted prediction into perioperative clinical workflows.
3 Aug 2026
Read appraisal →International urology and nephrology
Acute kidney injury prediction and prognostication using machine learning
Acute Kidney Injury (AKI) is a major health concern with high costs and poor outcomes, partly due to late diagnosis. This paper reviews the application of artificial intelligence (AI) and machine learning (ML) to overcome these limitations by enabling earlier and more accurate prediction and prognostication of AKI. Many studies on AI and ML models used to predict and prognosticate AKI were included in the review. The focus was on models that analyze complex datasets, including real-time data streams like novel biomarkers and continuous vital signs, to achieve earlier and more accurate predictions than conventional methods. ML models show high predictive accuracy for AKI onset and outcomes across various clinical settings, including intensive care units, sepsis, and postoperative and postcontrast situations, with key findings like the successful integration of real-time data to reflect the evolving nature of kidney injury. AI and ML offer a powerful, proactive solution for AKI management. By leveraging diverse data, these technologies can significantly improve patient outcomes and reduce healthcare costs. While their promising performance warrants further exploration, successful clinical integration will require user-friendly platforms and continued validation.
1 Aug 2026
Read appraisal →JMIR formative research
Effects of Virtual Reality on Postoperative Pain Management Following Minimally Invasive Gynecologic Surgery: Randomized Controlled Trial
BACKGROUND: Postoperative pain and anxiety remain common concerns after minimally invasive gynecologic surgery despite advances in surgical techniques and analgesic strategies. Virtual reality (VR) has been investigated as a potential nonpharmacological intervention for pain management; however, evidence in gynecologic postoperative settings is limited. OBJECTIVE: This study aims to evaluate the efficacy and safety of VR technology compared with standard postoperative analgesia for pain and anxiety management in patients undergoing minimally invasive gynecologic surgery. METHODS: This randomized controlled trial was conducted at Sun Yat-sen Memorial Hospital of Sun Yat-sen University in China. A total of 131 patients undergoing laparoscopy or combined hysteroscopy for benign gynecologic diseases were randomly assigned in a 1:1 ratio to either a VR group (n=68) or a control group (n=63). All patients received a standardized general anesthesia protocol intraoperatively. The control group received conventional analgesic therapy after surgery, and the VR group received a 20-minute VR intervention 6 hours postoperatively. The pain and anxiety levels were evaluated using a visual analog scale at 6 and 7 hours postoperatively. The primary outcome was the change in pain scores between 6 and 7 hours. Secondary outcomes included maximum pain score, anxiety score changes, length of hospital stay, hospitalization costs, and occurrence of adverse events. Analyses were performed according to the intention-to-treat principle. RESULTS: There was no statistically significant difference in the primary outcome between the VR and control groups (mean difference 0.169, 95% CI -0.271 to 0.608; P=.45). Similarly, no significant differences were observed in the maximum pain score (mean difference 0.839, 95% CI -0.101 to 1.779; P=.08), and no improvement was observed in the anxiety score (mean difference 0.042, 95% CI -0.365 to 0.449; P=.84). No significant differences were found in length of hospital stay, hospitalization costs, or incidence of adverse events, including dizziness, nausea, and vomiting (all P>.05). CONCLUSIONS: A single 20-minute VR intervention did not provide additional analgesic or anxiolytic benefit compared with standard postoperative care after minimally invasive gynecologic surgery. VR was well tolerated, and its role in postoperative recovery requires further investigation. TRIAL REGISTRATION: Chinese Clinical Trial Registry ChiCTR2400091244; https://tinyurl.com/4b92a9td.
5 July 2026
Read appraisal →Lancet (London, England)
Diagnosis and treatment of postpartum haemorrhage: a race against time.
Postpartum haemorrhage (PPH) is common, affecting an estimated 13% of women having vaginal birth and 31% of women having caesarean birth. Successful management of PPH requires early and accurate diagnosis and effective treatment. A systematic review found that subjective visual estimation of blood loss misses 52% of PPH diagnoses at vaginal birth (pooled sensitivity 48%, 95% CI 44-53), and probably more at caesarean birth. The WHO-International Federation of Gynecology and Obstetrics-International Confederation of Midwives consolidated guidelines on PPH therefore recommend objective quantification of blood loss with products such as a calibrated blood collection drape. When supported by a robust implementation strategy and a first-response treatment bundle, objective measurement of blood loss and monitoring of vital signs has been shown to diagnose PPH accurately and early, and improve clinical outcomes. Refractory PPH can progress to life-threatening PPH, which should be managed by a multidisciplinary team providing aggressive resuscitation and targeted treatment. Saving the life of a woman with excessive postpartum bleeding is a race against time. The six delays to avoid are: (1) in the diagnosis (by use of objective cumulative blood loss measurement and early trigger criteria), (2) in the first-response treatment (by authorising midwives to administer all components of a standardised bundle of interventions), (3) in the escalation (by use of explicit escalation criteria and red flags), (4) in the use of temporising measures (eg, non-pneumatic anti-shock garment), (5) in the identification and targeted management of any specific causes of bleeding, and (6) in the provision of blood and blood products. Quick actions to avoid these delays can mean the difference between life and death for a woman with PPH.
4 July 2026
Read appraisal →Acta orthopaedica
Predicting persistent pain after total knee arthroplasty using different machine learning algorithms
BACKGROUND AND PURPOSE: After total knee arthroplasty (TKA), 10-20% of patients remain unsatisfied. Well-performing clinical prediction models can provide individualized risk estimates and stratification in terms of poor outcomes, resulting in unnecessary surgeries being avoided and patients being counseled preoperatively. We aimed to create a precise, well-performing prediction model for clinical application using different machine learning algorithms to predict those patients who will have residual pain, a low total Oxford Knee Score (OKS) and the patient group who do not achieve minimally clinical important difference (MCID) in OKS 1 year after TKA. METHODS: We conducted a retrospective cohort study based on patients who had undergone primary TKA at our institution combined with 751 patient-related variables. The multivariable models used were based on the results of univariate analysis. We used the machine learning method Extreme Gradient Boosting (XGBoost). The discrimination capability of the models was measured with the area under the curve (AUC). RESULTS: 11,755 patients were included in this study. There were 850 (7.2%) patients who experienced persistent pain 1 year after TKA. The AUC was 0.67. For the secondary outcomes, the AUC values were similar. The most important variables in the model were lower preoperative OKS, younger age, valgus malalignment, lower preoperative pain OKS, use of mild opioid, neuropathic pain medicine and thyroxine, and higher body mass index. CONCLUSION: The prediction models achieved poor AUCs. It seems clear that the prediction of pain and functional outcome after TKA is difficult, even with a large patient cohort combined with 751 patient-related variables and sophisticated machine-learning algorithms.
23 June 2026
Read appraisal →