Validation of a natural language processing algorithm using national reporting data to improve identification of anesthesia-related ADVerse evENTs: The "ADVENTURE" study.

BACKGROUND: Reporting and analysis of adverse events (AE) is associated with improved health system learning, quality outcomes, and patient safety. Manual text analysis is time-consuming, costly, and prone to human errors. We aimed to demonstrate the feasibility of novel machine learning and natural language processing (NLP) approaches for early predictions of adverse events and provide input to direct quality improvement and patient safety initiatives. METHODS: We used machine learning to analyze 9559 continuously reported AE by clinicians and healthcare systems to the French National Health accreditor (HAS) between January 1, 2009, and December 31, 2020 . We validated the labeling of 135,000 unique de-identified AE reports and determined the associations between different system's root causes and patient consequences. The model was validated by independent expert anesthesiologists. RESULTS: The machine learning (ML) and Artificial Intelligence (AI) model trained on 9559 AE datasets accurately categorized 8800 (88%) of reported AE. The three most frequent AE types were "difficult orotracheal intubation" (16.9% of AE reports), "medication error" (10.5%), and "post-induction hypotension" (6.9%). The accuracy of the AI model reached 70.9% sensitivity, 96.6% specificity for "difficult intubation", 43.2% sensitivity, and 98.9% specificity for "medication error." CONCLUSIONS: This unsupervised ML method provides an accurate, automated, AI-supported search algorithm that ranks and helps to understand complex risk patterns and has greater speed, precision, and clarity when compared to manual human data extraction. Machine learning and Natural language processing (NLP) models can effectively be used to process natural language AE reports and augment expert clinician input. This model can support clinical applications and methodological standards and used to better inform and enhance decision-making for improved risk management and patient safety. TRIAL REGISTRATION: The study was approved by the ethics committee of the French Society of Anesthesiology (IRB 00010254-2020-20) and the CNIL (CNIL: 118 58 95) and the study was registered with ClinicalTrials.gov (NCT: NCT05185479).

Peripheral obstruction without airflow limitation is rare and not specific to asthma in children.

Peripheral (or small) airway obstruction (PAO) is considered a marker of childhood asthma but the techniques able to directly measure it are rarely used in routine. Usual spirometry and plethysmography can detect a certain degree of PAO when reduced forced vital capacity (FVC) is associated to normal forced expiratory volume in 1 s (FEV1 ) to FVC ratio, and normal total lung capacity (TLC). The frequency of this functional pattern has never been studied in different pediatric respiratory conditions. To assess the prevalence and outcome of PAO in children with different diseases or symptoms, we retrospectively extracted from our database all files of Caucasian subjects encompassing spirometry and plethysmography measurements. Spirometry patterns (normal, airflow limitation [AFL; low FEV1 /FVC], low FVC [with normal FEV1 /FVC]) and final functional patterns (normal, AFL, PAO, restrictive [low TLC], or mixed) were described. We included 4394 files recorded in 1794 children (median [IQR] age: 10.7 [9.2-12.9] years). At inclusion, 125 (7%) children had low FVC of which 56 (44.8%, and 3.1% [95% CI 2.3-3.9] of the whole population) had PAO. PAO prevalence increased with age (OR (95% CI) per 2-year-increase: 2.26 (1.59-3.23); p < .001), and was more frequent in chronic bronchial diseases other than asthma (1.6% vs. 8.0%). On repeated tests, PAO frequently normalized (26.1%) or persisted (43.5%), but it less often progressed into AFL (13%) or restrictive (13%) patterns. PAO is an infrequent nonspecific and persistent functional pattern. Its prevalence increases with age and in diagnosis of chronic respiratory diseases other than asthma.

Experimental Evaluation of Endovascular Fenestration Scissors in an Ovine Model of Aortic Dissection.

OBJECTIVE/BACKGROUND: To evaluate the experimental feasibility of endovascular fenestration using specific endovascular scissor prototypes in an ovine model of acute aortic dissection (AD). METHODS: A previously described endovascular technique was used to create a model of acute type B AD in sheep. Endovascular fenestrations using either endovascular scissor prototypes or a long sheath were compared. Four prototypes of endovascular fenestration scissors were evaluated. Both validity of the experimental model of AD and technical success of endovascular fenestration were assessed by haemodynamic criteria, completion angiography, transesophageal echocardiography, and post-procedural analysis of harvested aortas. RESULTS: Experimental acute AD was created by endovascular means in 17 sheep, with a technical success rate of 82%. Systolic blood pressure was lower in the false lumen than in the true lumen (58 ± 5 vs. 79 ± 3 mmHg, respectively; p < .001). Endovascular fenestration was performed in 11 models (endovascular scissors n = 8; long sheath n = 3). Controlled endovascular fenestration was obtained by the use of endovascular scissors (n = 5/8), resulting in a significant rise in false lumen systolic blood pressure after fenestration (60 ± 2 vs. 67 ± 9 mmHg before and after fenestration, respectively; p < .047). Long sheath fenestration resulted in an uncontrolled flap motion, leading to either pseudo-coarctation syndrome or aortic rupture (58 ± 6 vs. 40 ± 2 mmHg before and after fenestration, respectively; p < .001). CONCLUSION: In this experimental study, a reproducible AD model has been developed in sheep using endovascular procedures exclusively to evaluate endovascular fenestration techniques. Endovascular fenestration using a long sheath appeared hazardous and risky in vivo. Endovascular scissors constitute a dedicated and suitable tool to perform a safe controlled and effective endovascular fenestration in an ovine model.

Randomized controlled trials in pediatric patients had higher completion rates than adult trials: a cross-sectional study.

OBJECTIVE: Conduct of clinical trials is perceived to be more challenging in children than in adults. This study aimed to evaluate the impact of the age of participants on completion rates of randomized controlled trials (RCTs). STUDY DESIGN AND SETTING: A cross-sectional study on RCTs registered in the ClinicalTrials.gov database. All RCTs registered up to December 31, 2016, were extracted and were classified according to their recruitment status: active, completed, or discontinued and according to the age of participants: children (<17 years), adults (≥18 years), and mixed-age population. A logistic regression model was applied to assess the impact of participant's age category on trial completion while controlling for other relevant trial features. RESULTS: A total of 65,095 registered RCTs were identified. Among pediatric trials, 49.9% were completed and 8.5% were discontinued. Among adult and mixed age RCTs, respectively, 49.7% and 47.9% were completed whereas, 10.2% and 9.4% were discontinued. Overall, pediatric and mixed age RCTs were more likely to be registered as completed than adult RCTs (odds ratio: 1.16, 95% CI: 1.02-1.30; odds ratio: 1.15, 95% CI: 1.04-1.27, respectively). Also, funding source, type of intervention under evaluation, primary trial purpose, use of a blinding procedure, use of a placebo, and participants' assignment model were identified as independent predictors of RCT completion. CONCLUSION: Contrary to current perceptions and despite several specific challenges, recruitment of children and adolescents is not a limiting factor to completing a RCT. Other study features such as funding source, impact completeness and should be carefully considered before initiating research.