Clinical Decision Research Articles

Analytical and clinical evaluations of Snibe Maglumi® S100B assay.

To assess the analytical performances of Snibe Maglumi® S100 assay and compare it with the Roche Elecsys® S100B assay in adults with mild traumatic brain injury (mTBI) focusing on reducing unnecessary cranial computed tomography (CT) scans per Scandinavian and French guidelines. Analytical performance of the Maglumi® S100 kit was assessed using quality controls from both Snibe and Roche, as well as pooled serums. Clinical performances were assessed using serum from 89 adult mTBI patients presenting to the adult emergency department of Clermont-Ferrand University Hospital with a Glasgow Coma Scale score of 14-15. CT scans were performed according to the Elecsys® S100 measurement, with a decision threshold of 0.10 μg/L. Repeatability and reproducibility coefficients of variation determined using Elecsys® S100B, Maglumi® S100 controls and pooled serums were below 8 %. Six (7 %) mTBI patients included had clinically relevant intracranial lesions observed on CT scan (CT+), and eighty-three (93 %) patients had no lesions (CT-). S100B medians in CT- and CT+ patients were significantly different: 0.125 (0.085-0.219) vs. 0.368 (0.231-0.489) (p=0.006) for Elecsys®, and 0.073 (0.046-0.140) vs. 0.327 (0.230-0.353) for Maglumi® (p=0.004). The areas under the ROC curves for intracranial lesion detection were similar: 0.82 (0.73-0.91; p=0.0084) and 0.83 (0.75-0.92; p=0.0063) for Elecsys® and Maglumi®, respectively. The Maglumi® S100B assay can be used in the management of mTBI patients to exclude unnecessary CT scans. Further studies are needed to validate a clinical decision threshold for CT scan decisions.

The feasibility of preventing Hospital-Acquired Pressure Ulcers (HAPUs)

Introduction: Hospital-acquired pressure ulcers (HAPUs) are a significant healthcare issue, leading to physical complications, extended hospital stays, and increased healthcare costs. Preventive measures are crucial for managing this problem. With the growing integration of technology in healthcare, mobile e-book applications and chatbots offer promising tools for improving clinical practice and decision-making. This study evaluates the feasibility of using these technologies in terms of the Technology Acceptance Model (TAM), focusing on perceived usefulness, ease of use, and intention to use among nurses. Methods: A quantitative research design was employed to assess the perceptions of 30 registered nurses working in a hospital setting. The Technology Acceptance Model (TAM) framework guided the measurement of key variables, including perceived usefulness, ease of use, and intention to use. Data were collected through a self-administered online survey and semi-structured interviews. Descriptive statistics and thematic analysis were used to analyze the data. Results: The study revealed positive perceptions of the chatbot and mobile e-book app in terms of their usefulness, ease of use, and intention to use. The mean scores for usefulness, ease of use, and intention to use were 16.55 ± 1.50, 17.32 ± 0.79, and 13.03 ± 1.74, respectively, indicating high feasibility for adoption. Discussion: The findings suggest that mobile e-books and chatbots are well-received by healthcare professionals, with perceived usefulness and ease of use being significant factors for their adoption. These technologies can support clinical decision-making, reduce medical errors, and improve healthcare outcomes. Future developments should focus on integrating these tools into healthcare training programs and enhancing their functionality with real-time updates and interactive features. Keywords: Hospital-acquired pressure ulcers, Technology Acceptance Model, mobile e-books, chatbots, healthcare technology.

Structured reporting of neuroendocrine tumors in PET/CT using [18F]SiTATE - impact on interdisciplinary communication

Our retrospective single-center study aims to evaluate the impact of structured reporting (SR) using a self-developed template on report quality compared to free-text reporting (FTR) in [18F]SiTATE Positron Emission Tomography/Computer Tomography (PET/CT) for the primary staging and therapy monitoring of patients diagnosed with neuroendocrine tumors (NET). In total 50 patients were included. FTRs and SRs were generated post-examination. All reports were evaluated by a radiologist and a surgeon through a questionnaire to determine their contribution to facilitating clinical decision-making and to assess their completeness, linguistic quality, and overall quality. SR significantly increased the capacity of facilitating therapy decision-making from 32% in FTR to 55% in SR (p < 0.001). Trust in the report was significantly higher in SR with a mean of 5.0 (SD = 0.5) vs. 4.7 (SD = 0.5) for FTR (p < 0.001). SR received significantly higher mean ratings regarding linguistic quality with 4.7 for SR vs. 4.4 for FTR (p = 0.004) and overall report quality with a mean of 4.9 for SR vs. 4.6 for FTR (p < 0.001). Concluding that SR enhances the overall quality of reports in [18F]SiTATE-PET/CTs for NET staging, serving as a tool to streamline clinical decision-making and enhance interdisciplinary communication in the future.

Flow cytometry evaluation of acute myeloid leukemia minimal residual disease based on an understanding of the normal maturation patterns in the blast compartments.

Detection of minimal/measurable disease (MRD) in acute myeloid leukemia (AML) is critical for both clinical decision-making and prognostication, yet remains a challenge. Flow cytometry is a well-established method for MRD detection. Flow cytometric (FC) evaluation of MRD must consider a complex maturational pattern of normal hematopoietic development to separate normal from abnormal progenitors. Here, we offer an example of an interpretive approach based on a thorough understanding of stage- and lineage-specific hematopoietic maturation. We provide a comprehensive overview of blast maturation from early precursors (hematopoietic stem cells) to committed late-stage unilineage progenitors and commonly observed stage-specific abnormalities based on cases we have encountered in practice. We emphasize the importance of stage-specific comparisons for accurate MRD detection by flow cytometry. The AML blasts almost invariably show abnormal phenotypes, and the phenotypes may evolve upon therapy. The detected phenotypes are necessarily confined to the target antigens included in the panel. It is therefore critical to evaluate a range of antigens to establish a specific stage/state of lineage commitment and detect potential common abnormalities. Moreover, enough cells must be acquired to allow for the detection of MRD at desired levels. Significant technical and analytical validation is critical. Flow cytometry offers a powerful single-cell-based platform for MRD detection in AML, and the results have been proven critical for disease management. Leukemia-associated phenotype-informed difference from the normal approach presented in this review presents an analytical framework for sensitive and accurate MRD detection.

Risk prediction models for stress urinary incontinence after pelvic organ prolapse (POP) surgery: a systematic review and meta-analysis

ObjectiveTo systematically evaluate existing developed and validated predictive models for stress urinary incontinence after pelvic floor reconstruction.MethodsRelevant literature in PubMed, Embase, Web of Science, Cochrane Library, OVID, China National Knowledge Infrastructure(CNKI), Wan Fang Database, VIP database and Chinese Biomedical Literature Service System (SinoMed) were search from inception to 1 March 2024. Literature screening and data extraction were performed independently by two researchers. The chosen study’s statistics included study design, data sources, outcome definitions, sample size, predictors, model development, and performance. The Predictive Modelling Risk of Bias Assessment Tool (PROBAST) checklist was used to assess risk of bias and applicability.ResultsA total of 7 studies containing 9 predictive models were included. All studies had a high risk of bias, primarily due to retrospective design, small sample sizes, single-center trials, lack of blinding, and missing data reporting. The meta-analysis revealed moderate heterogeneity (I² = 68.8%). The pooled AUC value of the validated models was 0.72 (95% CI: 0.65, 0.79), indicating moderate predictive ability.ConclusionThe prediction models evaluated demonstrated moderate discrimination, but significant bias and methodological flaws. The meta-analysis revealed moderate heterogeneity (I² = 68.8%) among the included studies, reflecting differences in study populations, predictors, and methods, which limits the generalizability of the findings. Despite these challenges, these models highlight the potential to identify high-risk patients for targeted interventions to improve surgical outcomes and reduce postoperative complications. The findings suggest that by integrating these models into clinical decision-making, clinicians can better tailor surgical plans and preoperative counseling, thereby improving patient satisfaction and reducing the incidence of postoperative stress urinary incontinence. Future research should follow TRIPOD and PROBAST principles, focus on addressing sources of heterogeneity, improve model development through robust designs, large sample sizes, comprehensive predictors, and novel modelling approaches, and validate tools that can be effectively integrated into clinical decision-making to manage stress urinary incontinence after pelvic floor reconstruction.

Characterization of studies considered and required under Medicare's coverage with evidence development program.

In 2005, the Centers for Medicare and Medicaid Services introduced the Coverage with Evidence Development program for items and services with limited evidence of benefit and harm for Medicare beneficiaries, aiming to generate evidence to determine whether they meet the statutory "reasonable and necessary" criteria for coverage. Coverage with Evidence Development requires participation in clinical studies approved by the Centers for Medicare and Medicaid Services (i.e. Coverage with Evidence Development-approved studies) as a condition of coverage. We examined the quality of evidence generated by Coverage with Evidence Development-approved studies compared with those that informed Centers for Medicare and Medicaid Services' initial Coverage with Evidence Development decisions (i.e. National Coverage Determination studies). Using Centers for Medicare and Medicaid Services' webpage, we identified all items and services covered under Coverage with Evidence Development and their Coverage with Evidence Development-approved studies. Through searching PubMed and Google Scholar, we identified original research articles that reported results for primary endpoints of Coverage with Evidence Development-approved studies. We then reviewed the initial Coverage with Evidence Development decision memos and identified National Coverage Determination studies that were original research.We characterized and compared Coverage with Evidence Development-approved studies and National Coverage Determination studies. From 2005 to 2023, 26 items and services were covered under the Coverage with Evidence Development program, associated with 196 National Coverage Determination studies (170 (86.7%) clinical trials and 26 (13.3%) registries) and 116 unique Coverage with Evidence Development-approved studies (86 (74.1%) clinical trials, 23 (19.8%) registries, 4 (3.4%) claims-based studies, and 3 (2.6%) expanded access studies). Among clinical trial studies, National Coverage Determination studies and Coverage with Evidence Development-approved studies did not differ with respect to multi-arm design (59.4% vs 68.6%; p = 0.15). However, among multi-arm clinical trial studies, National Coverage Determination studies were less likely than Coverage with Evidence Development-approved studies to be randomized (52.5% vs 93.2%; p < 0.001). Overall, National Coverage Determination studies less frequently had ≥ 1 primary endpoint focused on a clinical outcome measure (65.8% vs 87.9%; p = 0.006) and less frequently exclusively enrolled Medicare beneficiaries (3.1% vs 25.9%; p < 0.001). In addition, National Coverage Determination studies had smaller population sizes than Coverage with Evidence Development-approved studies (median 100 (interquartile range, 45-414) vs 302 (interquartile range, 93-1000) patients; p = 0.002). Among Coverage with Evidence Development-approved studies, 59 (50.9%) had not yet publicly reported results for the primary endpoint. Studies required under Medicare's Coverage with Evidence Development program more often used randomized study designs, had larger patient populations, enrolled US patients, and focused on clinical outcomes as primary endpoints than studies used to inform initial National Coverage Determinations. However, not all Coverage with Evidence Development-approved studies have thus far reported results, which could create uncertainty for patients, physicians, and payers regarding the clinical benefits of covered items and services. The Centers for Medicare and Medicaid Services' Coverage with Evidence Development program has successfully prompted generation of more robustly designed clinical studies to better inform clinical, regulatory, and coverage decisions compared to studies informing initial coverage decisions. Yet, opportunities exist to further strengthen the design and dissemination of studies required under this program.

Clinical decision fatigue: a systematic and scoping review with meta-synthesis

ObjectiveDecision fatigue (DF) can lead to impaired judgement, decreased diagnostic accuracy and increased likelihood of medical errors. Research on DF is scarce, and little is known about its nature in...

Fuzzy Hybrid Meta-optimized Learning-based Medical Image Segmentation System for Enhanced Diagnosis

This medical image segmentation plays a fundamental role in the diagnosis of diseases related to the correct identification of internal structures and pathological regions in different imaging modalities. The conventional fuzzy-based segmentation approaches, though quite useful, still have some drawbacks regarding handling uncertainty, parameter optimization, and high accuracy of segmentation with diverse datasets. Because of these facts, it generally leads to poor segmentations, which can give less reliability to the clinical decisions. In addition, the paper is going to propose a model, FTra-UNet, with advanced segmentation of medical images by incorporating fuzzy logic and transformer-based deep learning. The model would take complete leverage of the strengths of FIS concerning the handling of uncertainties in segmentation. Besides, it integrates SSHOp optimization technique to fine-tune the weights learned by the model to ensure improvement in adaptability and precision. These integrated techniques ensure faster convergence rates and higher accuracy of segmentation compared to state-of-the-art traditional methods. The proposed FTra-UNet is tested on BRATS, CT lung, and dermoscopy image datasets and ensures exceptional results in segmentation accuracy, precision, and robustness. Experimental results confirm that FTra-UNet yields consistent, reliable segmentation outcomes from a practical clinical application perspective. The architecture and implementation of the model, with the uncertainty handled by FIS and the learning parameters optimization handled by the SSHOp method, increase the power of this model in segmenting medical images.

Evaluation of statin-induced muscle and liver adverse drug reactions in the Chinese population: a retrospective analysis of clinical trial data from 1992 to 2023

ObjectivesThis study addressed the gaps in the disclosure of statin-associated adverse drug reactions (ADRs) in China’s official database and the inadequacy of cases reported relative to the population size in...

Beyond the upgrade: unraveling the complexities of health information system migration

The need for efficient health information systems (HIS) becomes more critical as the healthcare industry and information technology era evolve. Some hospitals opt for in-house HIS implementation, while others prefer the software-as-a-service (SaaS) model. Although one solution may be more cost-effective than others, all HIS operations share the need for numerous management efforts to maintain regulatory compliance and ensure integrity with data manipulation processes. This becomes more significant during HIS migration when a hospital undergoes an extensive data transition from one system to another. In other words, hospital IT and service providers must ensure the seamless flow of patient data and system users during the HIS migration. This paper presents a real-life case study, including a framework of the implementation method, on a complete switch from a two-decade-old legacy HIS system to a state-of-the-art web-based HIS system in a non-incremental approach. Even though the HIS migration project achieved its primary objectives and can be deemed successful, the implementation process encountered several hurdles requiring problem-solving, decision-making and adaptability skills. Examples of such challenges include data integrity issues, user resistance, and the complexities of integrating multiple hospital departments. This case study highlights in a detailed framework how careful planning, stakeholder engagement, thorough testing, and continuous support were essential to overcoming obstacles. The key outcomes of this research besides the migration framework include an improvement in data accuracy, better clinical decision support, and enhanced operational efficiency, which could offer lessons for healthcare organizations or similar sectors looking to transition from legacy systems to modern ones.

Good-Quality mHealth Apps for Endometriosis Care: Systematic Search.

Mobile health (mHealth) apps are increasingly being used by community members to track symptoms and manage endometriosis. In addition, clinicians use mHealth apps for continued medical education and clinical decision-making and recommend good-quality apps to patients. However, poor-quality apps can spread misinformation or provide recommendations that are not evidence-based. Therefore, a critical evaluation is needed to assess and recommend good-quality endometriosis mHealth apps. This study aimed to evaluate the quality and provide recommendations for good quality endometriosis mHealth apps for the community and clinicians. PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) 2020 guidelines informed the search of mHealth apps on the Google Play Store and Apple App Store. The search terms included "endometriosis," "adenomyosis," and "pelvic pain." mHealth apps were eligible if they were (1) related to the search terms, (2) were in the English language, and (3) were available free of cost. Only the free content of the eligible mHealth apps was assessed. ENLIGHT, a validated evaluation tool for mobile and web-based interventions, was used to assess the quality across 7 domains such as usability, visual design, user engagement, content, therapeutic persuasiveness, therapeutic alliance, and general subjective evaluation. mHealth apps with a total score of ≥3.5 were classified as "good" according to the ENLIGHT scoring system and are recommended as good-quality mHealth apps for endometriosis care. In total, 42 mHealth apps were screened, and 19 were included in the quality assessment. A total of 6 good-quality mHealth apps were identified (QENDO, Bearable, Luna for Health, Matilda Health, Branch Health: Pain Management, and CHARLI Health). These apps provided symptom-tracking functions and self-management support. A total of 17 apps were designed for community use, while 2 apps provided a digital endometriosis classification tool to clinicians. Most mHealth apps scored well (≥3.5) in the domains of usability (16/19, 84.2%), visual design (14/19, 73.7%), user engagement (11/19, 57.9%), and content (15/19, 78.9%). Few eHealth websites scored well on therapeutic persuasiveness (6/19, 31.6%), therapeutic alliance (9/19, 47.4%), and general subjective evaluation (6/19, 31.6%). Although time and geographical location can influence the search results, we identified 6 "good-quality" endometriosis mHealth apps that can be recommended to the endometriosis community. mHealth apps designed for community use should evaluate their effectiveness on user's endometriosis knowledge, self-recommended management strategies, pain self-efficacy, user satisfaction, and user quality of life. Digital technology should be leveraged to develop mHealth apps for clinicians that contribute to continued medical education and assist clinical decision-making in endometriosis management. Factors that enhance usability, visual design, therapeutic persuasiveness, and therapeutic alliance should be incorporated to ensure successful and long-term uptake of mHealth apps. PROSPERO CRD42020185475; https://tinyurl.com/384dkkmj.

Diffusion transformer model with compact prior for low-dose PET reconstruction

Objective.Positron emission tomography (PET) is an advanced medical imaging technique that plays a crucial role in non-invasive clinical diagnosis. However, while reducing radiation exposure through low-dose PET scans is beneficial for patient safety, it often results in insufficient statistical data. This scarcity of data poses significant challenges for accurately reconstructing high-quality images, which are essential for reliable diagnostic outcomes.Approach.In this research, we propose a diffusion transformer model (DTM) guided by joint compact prior to enhance the reconstruction quality of low-dose PET imaging. In light of current research findings, we present a pioneering PET reconstruction model that integrates diffusion and transformer models for joint optimization. This model combines the powerful distribution mapping abilities of diffusion model with the capacity of transformers to capture long-range dependencies, offering significant advantages for low-dose PET reconstruction. Additionally, the incorporation of the lesion refining block and alternating direction method of multipliers enhance the recovery capability of lesion regions and preserves detail information, solving blurring problems in lesion areas and texture details of most deep learning frameworks.Main results. Experimental results validate the effectiveness of DTM in reconstructing low-dose PET image quality. DTM achieves state-of-the-art performance across various metrics, including PSNR, SSIM, NRMSE, CR, and COV, demonstrating its ability to reduce noise while preserving critical clinical details such as lesion structure and texture. Compared with baseline methods, DTM delivers best results in denoising and lesion preservation across various low-dose levels, including 10%, 25%, 50%, and even ultra-low-dose level such as 1%. DTM shows robust generalization performance on phantom and patient datasets, highlighting its adaptability to varying imaging conditions.Significance. This approach reduces radiation exposure while ensuring reliable imaging for early disease detection and clinical decision-making, offering a promising tool for both clinical and research applications.

Validation of the shortest version of the eating attitude test (EAT-7) as a screening tool for disordered eating in patients with first-episode schizophrenia

BackgroundIndividuals diagnosed with schizophrenia have a two to five-fold higher risk of developing obesity and up to threefold higher risk of developing disordered eating behaviors relative to the general population. Over the past decades, the Eating Attitude Test (EAT) has been well-recognized and widely used in clinical practice. However, only little psychometric information is available on the EAT for researchers and clinicians who deal with patients with schizophrenia, and it is still unclear whether the scale could fulfill measurement needs and help clinical and research decision-making in the context of psychotic disorders. This study aimed to investigate the psychometric properties of the shortest version of the EAT (EAT-7) among newly-diagnosed patients with schizophrenia who had minimal exposure to antipsychotics at the start of the study.MethodsThis is an observational, cross-sectional survey that was carried out in the department of psychiatry at Razi Hospital, Manouba, Tunisia. The study took place between January and June 2024, and involved clinically-stabilized outpatients with schizophrenia who have had < 3 months of treatment with antipsychotics (N = 112; mean age of 24.44 ± 5.41 years).ResultsIn terms of factor analysis, the EAT-7 presented a one-dimensional structure. All seven items strongly loaded on a single factor, with all indices of the CFA suggesting a good fit to the data. In addition, the EAT-7 yielded excellent reliability coefficients, with both a McDonald’s ω and a Cronbach’s α of 0.88. Measurement invariance of the EAT-7 across sex groups was tested using multi-group CFA, and established at the configural, scalar, and metric levels. No significant differences in EAT-7 scores between males and females were found. Finally, EAT-7 scores positively correlated with scores of depression and anxiety, supporting concurrent validity of the scale.ConclusionThe shortness, items’ clarity and conciseness of the EAT-7 make it an efficient tool appropriate as a first-step screening tool to detect disordered eating in patients with schizophrenia at low cost and burden. We hope that this study will facilitate the widespread application of the EAT-7 in routine assessment and monitoring of disordered eating in patients with schizophrenia, in both clinical and research practices.

Trends and Perioperative Outcomes of Surgical Treatments for Benign Prostatic Hyperplasia in Germany: Results from the GRAND Study.

Real-world data exploring the perioperative outcomes of different surgical treatments for benign prostatic hyperplasia (BPH) are lacking. We aimed to assess the trends of BPH surgeries and compare their perioperative outcomes. We used the GeRmAn Nationwide inpatient Data (GRAND), provided by the Research Data Center of the Federal Bureau of Statistics (2005-2022), and performed multiple patient-level analyses. Our study included 1 355 845 surgical treatments for BPH. Of these, 1 084 650 (79.9%) were transurethral resection of the prostate (TURP), 90 735 (6.9%) simple prostatectomy, 64 325 (4.7%) holmium laser enucleation of the prostate (HoLEP), 58 406 (4.3%) laser vaporization, 25 747 (1.9%) electrovaporization, 15 241 (1.1%) thulium laser enucleation of the prostate (ThuLEP), 7873 (0.58%) prostate incision, 3298 (0.24%) water ablation, 2724 (0.2%) water vapor thermal therapy, 1235 (0.09%) prostate stent, 1100 (0.08%) transurethral needle ablation of the prostate, and 511 (0.03%) prostatic urethral lift. The use of HoLEP, ThuLEP, water ablation, and water vapor thermal therapy has increased exponentially in the last years. After adjusting for the major risk factors in the multivariate regression analysis, simple prostatectomy was associated with the worst outcomes compared with TURP, followed by electrovaporization. On the contrary, HoLEP and ThuLEP were associated with favorable outcomes, in terms of sepsis, ureteral stent placement, transfusion, admission to the intensive care unit, hospital stay, and mortality, compared with TURP. Similarly, minimally invasive surgical therapies were associated with high safety compared with TURP. Nevertheless, our findings are based on retrospective billing data and are prone to a selection bias. The trends and perioperative outcomes of BPH surgery should be taken into consideration to improve clinical decision-making and patient outcomes.

Leveraging transformer models to predict cognitive impairment: accuracy, efficiency, and interpretability

ObjectiveThis study aims to develop an enhanced Transformer model for predicting mild cognitive impairment (MCI) using data from the China Health and Retirement Longitudinal Study (CHARLS), focusing on handling mixed data types and improving predictive accuracy.MethodsThe Transformer integrates categorical (integer-encoded) and continuous (floating-point) data, using multi-head attention with four heads to capture complex relationships. Preprocessing involved separate embedding layers for categorical data and feed-forward networks for continuous data. The model was compared with SVM and XGBoost, trained for 150 epochs with RMSProp and a cosine annealing scheduler. Key metrics included accuracy, Mean Absolute Error (MAE) tolerance, and training loss. An attention heatmap was generated to visualize feature importance.ResultsThe Transformer outperformed SVM and XGBoost, achieving over 90% accuracy at an MAE tolerance of 3.5. The model showed rapid convergence, with loss stabilizing within 20 epochs. The attention heatmap highlighted key features, confirming the effectiveness of the multi-head attention mechanism in identifying relevant variables.ConclusionThe enhanced Transformer model offers superior accuracy and efficiency in predicting cognitive decline compared to traditional models. Its capacity to process both continuous and categorical data and its interpretability through attention mechanisms make it a promising tool for early detection of neurodegenerative diseases, potentially improving clinical decision-making and interventions.

Establishing methodological standards for the development of artificial intelligence-based Clinical Decision Support in emergency medicine.

Artificial intelligence (AI) offers opportunities for managing the complexities of clinical care in the emergency department (ED), and Clinical Decision Support has been identified as a priority application. However, there is a lack of published guidance on how to rigorously develop and evaluate these tools. We sought to answer the question, "What methodological standards should be applied to the development of AI-based Clinical Decision Support tools in the ED?". We conducted an iterative consensus-establishing activity involving a subcommittee with AI expertise followed by surveys and a live facilitated discussion with participants of the 2024 Canadian Association of Emergency Physicians Research Symposium in Saskatoon. We augmented analysis of participant feedback with large language models. We established 11 recommendations AI-based Clinical Decision Support development including the selection of a relevant problem and team of experts, standards of data quality and quantity, novel AI-specific reporting guidelines, and adherence to principles of ethics and privacy. We removed the recommendation regarding model interpretability from the final list due to a lack of consensus. These 11 recommendations provide guiding principles and methodological standards for emergency medicine researchers to rigorously develop AI-based Clinical Decision Support tools and for clinicians to gain knowledge and trust in using them.

Predicting hematoma expansion after intracerebral hemorrhage: a comparison of clinician prediction with deep learning radiomics models.

Early prediction of hematoma expansion (HE) following nontraumatic intracerebral hemorrhage (ICH) may inform preemptive therapeutic interventions. We sought to identify how accurately machine learning (ML) radiomics models predict HE compared with expert clinicians using head computed tomography (HCT). We used data from 900 study participants with ICH enrolled in the Antihypertensive Treatment of Acute Cerebral Hemorrhage 2 Study. ML models were developed using baseline HCT images, as well as admission clinical data in a training cohort (n = 621), and their performance was evaluated in an independent test cohort (n = 279) to predict HE (defined as HE by 33% or > 6mL at 24h). We simultaneously surveyed expert clinicians and asked them to predict HE using the same initial HCT images and clinical data. Area under the receiver operating characteristic curve (AUC) were compared between clinician predictions, ML models using radiomic data only (a random forest classifier and a deep learning imaging model) and ML models using both radiomic and clinical data (three random forest classifier models using different feature combinations). Kappa values comparing interrater reliability among expert clinicians were calculated. The best performing model was compared with clinical predication. The AUC for expert clinician prediction of HE was 0.591, with a kappa of 0.156 for interrater variability, compared with ML models using radiomic data only (a deep learning model using image input, AUC 0.680) and using both radiomic and clinical data (a random forest model, AUC 0.677). The intraclass correlation coefficient for clinical judgment and the best performing ML model was 0.47 (95% confidence interval 0.23-0.75). We introduced supervised ML algorithms demonstrating that HE prediction may outperform practicing clinicians. Despite overall moderate AUCs, our results set a new relative benchmark for performance in these tasks that even expert clinicians find challenging. These results emphasize the need for continued improvements and further enhanced clinical decision support to optimally manage patients with ICH.

Discrepancies in safety reporting for chronic back pain clinical trials: an observational study from ClinicalTrials.gov and publications

IntroductionChronic back pain (CBP) is a leading cause of disability worldwide and is commonly managed with pharmacological, non-pharmacological, and procedural interventions. However, adverse event (AE) reporting for these therapies often lacks transparency, raising concerns about the accuracy of safety data. This study aimed to quantify inconsistencies in AE reporting between ClinicalTrials.gov and corresponding randomized controlled trial (RCT) publications, emphasizing the importance of comprehensive safety reporting to improve clinical decision-making and patient care.MethodsWe retrospectively analyzed Phase 2–4 CBP RCTs registered on ClinicalTrials.gov from 2009 to 2023. Extracted data included AE reporting, trial sponsorship, and discrepancies in serious adverse events (SAEs), other adverse events (OAEs), mortality, and treatment-related withdrawals between registry entries and publications. Statistical analyses assessed reporting inconsistencies, following STROBE guidelines.ResultsA total of 114 registered trials were identified, with 40 (35.1%) corresponding publications. Among these, 67.5% were industry-sponsored. Only 4 (10%) publications fully reported adverse events (AEs) without discrepancies, while 36 (90%) contained at least one inconsistency compared to ClinicalTrials.gov. Discontinuation due to AEs was explicitly reported in 24 (60%) of ClinicalTrials.gov entries and in 30 (75%) of publications, with discrepancies in 16 trials (40%). Serious adverse events (SAEs) were reported differently in 15 (37.5%) publications; 80% reported fewer SAEs than ClinicalTrials.gov. Other adverse events (OAEs) showed discrepancies in 37 (92.5%) publications, with 43.2% reporting fewer and 54.1% reporting more OAEs.DiscussionThis study highlights pervasive discrepancies in AE reporting for CBP trials, undermining the reliability of published safety data. Inconsistent reporting poses risks to clinical decision-making and patient safety. Adopting standardized reporting guidelines, such as CONSORT Harms, and ensuring transparent updates in publications could enhance the accuracy and trustworthiness of safety data. Journals and regulatory bodies should enforce compliance and future efforts should develop mechanisms to monitor and correct reporting inconsistencies, enhancing the trustworthiness of safety data in clinical research.

Preoperative screening thoracic radiographs yield few significant abnormalities in dogs with no history or exam findings suggestive of thoracic disease.

To evaluate the utility of routine preanesthetic screening thoracic radiographs in dogs without a specific clinical indication for this test. All patients presented for elective tibial plateau leveling osteotomy between January 1, 2022, and December 31, 2022, were enrolled if there were no clinical signs or history of neoplasia or cardiopulmonary or thoracic disease. Three-view thoracic radiographic studies were performed and evaluated by the attending surgeon and a board-certified veterinary radiologist. The incidence of radiographic abnormalities and agreement between the attending clinician and the radiologist were analyzed. Of the 281 cases evaluated in this study, 10 (3.6%) were found to have radiographic abnormalities that were likely to affect the clinician's decision to proceed with elective surgery. There was no association between patient age and the probability of diagnosing a significant radiographic abnormality. There was poor agreement between the findings of the clinician and radiologist for cardiovascular and pulmonary abnormalities. There was slight agreement between the findings of the clinician and the radiologist for extrathoracic abnormalities. Our data suggest that the utility of preoperative screening thoracic radiographs in dogs with no specific clinical indications for this test appears to be low. Given the poor agreement between clinician and radiologist findings, clinicians should consider routine evaluation of thoracic radiographs by a radiologist. These findings should be considered by a clinician when deciding whether to recommend thoracic radiographs as a screening tool for a patient without history or physical examination findings suggestive of intrathoracic disease.

Rabies in humans: a treatment approach.

Only few rabies survivors have been described in the medical literature, of whom most suffered severe neurological sequelae. Published treatment protocols have not been applied successfully. Yet, experimental treatments may be of benefit when factors associated with survival are present. Here, we describe two patients who were hospitalised at Amsterdam UMC with clinical rabies and who died despite experimental treatments. We describe the clinical course and medical decisions in the treatment of two rabies patients at our hospital and compared this approach with published data on the treatment of clinical rabies, depending on the presence or absence of prognostic factors associated with survival, and regarded this information in the context of clinical practice. The most important factor associated with survival - the presence of high antibody titres in serum or cerebrospinal fluid (CSF) at the time of diagnosis - was not present in either of the two cases at our hospital. In addition to supportive treatment, both of our patients were treated unsuccesfully with a novel treatment approach with intrathecal and intravenous monoclonal rabies antibodies, which barely increased serum and CSF antibody levels. Higher-dosed treatments with monoclonal antibodies in serum may be needed to yield an effect. Any experimental treatment may be most promising in patients who have other factors associated with survival. In the absence of these, initiation of palliative care still seems to remain the most rational strategy.