Novel tech throws knock-out punch to ECG improving GP referral decisions to cardiology.

PURPOSE: In a comparator study, designed with assistance from the Food and Drug Administration, a State-of-the-Art (SOTA) ECG device augmented with automated analysis, the comparator, was compared with a breakthrough technology, Cardio-HART (CHART). METHODS: The referral decision defined by physician reading biosignal-based ECG or CHART report were compared for 550 patients, where its performance is calculated against the ground truth referral decision. The ground truth was established by cardiologist consensus based on all the available measurements and findings including echocardiography (ECHO). RESULTS: The results confirmed that CHART analysis was far more effective than ECG only analysis: CHART reduced false negative rates 15.8% and false positive (FP) rates by 5%, when compared with SOTA ECG devices. General physicians (GP's) using CHART saw their positive diagnosis rate significantly increased, from ~10% to ~26% (260% increase), and the uncertainty rate significantly decreased, from ~31% to ~1.9% (94% decrease). For cardiology, the study showed that in 98% of the cases, the CHART report was found to be a good indicator as to what kind of heart problems can be expected (the 'start-point') in the ECHO examination. CONCLUSIONS: The study revealed that GP use of CHART resulted in more accurate referrals for cardiology, resulting in fewer true negative or FP-healthy or mildly abnormal patients not in need of ECHO confirmation. The indirect benefit is the reduction in wait-times and in unnecessary and costly testing in secondary care. Moreover, when used as a start-point, CHART can shorten the echocardiograph examination time.

Rapport building and witness memory: Actions may 'speak' louder than words.

Building rapport during police interviews is argued as important for improving on the completeness and accuracy of information provided by witnesses and victims. However, little experimental research has clearly operationalised rapport and investigated the impact of rapport behaviours on episodic memory. Eighty adults watched a video of a mock crime event and 24-hours later were randomly allocated to an interview condition where verbal and/or behavioural (non-verbal) rapport techniques were manipulated. Memorial performance measures revealed significantly more correct information, without a concomitant increase in errors, was elicited when behavioural rapport was present, a superiority effect found in both the free and probed recall phase of interviews. The presence of verbal rapport was found to reduce recall accuracy in the free recall phase of interviews. Post-interview feedback revealed significant multivariate effects for the presence of behavioural (only) rapport and combined (behavioural + verbal) rapport. Participants rated their interview experience far more positively when these types of rapport were present compared to when verbal (only) rapport or no rapport was present. These findings add weight to the importance of rapport in supporting eyewitness cognition, highlighting the potential consequences of impoverished social behaviours for building rapport during dyadic interactions, suggesting 'doing' rather than simply 'saying' may be more beneficial.

A Gastroenterologist' clinical experience in COVID 19 and in-hospital mortality and length of stay analysis.

OBJECTIVES: COVID19 pandemic has forced physicians from different specialties to assist cases overload. Our aim is to assess gastroenterologist's assistance in COVID-19 by assessing mortality, ICU admission, and length of stay, and seek for risk factors for in-hospital mortality and longer hospital stay. METHODS: A total of 41 COVID-19 patients assisted by gastroenterologist (GI cohort) and 137 assisted by pulmonologist, internal medicine practitioners, and infectious disease specialists (COVID expert cohort) during October-November 2020 were prospectively collected. Clinical, demographic, imaging, and laboratory markers were collected and compared between both cohorts. Bivariate analysis and logistic regression were performed to search for risk factors of mortality and longer hospital stays. RESULTS: A total of 27 patients died (15.1%), 11 were admitted to ICU (6.1%). There were no differences between cohorts in mortality (14.6% vs 15.4%;p = 0.90), ICU admission (12.1% vs 4%;p = 0.13), and length of stay (6.67 ± 4 vs 7.15 ± 4.5 days; p = 0.58). PaO2/FiO2 on admission (OR 0.991;CI95% 0.984-0.998) and age > 70 (OR 17.54;CI95% 3.93-78.22) were independently related to mortality. Age > 70, history of malignancy, diabetes, and cardiovascular disease were related to longer hospital stays (p < 0.001, p = 0.03, p = 0.04, p = 0.02 respectively). CONCLUSIONS: COVID-19 assistance was similar between gastroenterologist and COVID experts when assessing mortality, ICU admission, and length of stay. Age>70 and decreased PaO2/FiO2 on admission were independent risk factors of mortality. Age and several comorbidities were related to longer hospital stay.

Radiologists versus Deep Convolutional Neural Networks: A Comparative Study for Diagnosing COVID-19.

The reverse transcriptase polymerase chain reaction (RT-PCR) is still the routinely used test for the diagnosis of SARS-CoV-2 (COVID-19). However, according to several reports, RT-PCR showed a low sensitivity and multiple tests may be required to rule out false negative results. Recently, chest computed tomography (CT) has been an efficient tool to diagnose COVID-19 as it is directly affecting the lungs. In this paper, we investigate the application of pre-trained models in diagnosing patients who are positive for COVID-19 and differentiating it from normal patients, who tested negative for coronavirus. The study aims to compare the generalization capabilities of deep learning models with two thoracic radiologists in diagnosing COVID-19 chest CT images. A dataset of 3000 images was obtained from the Near East Hospital, Cyprus, and used to train and to test the three employed pre-trained models. In a test set of 250 images used to evaluate the deep neural networks and the radiologists, it was found that deep networks (ResNet-18, ResNet-50, and DenseNet-201) can outperform the radiologists in terms of higher accuracy (97.8%), sensitivity (98.1%), specificity (97.3%), precision (98.4%), and F1-score (198.25%), in classifying COVID-19 images.

Sustained type 1 diabetes self-management: Specifying the behaviours involved and their influences.

AIMS: Sustained engagement in type 1 diabetes self-management behaviours is a critical element in achieving improvements in glycated haemoglobin (HbA1c) and minimising risk of complications. Evaluations of self-management programmes, such as Dose Adjustment for Normal Eating (DAFNE), typically find that initial improvements are rarely sustained beyond 12 months. This study identified behaviours involved in sustained type 1 diabetes self-management, their influences and relationships to each other. METHODS: A mixed-methods study was conducted following the first two steps of the Behaviour Change Wheel framework. First, an expert stakeholder consultation identified behaviours involved in self-management of type 1 diabetes. Second, three evidence sources (systematic review, healthcare provider-generated 'red flags' and participant-generated 'frequently asked questions') were analysed to identify and synthesise modifiable barriers and enablers to sustained self-management. These were characterised according to the Capability-Opportunity-Motivation-Behaviour (COM-B) model. RESULTS: 150 distinct behaviours were identified and organised into three self-regulatory behavioural cycles, reflecting different temporal and situational aspects of diabetes self-management: Routine (e.g. checking blood glucose), Reactive (e.g. treating hypoglycaemia) and Reflective (e.g. reviewing blood glucose data to identify patterns). Thirty-four barriers and five enablers were identified: 10 relating to Capability, 20 to Opportunity and nine to Motivation. CONCLUSIONS: Multiple behaviours within three self-management cycles are involved in sustained type 1 diabetes self-management. There are a wide range of barriers and enablers that should be addressed to support self-management behaviours and improve clinical outcomes. The present study provides an evidence base for refining and developing type 1 diabetes self-management programmes.

How fixed versus variable gonadotropin dose during controlled ovarian stimulation could influence the management of infertility patients undergoing IVF treatment: a national Delphi consensus.

AIM: Define how and when fixed starting gonadotropin doses can be used in current clinical ART practices in Italy. METHODS: A Delphi conference consisting of three rounds was performed in order to define the ideal clinical conditions in which fixed-gonadotropin-dose during COS should be applied. During the conference, 19 statements about the current ART practice were provided to a panel of twenty-nine national experts. Median score was 5 (IQ:4-6) in all Delphi rounds. RESULTS: Eleven statements (57.9%) were classified as shareable with high-degree of convergence, 2 (10.5%) as shareable with low convergence and 6 (31.6%) as un-shareable with high convergence. The panel reached high consensus regarding some statements: (i) fixed FSH-dose in normoresponders and poor-responder, (ii) importance of predicting ovarian response before COS, considering multiple markers to select the right stimulation protocol for each patient, (iii) importance of therapy simplification and standardization to improve efficiency during COS. Moreover, a low-convergence was reached about use of GnRH antagonist as first treatment line and drug storage at room temperature. However from these findings, the debate remains open regarding some other statements: (a) usefulness of Bologna-criteria to define poor-responders; (b) efficacy to change always stimulation protocol after a failure IVF; (c) utility of AMH-dosed with standardized automatic mode to define normo-responder patients; (d) usefulness to modify the dosage of 12.5 IU/die during COS to improve stimulation effectiveness. CONCLUSION: Controlled ovarian stimulation remains a challenging clinical step in Assisted Reproductive Technique, especially in some specific patient groups for which no clinical consensus is available. This study is the first attempt to describe the shared clinical opinion regarding the fixed versus variable gonadotropin dose in the real IVF practice.

East Asia expert opinion on treatment initiation for chronic hepatitis B.

BACKGROUND: Globally, chronic hepatitis B (CHB) is a major public health concern. Timely and effective management can prevent disease progression to cirrhosis and reduce the risk of hepatocellular carcinoma (HCC). Currently, there is no consensus on the clinical management of CHB in East Asia. AIM: To establish an East Asia expert opinion on treatment initiation for CHB based on alanine aminotransferase (ALT) level, hepatitis B virus (HBV) deoxyribonucleic acid (DNA) level, cirrhosis and HCC risk scores. METHODS: A meeting was held online with a panel of 10 experts from East Asia to discuss ALT, HBV DNA, cirrhosis and HCC risk scores. Indications for CHB treatment in the latest international guidelines were reviewed. Consensus was summarised to provide recommendations on the initiation of treatment for CHB. RESULTS: Anti-viral therapy is recommended for CHB patients with (a) HBV DNA ≥ 2000 IU/mL and ALT ≥ 1× upper limit of normal (ULN); (b) HBV DNA ≥ 2000 IU/mL, ALT < 1× ULN and ≥ F2 fibrosis and/or ≥ A2 necroinflammation occurs; (c) cirrhosis and detectable HBV DNA; or (d) HBV DNA ≥ 2000 IU/mL, ALT < 1× ULN and a family history of cirrhosis or HCC, extrahepatic manifestations or age > 40 years. Patients with cirrhosis and/or HCC should be treated regardless of ALT levels if HBV DNA level is detectable. Initiating anti-viral therapy or close monitoring at 3-month intervals is recommended for CHB patients with at least two HCC risk factors. CONCLUSIONS: These expert recommendations will contribute to a new standard of daily clinical practice in East Asia.

Expert Opinion on Benefits of Long-Chain Omega-3 Fatty Acids (DHA and EPA) in Aging and Clinical Nutrition.

Life expectancy is increasing and so is the prevalence of age-related non-communicable diseases (NCDs). Consequently, older people and patients present with multi-morbidities and more complex needs, putting significant pressure on healthcare systems. Effective nutrition interventions could be an important tool to address patient needs, improve clinical outcomes and reduce healthcare costs. Inflammation plays a central role in NCDs, so targeting it is relevant to disease prevention and treatment. The long-chain omega-3 polyunsaturated fatty acids (omega-3 LCPUFAs) docosahexaenoic acid (DHA) and eicosapentaenoic acid (EPA) are known to reduce inflammation and promote its resolution, suggesting a beneficial role in various therapeutic areas. An expert group reviewed the data on omega-3 LCPUFAs in specific patient populations and medical conditions. Evidence for benefits in cognitive health, age- and disease-related decline in muscle mass, cancer treatment, surgical patients and critical illness was identified. Use of DHA and EPA in some conditions is already included in some relevant guidelines. However, it is important to note that data on the effects of omega-3 LCPUFAs are still inconsistent in many areas (e.g., cognitive decline) due to a range of factors that vary amongst the trials performed to date; these factors include dose, timing and duration; baseline omega-3 LCPUFA status; and intake of other nutrients. Well-designed intervention studies are required to optimize the effects of DHA and EPA in specific patient populations and to develop more personalized strategies for their use.

Expert Opinion Regarding Appropriately Timing Dermatological Surgeries in Patients With Recent Myocardial Infarctions.

BACKGROUND: There is limited data on the risk of perioperative myocardial infarctions (MIs) in patients with a recent MI who undergo dermatologic surgeries. OBJECTIVE: Present the recommendations of dermatologic surgeons and cardiologists to determine the safety of dermatologic surgeries after a recent MI. METHODS: An electronic survey was distributed to Mohs surgeons and cardiologists to infer the risk of major adverse cardiac events (MACE) inherent to dermatologic surgery and determine timing of dermatologic surgery in patients with a recent MI. RESULTS: One hundred twenty Mohs surgeons and 30 cardiologists were surveyed. Ninety-seven percent of cardiologists and 87% of Mohs surgeons deemed cutaneous excisions and Mohs micrographic surgery as low-risk procedures with less than one-percent chance of MACE. Seventy-seven percent of cardiologists and 46% of Mohs surgeons stated dermatologic surgery should either not be delayed or be delayed up to 1 month after an MI. Responses between cardiologists and Mohs surgeons did not significantly differ. CONCLUSION: A preponderance of surveyed experts believe that most dermatologic surgeries may be safely performed in patients with a history of an MI within 1 month. The decision to implement urgent dermatologic surgery in patients with a recent MI should account for all clinically significant factors.

[Antibiotic strategy in pleural empyema in children: Consensus by the DELPHI method]. / Stratégie antibiotique dans les pleurésies en pédiatrie : consensus par méthode DELPHI.

INTRODUCTION: The evolution of the microbial epidemiology of pleuropulmonary infections complicating community-acquired pneumonia has resulted in a change in empirical or targeted antibiotic therapy in children in the post Prevenar 13 era. The three main pathogens involved in pleural empyema in children are Streptococcus pneumoniae, Staphylococcus aureus and group A Streptococcus. METHODS: A questionnaire according to the DELPHI method was sent to experts in the field (paediatric pulmonologists and infectious disease specialists) in France with the purpose of reaching a consensus on the conservative antibiotic treatment of pleural empyema in children. Two rounds were completed as part of this DELPHI process. RESULTS: Our work has shown that in the absence of clinical signs of severity, the prescription of an intravenous monotherapy is consensual but there is no agreement on the choice of drug to use. A consensus was also reached on treatment adjustment based on the results of blood cultures, the non-systematic use of a combination therapy, the need for continued oral therapy and the lack of impact of pleural drainage on infection control. On the other hand, after the second round of DELPHI, there was no consensus on the duration of intravenous antibiotic therapy and on the treatment of severe pleural empyema, especially when caused by Staphylococci. CONCLUSIONS: The result of this work highlights the needed for new French recommendations based on the evolution of microbial epidemiology in the post PCV13 era.

Current approaches to the integration of sex- and gender-specific medicine in teaching: a qualitative expert survey.

Aim: Although criteria and recommendations for the successful integration of sex- and gender-sensitive aspects in medical teaching have already been published, only a few medical faculties in Germany have conducted the systematic integration of sex- and gender-sensitive medicine. The aim of this expert survey, therefore, was to describe the current approaches to the integration of sex- and gender-sensitive medicine in teaching in the sense of Good Practice. Method: Between April and June 2018, guided interviews were conducted with nine experts in the field of sex- and gender-sensitive medicine. Each of the experts had had experience of implementing sex- and gender-sensitive medicine at their universities. The expert interviews were then evaluated by means of quality content analysis, and frequency analyses were carried out. Results: Aspects of sex- and gender-sensitive medicine were integrated both longitudinally and selectively into the compulsory curriculum or elective fields of various medical, health and nursing science courses. In the opinion of the experts, medical studies should promote the students' gender sensitivity and in particular impart knowledge about the psychosocial and biological aspects of sex- and gender-related differences and sex- and gender-sensitive communication. For the methodological implementation of the integrated contents, didactic resources were partly adapted or developed. The players in the implementation process were confronted with various challenges, e.g. the involvement of the lecturers, the perception of sex- and gender-sensitive medicine as a women's theme as well as ensuring the sustainable integration of sex- and gender-sensitive medicine, which is also structurally anchored in the faculty. Aspects of the curricular integration (e.g. evidence-basing, relevance in examinations) and the structural anchoring (e.g. central organization, staff availability) were mentioned i.a. as being crucial for success. A combination of top-down and bottom-up processes, e.g. by involving the faculty management but also by supporting student initiatives, was described as conducive to success. Conclusion: The depicted approaches to the integration of sex- and gender-sensitive teaching contents give insight as to how sex- and gender-sensitive medicine can be integrated into the curricula. The interviews with the experts point to current themes related to sex- and gender-sensitive medicine and didactic resources. Moreover, it becomes clear which challenges are to be expected for the integration of sex- and gender-sensitive medicine in teaching and how these can be addressed. Particularly the involvement of the faculty's lecturers but also the sustainable integration and continual quality assurance of sex- and gender-sensitive contents present challenges of a crucial nature.

Expert Consensus on clinical application of FDG PET/CT in infection and inflammation.

To further promote the clinical application of 18F-fluorodeoxyglucose positron emission tomography/computed tomography (FDG PET/CT) in infection and inflammation and standardize the diagnostic process, the experts in relevant fields in China carried out discussion and formed the Expert Consensus on the clinical application of FDG PET/CT in infection and inflammation. This consensus is intended to provide a reference for imaging physicians to select a reasonable diagnostic plan. However, it should be noted that it couldn't include or solve all the problems in clinical operation. Imaging physicians and technicians should develop a comprehensive and reasonable diagnostic procedure according to their professional knowledge, clinical experience and currently available medical resources when facing specific patients.

Incorporation of expert knowledge in the statistical detection of diagnosis related group misclassification.

BACKGROUND: In activity based funding systems, the misclassification of inpatient episode Diagnostic Related Groups (DRGs) can have significant impacts on the revenue of health care providers. Weakly informative Bayesian models can be used to estimate an episode's probability of DRG misclassification. METHODS: This study proposes a new, Hybrid prior approach which utilises guesses that are elicited from a clinical coding auditor, switching to non-informative priors where this information is inadequate. This model's ability to detect DRG revision is compared to benchmark weakly informative Bayesian models and maximum likelihood estimates. RESULTS: Based on repeated 5-fold cross-validation, classification performance was greatest for the Hybrid prior model, which achieved best classification accuracy in 14 out of 20 trials, significantly outperforming benchmark models. CONCLUSIONS: The incorporation of elicited expert guesses via a Hybrid prior produced a significant improvement in DRG error detection; hence, it has the ability to enhance the efficiency of clinical coding audits when put into practice at a health care provider.

The experts' perspective of "ask-an-expert": An interview-based study of online nutrition and vaccination outreach.

Social media allow experts to form communities and engage in direct dialogue with publics, which can promote mutual understanding between sciences and publics. However, little is known about experts' participation in online communities, or effective ways to prepare them for public engagement. Here, we explored these issues with experts who voluntarily engage with publics on social media, to understand their public engagement practices. Stimulated recall interviews were conducted with 20 experts who participate in question-and-answer Facebook groups dedicated to vaccines and nutrition. The findings suggest that experts employ diverse considerations in their outreach, partly to establish epistemic trustworthiness. These can be grouped into three goals and two constraints: countering misinformation, establishing benevolence, and establishing competence while maintaining integrity and clarity. Empathic failure and burnout both emerged as factors that impair establishing benevolence. We discuss implications for community-level science literacy and for preparing scientists for "bounded engagement with publics."

Integration of elicited expert information via a power prior in Bayesian variable selection: Application to colon cancer data.

BACKGROUND: Building tools to support personalized medicine needs to model medical decision-making. For this purpose, both expert and real world data provide a rich source of information. Currently, machine learning techniques are developing to select relevant variables for decision-making. Rather than using data-driven analysis alone, eliciting prior information from physicians related to their medical decision-making processes can be useful in variable selection. Our framework is electronic health records data on repeated dose adjustment of Irinotecan for the treatment of metastatic colorectal cancer. We propose a method that incorporates elicited expert weights associated with variables involved in dose reduction decisions into the Stochastic Search Variable Selection (SSVS), a Bayesian variable selection method, by using a power prior. METHODS: Clinician experts were first asked to provide numerical clinical relevance weights to express their beliefs about the importance of each variable in their medical decision making. Then, we modeled the link between repeated dose reduction, patient characteristics, and toxicities by assuming a logistic mixed-effects model. Simulated data were generated based on the elicited weights and combined with the observed dose reduction data via a power prior. We compared the Bayesian power prior-based SSVS performance to the usual SSVS in our case study, including a sensitivity analysis using the power prior parameter. RESULTS: The selected variables differ when using only expert knowledge, only the usual SSVS, or combining both. Our method enables one to select rare variables that may be missed using only the observed data and to discard variables that appear to be relevant based on the data but not relevant from the expert perspective. CONCLUSION: We introduce an innovative Bayesian variable selection method that adaptively combines elicited expert information and real world data. The method selects a set of variables relevant to model medical decision process.

Recommended medical and non-medical factors to assess military preventable deaths: subject matter experts provide valuable insights.

INTRODUCTION: Historically, there has been variability in the methods for determining preventable death within the US Department of Defense. Differences in methodologies partially explain variable preventable death rates ranging from 3% to 51%. The lack of standard review process likely misses opportunities for improvement in combat casualty care. This project identified recommended medical and non-medical factors necessary to (1) establish a comprehensive preventable death review process and (2) identify opportunities for improvement throughout the entire continuum of care. METHODS: This qualitative study used a modified rapid assessment process that includes the following steps: (1) identification and recruitment of US government subject matter experts (SMEs); (2) multiple cycles of data collection via key informant interviews and focus groups; (3) consolidation of information collected in these interviews; and (4) iterative analysis of data collected from interviews into common themes. Common themes identified from SME feedback were grouped into the following subject areas: (1) prehospital, (2) in-hospital and (3) forensic pathology. RESULTS: Medical recommendations for military preventable death reviews included the development, training, documentation, collection, analysis and reporting of the implementation of the Tactical Combat Casualty Care Guidelines, Joint Trauma System Clinical Practice Guidelines and National Association of Medical Examiners autopsy standards. Non-medical recommendations included training, improved documentation, data collection and analysis of non-medical factors needed to understand how these factors impact optimal medical care. CONCLUSIONS: In the operational environment, medical care must be considered in the context of non-medical factors. For a comprehensive preventable death review process to be sustainable in the military health system, the process must be based on an appropriate conceptual framework implemented consistently across all military services.

The Complexity of Diabetic Foot Management: From Common Care to Best Practice. The Italian Expert Opinion by Delphi Survey.

Diabetic foot is a major public health issue, driven by diabetes complications (neuropathy, peripheral vascular disease, foot deformity, and abnormal leucocyte function), which may lead to leg amputation, thus resulting in severe disability, reduced quality of life, and high health costs. The prevention of diabetes complications and the early detection and proper management of diabetic foot wounds are the milestones to avoid major amputations. Unfortunately, in many areas, the prevention of diabetic foot lesions is inadequate and wounds may proceed toward infection and chronicity, with limb- and life-threatening results. Using the Delphi method, we conducted a survey on diabetic foot among Italian experts, selected across different Italian clinical centers. This method was used to achieve a consensus on current opinion and clinical leanings on the diagnosis and management of diabetic foot ulcers. Specifically, the aim of the survey was to evaluate the current management of the diabetic foot syndrome; highlight the differences in the approach among a group of experts; evaluate the role of wound bed preparation and antisepsis; and discuss any areas of disagreement in which evidences are sparse and the clinical judgment plays a crucial role in the decision-making process.

Effect of Different Probes and Expertise on the Interpretation Reliability of Point-of-Care Lung Ultrasound.

BACKGROUND: The effect of different probes and operator experience on the reliability of lung ultrasound (LU) interpretation has not been investigated. We studied the effect of probes and operator experience on the interpretation reliability of LU in critically ill neonates. METHODS: This was a prospective, blind, cohort study enrolling patients with basic patterns ("B," "severe B," consolidation). Patients were scanned with microlinear (15 MHz; L15), phased-array sectorial (6-12 MHz; S7), and microconvex (8 MHz; C8) probes, in random order. Static images were acquired in high resolution, anonymized, and included in a pictorial database in random sequences. Seventeen clinicians with different LU experience were asked to blindly assess the pictorial database. Interrater agreement and interpretation reliability were analyzed. Subanalyses according to expertise and probe, and multivariate linear regression (including an "expertise × probe" interaction factor), were also performed. RESULTS: The agreement tends to be lower and more heterogeneous for residents (intraclass correlation coefficient [ICC], 0.82 [95% CI, 0.74-0.9], P < .001; I2, 67%, P = .04) and for fellows (ICC, 0.93 [95% CI, 0.9-0.97], P < .001; I2, 69%, P = .04), especially when using nonlinear probes, compared with senior physicians (ICC, 0.95 [95% CI, 0.93-0.96], P < .001; I2, 0%, P = .433). Area under the curve (AUC) values were high for all probes (L15, 0.96 [95% CI, 0.93-0.99]; C8, 0.91 [95% CI, 0.85-0.98]; S7, 0.86 [95% CI, 0.82-0.91]) and physicians (senior physicians, 0.95 [95% CI, 0.83-0.99]; fellows, 0.95 [95% CI, 0.75-0.99]; residents, 0.86 [95% CI, 0.5-0.99]). Worse reliability and higher heterogeneity were found when the evaluation was performed by residents (AUC, 0.9 [95% CI, 0.85-0.94], P < .01; I2, 93.6%, P < .001) than by fellows (AUC, 0.99 [95% CI, 0.9-0.999], P < .001; I2, 34.3%, P = .09) and/or by senior physicians (AUC, 0.99 [95% CI, 0.9-0.999], P < .001; I2, 18%, P = .236). The "expertise × probe" interaction factor was associated with lower ICC (standardized regression coefficient ß, -0.69; P < .0001; adjusted R2, 0.99) and AUC (standardized regression coefficient ß, -0.76; P < .0001; adjusted R2, 0.98). CONCLUSIONS: LU interpretation in neonates shows good interrater agreement and reliability, irrespective of the probe and rater expertise. The use of nonlinear probes by novice operators is associated with the lowest agreement and reliability.

A Bayesian non-inferiority approach using experts' margin elicitation - application to the monitoring of safety events.

BACKGROUND: When conducing Phase-III trial, regulatory agencies and investigators might want to get reliable information about rare but serious safety outcomes during the trial. Bayesian non-inferiority approaches have been developed, but commonly utilize historical placebo-controlled data to define the margin, depend on a single final analysis, and no recommendation is provided to define the prespecified decision threshold. In this study, we propose a non-inferiority Bayesian approach for sequential monitoring of rare dichotomous safety events incorporating experts' opinions on margins. METHODS: A Bayesian decision criterion was constructed to monitor four safety events during a non-inferiority trial conducted on pregnant women at risk for premature delivery. Based on experts' elicitation, margins were built using mixtures of beta distributions that preserve experts' variability. Non-informative and informative prior distributions and several decision thresholds were evaluated through an extensive sensitivity analysis. The parameters were selected in order to maintain two rates of misclassifications under prespecified rates, that is, trials that wrongly concluded an unacceptable excess in the experimental arm, or otherwise. RESULTS: The opinions of 44 experts were elicited about each event non-inferiority margins and its relative severity. In the illustrative trial, the maximal misclassification rates were adapted to events' severity. Using those maximal rates, several priors gave good results and one of them was retained for all events. Each event was associated with a specific decision threshold choice, allowing for the consideration of some differences in their prevalence, margins and severity. Our decision rule has been applied to a simulated dataset. CONCLUSIONS: In settings where evidence is lacking and where some rare but serious safety events have to be monitored during non-inferiority trials, we propose a methodology that avoids an arbitrary margin choice and helps in the decision making at each interim analysis. This decision rule is parametrized to consider the rarity and the relative severity of the events and requires a strong collaboration between physicians and the trial statisticians for the benefit of all. This Bayesian approach could be applied as a complement to the frequentist analysis, so both Data Safety Monitoring Boards and investigators can benefit from such an approach.