Routine Decisions Research Articles

Fractional Flow Reserve Implications for Clinical Decision Making in Coronary Artery Disease.

Fractional flow reserve (FFR) is regarded as the gold standard for assessing the functional significance of coronary artery lesions. However, its utilization in clinical practice remains limited. This study aims to determine whether FFR results can influence treatment decisions for coronary artery disease compared to visual assessments of angiographic images. We conducted a retrospective study involving 63 patients diagnosed with either chronic coronary syndrome (n = 39, 61.9%) or acute coronary syndrome (n = 24, 38.1%) who underwent an FFR assessment. Three experienced interventional cardiologists (>300 PCI procedures/year) reevaluated 105 ambiguous coronary lesions in these patients, blinded to the FFR results. The objective was to assess lesion significance and determine the treatment strategy based on a visual angiographic evaluation. The three operators reached concordant agreement (≥two operators) to perform PCI in 60 (57.1%) of the evaluated lesions based on the angiographic assessment. Of these, nine lesions (15%) were deemed functionally non-significant by FFR (FFR > 0.80). Conversely, they agreed to defer PCI in 45 (42.9%) lesions, but 4 lesions (8.9%) were found to be functionally significant (FFR ≤ 0.80) and required a re-evaluation for PCI. Visual-guided decision making by interventional cardiologists shows variability and does not always align with the functional significance of coronary lesions as determined by FFR. Incorporating FFR into routine decision making could enhance treatment accuracy and patient outcomes.

Factors Guiding Clinical Decision-Making in Genitourinary Oncology.

Clinical decision-making in oncology is a complex process, with the primary goal of identifying the most effective treatment tailored to individual cancer patients. Many factors influence the treatment decision: disease- and patient-specific criteria, the increasingly complex treatment landscape, market authorization and drug availability, financial aspects, and personal treatment expertise. In the domain of genitourinary cancers, particularly prostate cancer, decision-making is challenging. Despite the prevalence of this malignancy, there are few in-depth explorations of these factors within real-world scenarios. Understanding and refining this intricate decision-making process is essential for future successful clinical decisions and the integration of computerized decision support into clinicians' workflows. The objective of this study is to improve the current knowledge base and evidence of the factors that influence treatment decision-making for patients with genitourinary cancers. Assessment of how routine treatment decisions are made for genitourinary cancers was performed by a mixed-methods study, encompassing field observations and focus group discussions. In total, we identified 59 factors that influence clinical decision-making in oncology, specifically for genitourinary and prostate cancer. Of these, 23 criteria can be classified as decision-maker-related criteria encompassing personal, cognitive, and emotional attributes and factors of both, healthcare professionals and patients. Moreover, 20 decision-specific criteria have been identified that refer to clinical and disease-related factors, followed by 16 contextual decision factors that describe the relevant criteria introduced by the specific circumstances and environment in which the treatment decision is made. By presenting an exhaustive set of decision factors and providing specific examples for genitourinary cancers, this observational study establishes a possible framework for a better understanding of decision-making. Moreover, we specify and expand the set of decision factors, while emphasizing the importance of cognitive, emotional, and human factors, as well as the quality and accessibility of decision-relevant information.

Activity Based Cost Calculation Decision Making Tool

This research aims to determine the relevant cost calculations carried out by companies in selling or further processing. Routine decision making generally occurs and is related to regular company operational activities. Relevant costs arise in situations where decision makers must choose between two or more options. Decision making will require information on relevant costs. When choosing between different alternatives, managers should concentrate only on costs and revenues that are among the decision systems.

Singapore's COVID-19 crisis decision-making through centralization, legitimacy, and agility: an empirical analysis

Decision-making during health crises differs from routine decision-making and is constrained by ambiguity about evolving epidemiological situations, urgency of response, lack of evidence, and fear. Recent analyses of governance and decision-making during COVID-19, focusing on leadership qualities, involvement of specific stakeholders, and effective resource management, do not adequately address a persisting gap in understanding the determinants of decision-making during health crises at the national level. We undertook a study to understand the processes and characteristics of decision-making during the COVID-19 pandemic in Singapore. We used a case study approach and collected empirical evidence about public health decision-making, using a combination of key informant interviews and focus group discussions with stakeholders from government, academia and civil society organizations. We argue that administrative centralization and political legitimacy played important roles in agile governance and decision-making during the pandemic in Singapore. We demonstrate the role of the Singapore government's centralization in creating a unified and coherent governance model for emergency response and the People's Action Party's (PAP) legitimacy in facilitating people's trust in the government. Health system resilience and financial reserves further facilitated an agile response, yet community participation and prioritization of vulnerable migrant populations were insufficient in the governance processes. Our analysis contributes to the theory and practice of crisis decision-making by highlighting the role of political and administrative determinants in agile crisis decision-making. This study is funded by the U.S. Centers for Disease Control and Prevention through a Cooperative Research Agreement (NU2HGH2020000037).

The contributions of executive functions to decision-making in sport

ABSTRACT In sports, when the game situations become complex, untypical, and uncertain, making good decisions requires controlled processing, i.e., executive functions to regulate thought and action. The current study aimed at comprehending the effects of three core executive functions on decision-making in sport. In study 1, 86 volleyball players (49 experts and 37 novices) completed a routine blocking decision task as well as three core executive function tasks assessing updating, inhibition and shifting. Results showed that the effects of executive functions on blocking decision accuracy were moderated by sports level, with significant predictivity observed only for experts. In study 2, 32 expert players engaged in a dual-task version of the blocking decision task. In this study, only updating was correlated with decision accuracy. Study 3, involving 26 experts, introduced interference during the decision task. The findings highlighted a closer, albeit not statistically significant, relationship between decision accuracy and inhibition compared to updating. Collectively, these findings demonstrated the nuanced yet crucial roles of executive functions in decision-making in sport, underscoring more efficient cognitive processing among higher-level athletes and the different roles served by each executive function.

Embracing Implementation Science to Enhance Fisheries and Aquatic Management and Conservation

The management and conservation of fisheries and aquatic resources are inherently applied activities. Therefore, when knowledge generated from research and monitoring, or knowledge that is held by practitioners and other actors (e.g., Indigenous elders, fishers), fails to inform those applied decisions, the persistent gap between knowledge and action is reinforced (i.e., the knowledge–action gap). In the healthcare realm, there has been immense growth in implementation science over the past decade or so with a goal of understanding and bridging the gap between knowledge and action and delivering on evidence‐based decision making. Yet, within fisheries and aquatic sciences, the concept of implementation science has not received the same level of attention. We posit, therefore, that there is an urgent need to embrace implementation science to enhance fisheries and aquatic management and conservation. In this paper, we seek to describe what implementation science is and what it has to offer to the fisheries and aquatic science and management communities. For our context, we define implementation science as the scientific study of processes and approaches to promote the systematic uptake of research and monitoring findings and other evidence‐based practices into routine practice and decision making to improve the effectiveness of fisheries management and aquatic conservation. We explore various frameworks for implementation science and consider them in the context of fisheries and aquatic science. Although there are barriers and challenges to putting implementation science into practice (e.g., lack of capacity for such work, lack of time to engage in reflection, lack of funding), there is also much in the way of opportunity and several examples of where such efforts are already underway. We conclude by highlighting the research needs related to implementation science in the fisheries and aquatic science realm that span methodological approaches, albeit a common theme is the need to involve practitioners (and other relevant actors) in the research. By introducing the concept and discipline of implementation science to the fisheries and aquatic science community, our hope is that we will inspire individuals and organizations to learn more about how implementation science can help deliver on the promise of evidence‐based management and decision making and narrow the gap between research and practice.

How transparency affects algorithmic advice utilization: The mediating roles of trusting beliefs

Although algorithms are increasingly used to support professional tasks and routine decision-making, their opaque nature invites resistance and results in suboptimal use of their advice. Scholars argue for transparency to enhance the acceptability of algorithmic advice. However, current research is limited in understanding how improved transparency enhances the use of algorithmic advice, such as the differences among various aspects of transparency and the underlying mechanism. In this paper, we investigate whether and how different aspects of algorithmic transparency (performance, process, and purpose) enhance the use of algorithmic advice. Drawing on the knowledge-based trust perspective, we examine the mediating roles of trusting beliefs in the relationships between transparency and the use of algorithmic advice. Using the “judge-advisor system” paradigm, we conduct a 2 × 2 × 2 experiment to manipulate the three aspects of transparency and examine their effects on the use of algorithmic advice. We find that performance and process transparency promote the use of algorithmic advice. However, the effect of process transparency gets attenuated when purpose transparency is high. Purpose transparency is only useful when process transparency is low. We also find that while all three aspects of transparency facilitate different trusting beliefs, only competence belief significantly promotes the use of algorithmic advice. It also fully mediates the facilitating effects of performance and process transparency. This study contributes to the emerging research on algorithmic decision support by empirically investigating the effects of transparency on the use of algorithmic advice and identifying the underlying mechanism. The findings also provide practical guidance on how to promote the acceptance of algorithmic advice that is valuable to both individual users and practitioners.

Cefiderocol for treating severe aerobic Gram-negative bacterial infections: technology evaluation to inform a novel subscription-style payment model.

To limit the use of antimicrobials without disincentivising the development of novel antimicrobials, there is interest in establishing innovative models that fund antimicrobials based on an evaluation of their value as opposed to the volumes used. The aim of this project was to evaluate the population-level health benefit of cefiderocol in the NHS in England, for the treatment of severe aerobic Gram-negative bacterial infections when used within its licensed indications. The results were used to inform the National Institute for Health and Care Excellence guidance in support of commercial discussions regarding contract value between the manufacturer and NHS England. The health benefit of cefiderocol was first derived for a series of high-value clinical scenarios. These represented uses that were expected to have a significant impact on patients' mortality risks and health-related quality of life. The clinical effectiveness of cefiderocol relative to its comparators was estimated by synthesising evidence on susceptibility of the pathogens of interest to the antimicrobials in a network meta-analysis. Patient-level costs and health outcomes of cefiderocol under various usage scenarios compared with alternative management strategies were quantified using decision modelling. Results were reported as incremental net health effects expressed in quality-adjusted life-years, which were scaled to 20-year population values using infection number forecasts based on data from Public Health England. The outcomes estimated for the high-value clinical scenarios were extrapolated to other expected uses for cefiderocol. Among Enterobacterales isolates with the metallo-beta-lactamase resistance mechanism, the base-case network meta-analysis found that cefiderocol was associated with a lower susceptibility relative to colistin (odds ratio 0.32, 95% credible intervals 0.04 to 2.47), but the result was not statistically significant. The other treatments were also associated with lower susceptibility than colistin, but the results were not statistically significant. In the metallo-beta-lactamase Pseudomonas aeruginosa base-case network meta-analysis, cefiderocol was associated with a lower susceptibility relative to colistin (odds ratio 0.44, 95% credible intervals 0.03 to 3.94), but the result was not statistically significant. The other treatments were associated with no susceptibility. In the base case, patient-level benefit of cefiderocol was between 0.02 and 0.15 quality-adjusted life-years, depending on the site of infection, the pathogen and the usage scenario. There was a high degree of uncertainty surrounding the benefits of cefiderocol across all subgroups. There was substantial uncertainty in the number of infections that are suitable for treatment with cefiderocol, so population-level results are presented for a range of scenarios for the current infection numbers, the expected increases in infections over time and rates of emergence of resistance. The population-level benefits varied substantially across the base-case scenarios, from 896 to 3559 quality-adjusted life-years over 20 years. This work has provided quantitative estimates of the value of cefiderocol within its areas of expected usage within the NHS. Given existing evidence, the estimates of the value of cefiderocol are highly uncertain. Future evaluations of antimicrobials would benefit from improvements to NHS data linkages; research to support appropriate synthesis of susceptibility studies; and application of routine data and decision modelling to assess enablement value. No registration of this study was undertaken. This award was funded by the National Institute for Health and Care Research (NIHR) Health Technology Assessment Policy Research Programme (NIHR award ref: NIHR135591), conducted through the Policy Research Unit in Economic Methods of Evaluation in Health and Social Care Interventions, PR-PRU-1217-20401, and is published in full in Health Technology Assessment; Vol. 28, No. 28. See the NIHR Funding and Awards website for further award information.

A novel nonparametric time-dependent precision-recall curve estimator for right-censored survival data.

In order to assess prognostic risk for individuals in precision health research, risk prediction models are increasingly used, in which statistical models are used to estimate the risk of future outcomes based on clinical and nonclinical characteristics. The predictive accuracy of a risk score must be assessed before it can be used in routine clinical decision making, where the receiver operator characteristic curves, precision-recall curves, and their corresponding area under the curves are commonly used metrics to evaluate the discriminatory ability of a continuous risk score. Among these the precision-recall curves have been shown to be more informative when dealing with unbalanced biomarker distribution between classes, which is common in rare event, even though except one, all existing methods are proposed for classic uncensored data. This paper is therefore to propose a novel nonparametric estimation approach for the time-dependent precision-recall curve and its associated area under the curve for right-censored data. A simulation is conducted to show the better finite sample property of the proposed estimator over the existing method and a real-world data from primary biliary cirrhosis trial is used to demonstrate the practical applicability of the proposedestimator.

Advanced Systems of Planning and Scheduling and the Limitations of their Implementation: A Study Case of Al kufa Old Cement Factory

The present study is based on research and investigation into planning and scheduling systems as a basic variable and the limitations of applying planning and scheduling systems as a dependent variable to investigate the most important obstacles that prevent their application due to the development of the technology in planning and scheduling systems in addition to the advances in the algorithms of optimization by using a computer to make most of the routine decisions and using interactivity to send the information and the real understanding of the efficiency of the planners. And by changing the competition between the manufacturing companies around the world, as well as the relationships between these companies, suppliers, and customers. Supply chain management has greatly reduced the planning problem, and the concepts of flexible production have contributed to changing the relationship between suppliers and consumers at every link in the supply chain. This was done through a questionnaire that was distributed to the company’s management to investigate the extent to which this factory benefits from these modern systems, the obstacles that curb applying these systems in this factory, and the extent to which this industrial factory seeks to obtain modern systems, to cope with development, and be able to improve their performance, which will contribute to increasing their efficiency and ability to compete in the labor market.

Development and validation of a blood routine-based extent and severity clinical decision support tool for ulcerative colitis

Monitoring extent and severity is vital in the ulcerative colitis (UC) follow-up, however, current assessment is complex and low cost-effectiveness. We aimed to develop a routine blood-based clinical decision support tool, Jin’s model, to investigate the extent and severity of UC. The multicentre retrospective cohort study recruited 975 adult UC inpatients and sub-grouped into training, internal validation and external validation set. Model was developed by logistics regression for the extent via Montreal classification and for the severity via Mayo score, Truelove and Witts score (TWS), Mayo endoscopic score (MES) and Degree of Ulcerative colitis Burden of Luminal Inflammation (DUBLIN) score. In Montreal classification, left-sided and extensive versus proctitis model achieved area under the receiver operating characteristic curve (AUROC) of 0.78 and 0.81 retrospectively. For severity, Mayo score model, TWS model, MES model and DUBLIN score model achieved an AUROC of 0.81, 0.70, 0.74 and 0.70 retrospectively. The models also were evaluated with satisfactory calibration and clinical unity. Jin’s model was free with open access at http://jinmodel.com:3000/. Jin’s model is a noninvasive, convenient, and efficient approach to assess the extent and severity of UC.

Can precision medicine be integrated into routine therapeutic decisions at the bedside of patients?*

Therapeutic strategies are shifting from a "one-size-fits-all" population-based approach to a stratified approach targeting groups with similar characteristics, or even individuals, tailoring treatments to the unique characteristics of each patient. Since such strategies rely on increasingly complex knowledge and healthcare technologies, along with an understanding of the tools of precision medicine, the appropriate dissemination and use of these strategies involves a number of challenges for the medical community. Having evaluation methodologies that have been jointly designed with the institutional, industrial, academic stakeholders, and also patients, like streamlining the processes and externally validating performances, could enhance the relevance of the "evaluation" aspect of precision medicine. Creating a network of expert precision-medicine centers and ensuring that precision-medicine procedures are reimbursed by social security would guarantee fair and sustainable access. Finally, training healthcare professionals, creating interfaces between precision-medicine expert centers and primary care professionals as well as patients, and integrating individual patient data into medical records are all key drivers that will enable information from precision-medicine to be made available and guarantee the proper use of these approaches.

A Study on Emotional Intelligence at Workplace

Abstract Your professional life and career may be significantly affected by emotional intelligence, so it's critical to comprehend what it is and why it matters. People in every workplace have varying personalities, strengths, and emotions, which can have a big impact on how they work. The capacity to recognize and control both your own and other people's emotions is known as emotional intelligence. Employers usually take emotional intelligence into account when making routine decisions about hiring, firing, and promotion. In order to determine which candidates have a higher level of emotional intelligence, hiring managers frequently research candidates' emotional intelligence by posing targeted questions during the hiring process. In order to assess leadership potential, they also evaluate the emotional intelligence of their current workforce.

Cardiac protective techniques in left breast radiotherapy: rapid selection criteria for routine clinical decision making

ObjectiveIn left breast radiotherapy (RT) desired heart doses may be achieved without heart-sparing RT techniques in some patients. We aimed to examine the existence of predictive factors and cutoff points to determine which patients are the main candidates for heart-sparing RT techniques.Material and methodDosimetric data for left breast cancer was examined. RT plans were made at conventional doses to the breast and peripheral lymph nodes. Statistical analyses were performed using SPSS 22.0 (SPSS Inc., IBM Corp., Armonk, NY).Result114 cases were evaluated by ROC (Receiver operating characteristic) analysis in the breast-conserving surgery (BCS) and mastectomy groups. While only left lung volume (AUC: 0.74, 95% CI 0.61–0.87, p = 0.002) was significant in BCS cases, in cases with mastectomy, left lung volume (AUC: 0.81, 95% CI 0.69–0.94, p = 0.002) and lung/heart volume ratio (AUC: 0.83, 95% CI 0.70–0.96, p = 0.001) had a significant relationship with the relevance of heart doses. The cutoff point of 1.92 was selected for the lung/heart volume ratio for the mastectomized patients. Moreover, the cutoff point 1154 cc and 1208 cc was determined for the left lung volume for the BCS and mastectomized patients, respectively.ConclusionVarious cutoff points in left breast RT can be used to predict whether RT plans will meet QUANTEC (Quantitative Analysis of Normal Tissue Effects in the Clinic) heart dose limits. Evaluating only these few cutoff points before planning makes it possible to eliminate 70% of patients with BCS and 40% of patients with mastectomy from respiratory-controlled methods, which require time and effort. Patients with lung volume and lung/heart volume ratio smaller than the cutoff values can be considered primary candidates for heart-sparing techniques.

Nano-topographical surface engineering for enhancing bioactivity of PEEK implants (in vitro—histomorphometric study)

ObjectivesDental implants are currently becoming a routine treatment decision in dentistry. Synthetic polyetheretherketone (PEEK) polymer is a prevalent component of dental implantology field. The current study aimed to assess the influence of Nd:YAG laser nano-topographical surface engineering combined with ultraviolet light or platelet rich fibrin on the bioactivity and osseointegration of PEEK implants in laboratory and animal testing model.Materials and methodsComputer Aided Design-Computer Aided Manufacturing (CAD CAM) discs of PEEK were used to fabricate PEEK discs (8 mm × 3 mm) N = 36 and implant cylinders (3 mm × 6 mm) N = 72. Specimens were exposed to Nd:YAG laser at wavelength 1064 nm, and surface roughness topography/Ra parameter was recorded in nanometer using atomic force microscopy. Laser modified specimens were divided into three groups: Nd:YAG laser engineered surfaces (control), Nd:YAG laser/UV engineered surfaces and Nd:YAG laser/PRF engineered surfaces (N = 12 discs–N = 24 implants). In vitro bioactivity test was performed, and precipitated apatite minerals were assessed with X-ray diffraction analysis (XRD) and scanning electron microscopy (SEM). In vivo histomorphometric analysis was performed in rabbits with BIC% calculation.ResultsRa mean value of PEEK laser engineered surfaces was 125.179 nm. For the studied groups, XRD patterns revealed distinctive peaks of different apatite minerals that were demonstrated by SEM as dispersed surface aggregations. There was a significant increase in the BIC% from control group 56.43 (0.97) to laser/UV surfaces 77.30 (0.78) to laser/PRF 84.80 (1.29) (< 0.0001).ConclusionsSuccessful engineered nano-topographical biomimetic PEEK implant could be achieved by Nd:YAG laser technique associated with improving bioactivity. The combination with UV or PRF could be simple and economic methods to gain more significant improvement of PEEK implant surface bioactivity with superior osteointegration.

Using design thinking to explore teaching problems in Chilean schools

ABSTRACT Educational systems face increasingly complex demands, confronting teachers with multidimensional people-centred problems rarely solved by linear or standardised solutions. Nevertheless, teachers must juggle multiple variables simultaneously in their daily work. This can lead to routine and unreflective decisions that do not consider unique situations. Considering that designers’ abductive reasoning could support problem-framing skills, this article discusses how a design thinking approach can contribute to developing reflective teaching practice. This case study explores how 20 Chilean teachers define, frame, and re-frame their pedagogical problems in a design-based teacher professional development programme. Findings revealed three problem-framing triggers that support teachers’ reflection: (a) collaborative discussions, (b) awareness of people and their context, and (c) visualising, making, and testing ideas. Combined, they articulate action and promote reflection, demonstrating the value of a design thinking approach in supporting teachers’ pedagogical decisions.

AUGMENTED INTELLIGENCE: HUMAN-AI COLLABORATION IN THE ERA OF DIGITAL TRANSFORMATION

Augmented Intelligence (AI) combines human and artificial intelligence to enhance decision-making. This paper reviews AI concepts, applications, and collaboration models like human-in-the-loop AI and cognitive computing AI. It examines AI's role in improving human judgment, handling large datasets, and making routine decisions. The paper explores AI's impact on sectors like healthcare through use cases in blood glucose monitoring. It applies the McKinsey 4D framework to implement AI for glucose monitoring. The paper also discusses emerging models like Hybrid Augmented Intelligence (HAI) that integrate human cognition within AI systems for optimal performance. Overall, it underscores AI's potential in complementing human capabilities and driving innovation across industries with responsible design.

Gastroscopy for dyspepsia: Understanding primary care and gastroenterologist mental models of practice: A cognitive task analysis approach.

Gastroscopy to investigate dyspepsia without alarm symptoms rarely results in clinically actionable findings or sustained health-related quality-of-life improvements among patients aged 18-60 years and is, therefore, not recommended. Despite this, referrals for and performance of gastroscopy among this patient population remain high. The purpose of this study was to understand family physicians' and gastroenterologists' mental models of dyspepsia and the drivers behind referring or performing gastroscopy. Cognitive task analysis routine critical decision method interviews with family physicians (n = 8) and gastroenterologists (n = 4). Family physicians and gastroenterologists hold rich mental models of dyspepsia that rely on sensemaking; however, gaps in information continuity affect their ability to plan and coordinate patient care. Drivers behind decisions to refer or perform gastroscopy were: eliminating risk for serious pathology, providing reassurance, perceived preference by patients to receive information and reassurance from gastroenterologists, maintaining relationships with patients, and saving costs to the health system. Family physicians refer for dyspepsia when they are seeking support from gastroenterologists, they believe that alternative factors may be impacting the patient's health or view it as a cost-saving measure. Likewise, gastroenterologists perform gastroscopy for dyspepsia when they perceive it as a cost-saving measure, they want to support their primary care colleagues and provide their colleagues and patients with reassurance. An improved degree of communication between speciality and primary care could allow for continuity in the transfer of information about patients and reduce referrals for dyspepsia.

The alcohol biomarker phosphatidylethanol (PEth) – test performance and experiences from routine analysis and external quality assessment

Phosphatidylethanol (PEth) are membrane molecules formed from phosphatidylcholine and ethanol through transphosphatidylation catalyzed by phospholipase D. Measurement of the main PEth form 16:0/18:1 is used as a specific and sensitive alcohol biomarker, since its formation requires ethanol, it accumulates in the blood upon repeated ethanol exposure, and it is only slowly eliminated during abstinence. PEth formation correlates with alcohol intake at the population level, albeit with considerable inter-individual variation as for the half-life during withdrawal. Over the past decade, the use of PEth has increased significantly and the applications have broadened. In Sweden, routine decision limits and the interpretation of test results for PEth were harmonized in 2013, using < 0.05 µmol/L (∼35 µg/L) as the recommended lower reporting limit and values > 0.30 µmol/L (∼210 µg/L) to indicate regular high alcohol intake. Routine test results show a large variation with about half being < 0.05 µmol/L and some even exceeding 10 µmol/L. In 2013, an external quality assessment (EQA) scheme for PEth 16:0/18:1 measurement in whole blood was also started (Equalis, Uppsala, Sweden), presently involving 56 laboratories from 13 countries. The agreement of PEth results between the laboratories has gradually improved to a CV < 15%. The current clinical and scientific information suggests that PEth values below the lower reporting limit (typically ∼0.03–0.05 µmol/L, or ∼20–35 µg/L) indicates sobriety or only low or occasional alcohol consumption, while regular high alcohol intake at levels corresponding to harmful drinking is required in most cases to reach PEth values > 0.30 µmol/L.

Testing the performance, adequacy, and applicability of an artificial intelligence model for pediatric pneumonia diagnosis

BackgroundCommunity-acquired Pneumonia (CAP) is a common childhood infectious disease. Deep learning models show promise in X-ray interpretation and diagnosis, but their validation should be extended due to limitations in the current validation workflow. To extend the standard validation workflow we propose doing a pilot test with the next characteristics. First, the assumption of perfect ground truth (100% sensitive and specific) is unrealistic, as high intra and inter-observer variability have been reported. To address this, we propose using Bayesian latent class models (BLCA) to estimate accuracy during the pilot. Additionally, assessing only the performance of a model without considering its applicability and acceptance by physicians is insufficient if we hope to integrate AI systems into day-to-day clinical practice. Therefore, we propose employing explainable artificial intelligence (XAI) methods during the pilot test to involve physicians and evaluate how well a Deep Learning model is accepted and how helpful it is for routine decisions as well as analyze its limitations by assessing the etiology. This study aims to apply the proposed pilot to test a deep Convolutional Neural Network (CNN)-based model for identifying consolidation in pediatric chest-X-ray (CXR) images already validated using the standard workflow. MethodsFor the standard validation workflow, a total of 5856 public CXRs and 950 private CXRs were used to train and validate the performance of the CNN model. The performance of the model was estimated assuming a perfect ground truth. For the pilot test proposed in this article, a total of 190 pediatric chest-X-ray (CXRs) images were used to test the CNN model support decision tool (SDT). The performance of the model on the pilot test was estimated using extensions of the two-test Bayesian Latent-Class model (BLCA). The sensitivity, specificity, and accuracy of the model were also assessed. The clinical characteristics of the patients were compared according to the model performance. The adequacy and applicability of the SDT was tested using XAI techniques. The adequacy of the SDT was assessed by asking two senior physicians the agreement rate with the SDT. The applicability was tested by asking three medical residents before and after using the SDT and the agreement between experts was calculated using the kappa index. ResultsThe CRXs of the pilot test were labeled by the panel of experts into consolidation (124/176, 70.4%) and no-consolidation/other infiltrates (52/176, 29.5%). A total of 31/176 (17.6%) discrepancies were found between the model and the panel of experts with a kappa index of 0.6. The sensitivity and specificity reached a median of 90.9 (95% Credible Interval (CrI), 81.2–99.9) and 77.7 (95% CrI, 63.3–98.1), respectively. The senior physicians reported a high agreement rate (70%) with the system in identifying logical consolidation patterns. The three medical residents reached a higher agreement using SDT than alone with experts (0.66±0.1 vs. 0.75±0.2). ConclusionsThrough the pilot test, we have successfully verified that the deep learning model was underestimated when a perfect ground truth was considered. Furthermore, by conducting adequacy and applicability tests, we can ensure that the model is able to identify logical patterns within the CXRs and that augmenting clinicians with automated preliminary read assistants could accelerate their workflows and enhance accuracy in identifying consolidation in pediatric CXR images.