Preemptive Intervention Research Articles

Links Between Early Prelinguistic Communication and Later Expressive Language in Toddlers With Autistic and Non-Autistic Siblings.

The present study explored the extent to which early prelinguistic communication skills predict expressive language in toddlers with autistic siblings (Sibs-autism), who are known to be at high likelihood for autism and language disorder, and a comparison group of toddlers with non-autistic older siblings (Sibs-NA). Participants were 51 toddlers (29 Sibs-autism, 22 Sibs-NA) aged 12-18 months at the first time point in the study (Time 1). Toddlers were seen again 9 months later (Time 2). Three prelinguistic communication skills (i.e., intentional communication, vocalization complexity, and responding to joint attention) were measured at Time 1 via the Communication and Symbolic Behavior Scales Developmental Profile-Behavior Sample. An expressive language aggregate was calculated for each participant at Time 2. A series of correlation and multiple regression models was run to evaluate associations of interest between prelinguistic communication skills as measured at Time 1 and expressive language as measured at Time 2. Vocalization complexity and intentional communication displayed significant zero-order correlations with expressive language across sibling groups. Vocal complexity and responding to joint attention did not have significant added value in predicting later expressive language, after covarying for intentional communication across groups. However, sibling group moderated the association between vocalization complexity and later expressive language, such that vocal complexity displayed incremental validity for predicting later expressive language, covarying for intentional communication, only within Sibs-NA. Results indicate that prelinguistic communication skills, in particular intentional communication, show promise for predicting later expressive language in siblings of autistic children. These findings provide additional empirical support for the notion that early preemptive interventions targeting prelinguistic communication skills, especially intentional communication, may have the potential to scaffold language acquisition and support more optimal language outcomes in this population at high likelihood for a future diagnosis of both autism and language disorder. https://doi.org/10.23641/asha.27745437.

‘Social defence’ and the resilience of the domestic war on terror: A genealogy of social security, national security, and defence

Abstract Russia’s invasion of Ukraine has moved attention within International Security back to Great Power Politics and interstate war, and away from phenomena like ‘foreign terrorist fighters’ and lone actor terrorism. Is the War on Terror paradigm receding? This paper argues that counter-radicalisation imaginaries and counter-extremism programming will all remain stubbornly present despite the shift back to interstate war. The resilience of domestic War on Terror policies will stem from their integration within local municipality activity and social policy during the second decade of the 21st century. These policy areas and levels are detached from the international security community, meaning that shifts in attention at the global level will have little effect upon localised national security practices. The paper then interrogates how counterterrorism agendas have so successfully entered the world of social policy and local public administration, using genealogical methods to trace how social security and crime prevention were historically prefaced upon defence. The paper traces the policy paradigm of ‘social defence’ which dominated mid-20th century international organisations, targeting non-criminal juveniles with pre-emptive interventions to prevent ‘dangerous persons’ imperilling society. Effectively, this untold history of crime prevention in Europe demonstrates the profound interlocking of national security, social policy and defence across the 20th century, with implications for the resilience of this triad.

Pre-emptive therapeutic decisions based on measurable residual disease status in acute myeloid leukemia: ready for prime time?

The use of measurable residual disease (MRD) as a biomarker for prognostication, risk stratification, and therapeutic decision-making in acute myeloid leukemia (AML) is gaining prominence. MRD monitoring for NPM1-mutated and core-binding factor AML using PCR techniques is well-established for assessing disease after intensive chemotherapy. AML with persistent FLT3-ITD MRD post-intensive chemotherapy and pre-allogeneic hematopoietic cell transplantation (pre-allo-HCT) is associated with an increased risk of relapse and lower survival. Pre-allo-HCT MRD is an independent risk factor for post-allo-HCT outcomes, including relapse and death. Therefore, preemptive interventions on the natural history of MRD positivity are an active area of research beyond its initial prognostic function. Targeting MRD in AML with innovative treatment strategies can improve patient outcomes.

Multimodal Intravascular Imaging of the Vulnerable Coronary Plaque.

Vulnerable coronary plaques are atherosclerotic lesions which, due to their specific phenotype, are prone to plaque rupture and to cause acute coronary syndromes, with subsequent relevant morbidity and mortality. Strategies to break the chain link between plaque vulnerability and adverse clinical events include optimized pharmacologic prevention and potentially also preemptive percutaneous coronary interventions (previously defined as "plaque sealing" or "plaque passivation"). Various morphologic features of the vulnerable plaques have been described, including aspects regarding the large necrotic lipid content, the thin fibrous cap, the presence and extent of the presence of calcifications with small size and calcification angle, and as well as the large macrophage infiltration within the plaque. The detection of these features of plaque vulnerability is possible with intravascular imaging modalities such as intravascular ultrasound (IVUS), near-infrared spectroscopy (NIRS), and optical coherence tomography (OCT). This review explores the peculiarities of these three imaging modalities for the detection of vulnerable coronary plaque features.

Liquid biopsy with plasma Epstein-Barr virus DNA characterizes biological relapse for the prediction of cancer recurrence in non-disseminated nasopharyngeal carcinoma

PurposeTo investigate whether a bounce in plasma Epstein-Barr virus (EBV) DNA during posttreatment surveillance of nasopharyngeal carcinoma (NPC) informs the risk of clinical recurrence and its implication for early therapeutic intervention. Methods950 non-disseminated NPC patients with completed remission in 3 months after treatment were retrospectively screened. Detectable EBV DNA with no evidence of clinical relapse during follow-up was deemed as DNA bounce. The diagnostic and prognostic performance of EBV DNA bounce was assessed for subsequent failures. ResultsTumor recurrence occurred in 6.6 %, 10.1 % and 65.8 % in the group with persistently negative EBV DNA, single positive test and ≥ 2 positive tests, respectively. EBV DNA bounce over twice was associated with worse disease-free survival (DFS), locoregional recurrence-free survival (LRRFS), and distant metastasis-free survival (DMFS) than the other two groups. Sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV) and accuracy for the prediction of recurrence were 0.56, 0.95, 0.66, 0.93 and 0.90 using two positive tests, which were hence deemed as biological relapse. Serial cutoffs (EBV DNA 1 ≥ 40 copies/ml or EBV DNA 2 ≥100 copies/ml) further defined a high-risk subgroup with an eventual recurrence rate of 77.9 % and 3-year DFS of merely 20.5 %. Prophylactic medical intervention with capecitabine or S1 significantly improved the 3-year DFS when compared to those with observation. ConclusionsThe earliest two positive tests of EBV DNA represent a biomarker of biological relapse that allows early detection of clinical recurrence in EBV-related NPC. For high-risk biological relapse, preemptive intervention provides potential survival benefits.

Chatbots in medicine: certification process and applied use case.

Chatbots are computer programs designed to engage in natural language conversations in an easy and understandable way. Their use has been accelerated recently with the advent of large language models. However, their application in medicine and healthcare has been limited due to concerns over data privacy, the risk of providing medical diagnoses, and ensuring regulatory and legal compliance. Medicine and healthcare could benefit from chatbots if their scope is carefully defined and if they are used appropriately and monitored long-term. The confIAnce chatbot, developed at the Geneva University Hospitals and the University of Geneva, is an informational tool aimed at providing simplified information to the general public about primary care and chronic diseases. In this paper, we describe the certification and regulatory aspects applicable to chatbots in healthcare, particularly in primary care medicine. We use the confIAnce chatbot as a case study to explore the definition and classification of a medical device and its application to chatbots, considering the applicable Swiss regulations and the European Union AI Act. Chatbots can be classified anywhere from non-medical devices (informational tools that do not handle patient data or provide recommendations for treatment or diagnosis) to Class III medical devices (high-risk tools capable of predicting potentially fatal events and enabling a pre-emptive medical intervention). Key considerations in the definition and certification process include defining the chatbot's scope, ensuring compliance with regulations, maintaining security and safety, and continuously evaluating performance, risks, and utility. A lexicon of relevant terms related to artificial intelligence in healthcare, medical devices, and regulatory frameworks is also presented in this paper. Chatbots hold potential for both patients and healthcare professionals, provided that their scope of practice is clearly defined, and that they comply with regulatory requirements. This review aims to provide transparency by outlining the steps required for certification and regulatory compliance, making it valuable for healthcare professionals, scientists, developers, and patients.

Machine learning for adverse event prediction in outpatient parenteral antimicrobial therapy: a scoping review.

This study aimed to conduct a scoping review of machine learning (ML) techniques in outpatient parenteral antimicrobial therapy (OPAT) for predicting adverse outcomes and to evaluate their validation, implementation and potential barriers to adoption. This scoping review included studies applying ML algorithms to adult OPAT patients, covering techniques from logistic regression to neural networks. Outcomes considered were medication intolerance, toxicity, catheter complications, hospital readmission and patient deterioration. A comprehensive search was conducted across databases including Cochrane Central, Cochrane Reviews, Embase, Ovid MEDLINE and Scopus, from 1 January 2000 to 1 January 2024. Thirty-two studies met the inclusion criteria, with the majority being single-centre experiences primarily from North America. Most studies focused on developing new ML models to predict outcomes such as hospital readmissions and medication-related complications. However, there was very little reporting on the performance characteristics of these models, such as specificity, sensitivity and C-statistics. There was a lack of multi-centre or cross-centre validation, limiting generalizability. Few studies advanced beyond traditional logistic regression models, and integration into clinical practice remains limited. ML shows promise for enhancing OPAT outcomes by predicting adverse events and enabling pre-emptive interventions. Despite this potential, significant gaps exist in development, validation and practical implementation. Barriers include the need for representative data sets and broadly applicable, validated models. Future research should address these barriers to fully leverage ML's potential in optimizing OPAT care and patient safety. Models must deliver timely, accurate and actionable insights to improve adverse event prediction and prevention in OPAT settings.

Literacy training vs. psychological inoculation? Explicating and comparing the effects of predominantly informational and predominantly motivational interventions on the processing of health statistics

Abstract Communicating statistics is challenging and fraught with mis-contextualization and causal misattributions. Can we train the public against statistical misrepresentations? Pre-emptive interventions against misinformation primarily include literacy tips/training and inoculation. In theory, inoculation has an additional motivational component (forewarning). However, forewarning has not been directly tested against literacy interventions, calling into question inoculation’s distinction. We critique the theoretical boundary work and compare these informational and motivational interventions in the context of health statistics. The longitudinal experiment compared the effects of interventions on processing accurate and inaccurate statistics about COVID-19 vaccines and/or genetically modified organisms across digital platforms. Both interventions prevented an elevation in risk perceptions following exposure to statistical misinformation at a later time. However, literacy intervention increased risk perceptions following exposure to accurate statistics too, suggesting an additional benefit of forewarning. Those with high levels of pre-existing misinformation concern exhibited inoculation effects more strongly. We discuss the theoretical, empirical, and practical implications.

Dynamic Relapse Prediction by Peripheral Blood WT1mRNA after Allogeneic Hematopoietic Cell Transplantation for Myeloid Neoplasms

Although various relapse prediction models based on pretransplant information have been reported, they cannot update the predictive probability considering post-transplant patient status. Therefore, these models are not appropriate for deciding on treatment adjustment and preemptive intervention during post-transplant follow-up. A dynamic prediction model can update the predictive probability by considering the information obtained during follow-up. This study aimed to develop and assess a dynamic relapse prediction model after allogeneic hematopoietic cell transplantation (allo-HCT) for acute myeloid leukemia (AML) and myelodysplastic syndrome (MDS) using peripheral blood Wilms' tumor 1 messenger RNA (WT1mRNA). We retrospectively analyzed patients with AML or MDS who underwent allo-HCT at our institution. To develop dynamic models, we employed the landmarking supermodel approach, using age, refined disease risk index, conditioning intensity, and number of transplantations as pretransplant covariates and both pre- and post-transplant peripheral blood WT1mRNA levels as time-dependent covariates. Finally, we compared the predictive performances of the conventional and dynamic models by area under the time-dependent receiver operating characteristic curves. A total of 238 allo-HCT cases were included in this study. The dynamic model that considered all pretransplant WT1mRNA levels and their kinetics showed superior predictive performance compared to models that considered only pretransplant covariates or factored in both pretransplant covariates and post-transplant WT1mRNA levels without their kinetics; their time-dependent areas under the curve were 0.89, 0.73, and 0.87, respectively. The predictive probability of relapse increased gradually from approximately 90 days before relapse. Furthermore, we developed a web application to make our model user-friendly. This model facilitates real-time, highly accurate, and personalized relapse prediction at any time point after allo-HCT. This will aid decision-making during post-transplant follow-up by offering objective relapse forecasts for physicians.

A comparative study across mule, equine and equine clone pregnancies regarding the determination of the day of birth

Monitoring equine parturition effectively is essential for preemptive intervention in periparturient issues and ensuring the overall well-being of both mares and foals. However, its implementation in breeding farms is challenging due to variable gestational lengths and nocturnal births. Predictive techniques have the potential to streamline the monitoring process, reduce labor intensity, and minimize costs. Research on foaling prediction in mares carrying mule or equine clone fetuses is scarce. Therefore, this study aimed to comparatively analyze foaling prediction parameters in mares pregnant with mule, equine, or equine clone fetus. The study included vulvar relaxation, sacroiliac ligament tension, pH, BRIX index, and concentrations of calcium, phosphorus, magnesium, sodium, and potassium in prepartum mammary secretions. Sixty pregnant mares were used for this study and grouped as follows: 25 mares with mule fetuses (MF), 20 with equine clone fetuses (CF), and 15 with equine control fetuses (EF). Results showed significant differences in vulvar relaxation and sacroiliac ligament tension only in MF group (p < 0.05) on the day of parturition compared to the other days evaluated, different from the other groups. Levels of pH notably decreased on parturition day (mean 5.7 ± 0.04, p < 0.0001), with lower values in MF (6.05 ± 0.02) and CF (6.08 ± 0.04) compared to EF (6.26 ± 0.04) (p < 0.03). The BRIX index showed variation across mares and was not a good parameter for foaling prediction. Electrolytes correlated positively with impending parturition, showing no significant differences among groups. The MF and CF groups exhibited a substantial increase (102.13 % and 110.66 %, respectively) in mean calcium concentrations on the day before foaling, unlike EF (38.29 %). In conclusion, the pH values were different in mammary secretions between mares carrying mule and clone fetuses, in contrast to equine control fetuses. Nevertheless, there was a trend of decreasing pH values closer to parturition in all groups. Conversely, the BRIX index serves as a valuable indicator of colostrum quality yet does not offer insights into the proximity of parturition. While electrolyte concentrations did not reveal significant differences among groups, it is worth noting that the evaluation of phosphorus emerges as a new parameter to explore in mares nearing parturition, since it obtained a pattern similar to calcium.

Apply multiple machine learning models to diabetes prediction

Diabetes mellitus, a pervasive and chronic metabolic disorder, imposes a substantial burden on global health systems due to its requirement for lifelong management and the myriad of complications associated with inadequate control. The ability to accurately forecast the onset of this disease is paramount, as it enables preemptive interventions and tailored treatment strategies that can significantly mitigate its impact. This paper investigates the application of machine learning techniques and deep learning models in diabetes prediction. This paper makes use of the Pima Indian Diabetes Dataset (PIDD) from Kaggle, which has 768 data entries and eight characteristics like blood pressure, blood sugar, and body mass index (BMI). Various algorithms, including Support Vector Machine (SVM), Decision Trees(DT), Random Forest(RF) , and Fully Connected Neural Network (FCNN), are implemented and compared. Identifying the strengths and limitations of each model, the results emphasize the potential of advanced computational models in improving the accuracy and clinical usefulness of diabetes prediction. The best-performing model is the FCNN model, with a test accuracy of 78.67% and an AUC value of 83.36%.

Multimodal Machine Learning-Based Marker Enables Early Detection and Prognosis Prediction for Hyperuricemia.

Hyperuricemia (HUA) has emerged as the second most prevalent metabolic disorder characterized by prolonged and asymptomatic period, triggering gout and metabolism-related outcomes. Early detection and prognosis prediction for HUA and gout are crucial for pre-emptive interventions. Integrating genetic and clinical data from 421287 UK Biobank and 8900 Nanfang Hospital participants, a stacked multimodal machine learning model is developed and validated to synthesize its probabilities as an in-silico quantitative marker for hyperuricemia (ISHUA). The model demonstrates satisfactory performance in detecting HUA, exhibiting area under the curves (AUCs) of 0.859, 0.836, and 0.779 within the train, internal, and external test sets, respectively. ISHUA is significantly associated with gout and metabolism-related outcomes, effectively classifying individuals into low- and high-risk groups for gout in the train (AUC, 0.815) and internal test (AUC, 0.814) sets. The high-risk group shows increased susceptibility to metabolism-related outcomes, and participants with intermediate or favorable lifestyle profiles have hazard ratios of 0.75 and 0.53 for gout compared with those with unfavorable lifestyles. Similar trends are observed for other metabolism-related outcomes. The multimodal machine learning-based ISHUA marker enables personalized risk stratification for gout and metabolism-related outcomes, and it is unveiled that lifestyle changes can ameliorate these outcomes within high-risk group, providing guidance for preventive interventions.

Could prophylactic antivirals reduce dengue incidence in a high-prevalence endemic area?

Prophylactic drugs against dengue are currently under development. In this study, we explored how such prophylactic approaches might affect dengue cases in four communes of Nha Trang City, Vietnam. A community level dengue transmission survey indicated high levels of previous exposure to dengue (89.7%; 95% CI: 87.2,92.0). We fitted a spatially explicit model to an observed outbreak and simulated likely effectiveness of Case-Area Targeted Interventions (CATI) and One-Time Mass Distribution (OTMD) of drug and vector control strategies. Increasing radius and effectiveness and decreasing delay of CATI was most effective, with drugs being more effective in averting dengue cases than vector control. Using an OTMD approach early in the outbreak required the least number of treatments to avert a case, suggesting that OTMD strategies should be considered as pre-emptive rather than reactive strategies. These findings show that pre-emptive interventions can substantially reduce the burden of dengue outbreaks in endemic settings.

Parent-infant interaction trajectories in infants with an elevated likelihood for autism in relation to 3-year clinical outcome.

Developmental antecedents of autism may affect parent-infant interactions (PII), altering the context in which core social skills develop. While studies have identified differences in PII between infants with and without elevated likelihood (EL) for autism, samples have been small. Here, we examined whether previously reported differences are replicable. From a longitudinal study of 113 EL and 27 typical likelihood infants (TL), 6-min videotaped unstructured PII was blind rated at 8 and 14 months on eight interactional qualities. Autism outcome was assessed at 36 months. Linear mixed-effects models found higher parent sensitive responsiveness, nondirectiveness, and mutuality ratings in TL than EL infants with and without later autism. PII qualities at 8 (infant positive affect, parent directiveness) and 14 months (infant attentiveness to parent, mutuality) predicted 3-year autism. Attentiveness to parent decreased between 8 and 14 months in EL infants with later autism. This larger study supports previous findings of emerging alterations in PII in this group and extends on this by detecting earlier (8-month) predictive effects of PII for autism outcome and a more marked trajectory of decreased social attentiveness. The findings strengthen the evidence base to support the implementation of early preemptive interventions to support PII in infants with early autism signs.

Predictive Role of Maternal Laboratory Parameters and Inflammatory Scores in Determining Systemic Inflammatory Response Syndrome in Newborns at Birth.

The incidence of Neonatal Systemic Inflammatory Response Syndrome (SIRS) is a critical concern in neonatal care. This study aimed to identify maternal laboratory parameters predictive of SIRS in newborns, focusing on the establishment of diagnostic cutoffs and evaluating the predictive power of these biomarkers. This prospective cohort study was conducted from January 2023 to January 2024 across several regional hospitals specializing in neonatal care. It included 207 mother-newborn pairs, divided into groups based on the neonatal development of SIRS (66 cases) or its absence (141 controls). Key maternal parameters measured included inflammatory markers and liver enzymes, analyzed using standard biochemical methods. The study applied receiver operating characteristic (ROC) analysis to establish optimal cutoff values and conducted multivariate logistic regression to determine hazard ratios (HRs) for SIRS prediction, with adjustments for potential confounders. The study identified significant ROC/AUC values for several biomarkers. The neutrophil-to-lymphocyte ratio (NLR) demonstrated an AUC of 0.926, with a cutoff value of 3.64, achieving 81.8% sensitivity and 90.9% specificity (p < 0.001). The systemic immune-inflammation index (SII) showed an AUC of 0.819 and a cutoff of 769.12, with 75.8% sensitivity and 81.8% specificity (p < 0.001). Multivariate regression analysis highlighted that neonates with maternal SII values above this cutoff were three times more likely to develop SIRS (HR 3.09, 95% CI 2.21-4.17, p < 0.0001). Other notable biomarkers included dNLR and ALRI, with respective HRs of 1.88 (p = 0.018) and 1.75 (p = 0.032). These findings confirm the significant predictive value of specific maternal inflammatory markers for neonatal SIRS. These findings support the utility of these biomarkers in prenatal screening to identify neonates at increased risk of SIRS, potentially guiding preemptive clinical interventions.

Utilizing machine learning to predict employee turnover in high-stress sectors

This study investigates the application of machine learning techniques to predict employee turnover in high-stress sectors. The primary objective is to enhance retention strategies by accurately identifying potential turnover risks. The research utilizes a comprehensive dataset comprising various factors, including employee demographics, job satisfaction, performance metrics, and stress levels. Multiple machine learning algorithms, such as logistic regression, decision trees, random forests, and neural networks, are employed to build predictive models. The methodology involves data preprocessing, feature selection, model training, and evaluation. Cross-validation and hyper parameter tuning are performed to ensure the robustness and accuracy of the models. The performance of each algorithm is assessed using metrics such as accuracy, precision, recall, and the area under the receiver operating characteristic curve (AUC-ROC). Key findings reveal that machine learning models can effectively predict employee turnover, with random forests and neural networks demonstrating superior performance. Significant predictors of turnover include job satisfaction, stress levels, and performance ratings. The study concludes that integrating machine learning models into human resource practices can provide valuable insights for preemptive interventions, ultimately reducing turnover rates in high-stress environments. Future research should explore the integration of real-time data and the potential of deep learning techniques to further enhance predictive accuracy. Additionally, the ethical implications of using predictive models in HR decisions warrant careful consideration to ensure fairness and transparency. Keywords: Machine Learning (ML), Employee Turnover, Predictive Analytics, Human Resources (HR), High-Stress Sectors, Decision Trees, Random Forests, Extreme Gradient Boosting (XGBoost), Personalized Retention Strategies, Business Intelligence (BI) Tools, Data Quality, Ethical Considerations, Data Privacy, Natural Language Processing (NLP), Deep Learning, Real-time Data Analysis, Employee Engagement, Work-Life Balance, Organizational Performance, Data-Driven Insights.

Dermatologic manifestations of hereditary hemochromatosis: A systematic review.

Hereditary hemochromatosis (HH) is a genetic disorder leading to excessive iron absorption, impacting multiple organs, notably the skin, nails and mucosae. The objective of this study is to elucidate the dermatologic manifestations, associated symptoms, pathophysiology and management recommendations of HH. We searched five primary databases (PubMed, Embase, Cochrane Library, Scopus and Web of Science) up to April 2023. Non-English articles were included to minimize language bias. The studies were evaluated using Oxford Centre for Evidence-based Medicine standards, with adherence to PRISMA guidelines. Inaccessible articles were directly sourced from authors. Out of the initial 1582 publications from 1904 to 2023, 22 studies (19 in English, 2 in French and 1 in German) were selected. Most reports were from the USA, UK and France and were predominantly case reports, covering 148 patients with skin symptoms related to hereditary hemochromatosis. We collected data on the cutaneous findings and, when available, their histopathological features. The current study highlights the scope, variety and traits of dermatologic symptoms in hereditary hemochromatosis, pinpointing research gaps and areas for future exploration. Our review accentuates the diverse dermatological manifestations of hereditary hemochromatosis, notably hyperpigmentation, hypertrichosis and resistant pruritus, often linked to excessive iron deposition and subsequent impairment of skin cell function. We also found controversial evidence indicating that skin cancers seem to be associated with hereditary hemochromatosis. Porphyria cutanea tarda and hereditary hemochromatosis were frequently reported together. Given hereditary hemochromatosis's genetic nature, early identification in one individual can substantially guide familial care and preemptive interventions. Clinicians should prioritize hereditary hemochromatosis as a differential when patients present with specific dermatological symptoms, especially in sun-exposed regions. A rigorous assessment ensures accurate diagnosis, facilitating optimal management for both the patient and their family.

Prospective comparison of cytomegalovirus quantification in whole blood and plasma samples among hematopoietic stem cell transplant and kidney transplant recipients

BackgroundCytomegalovirus (CMV) induces multi-organ pathogenesis in hematopoietic stem cell transplant (HSCT) and kidney transplant (KT) recipients. Effective management involves systematic monitoring for CMV reactivation by quantitative real-time PCR, allowing timely preemptive intervention. However, the optimal blood compartment for CMV surveillance remains undetermined. ObjectiveThe aim of the study was to compare the quantification of CMV DNA in paired plasma and whole blood samples. Study designFrom June and October 2022, we conducted a prospective study with 390 sets of paired plasma and whole blood specimens collected from 60 HSCT and 24 KT recipients. CMV DNA levels were compared between the cobas® CMV assay on the automated cobas® 6800 system for plasma and the reference assay, Abbott RealTime CMV assay on the m2000 RealTime platform for whole blood. ResultsThe sensitivity and specificity of CMV quantification in plasma using the cobas® CMV assay were 90.0 % (95 %CI: 81.5 to 95.9) and 94.8 % (95 %CI: 91.8 to 96.8), respectively, compared to whole blood quantification with the Abbott assay. The overall agreement between these two strategies was 0.89 (95 %CI: 0.86–0.91). In samples with quantifiable results, a correlation was observed between the two methods (R2 = 0.62, 95 %CI: 0.65–0.87, p < 0.0001). CMV loads were significantly higher in whole blood, with a mean bias of 0.42 log10 IU/mL (95 %CI: -0.32–1.15). ConclusionThe cobas® CMV assay in plasma showed significant concordance with the Abbott RealTime CMV assay in whole blood, confirming the relevance of plasma samples for CMV monitoring in HSCT and KT recipients.

LifeGuardAI-Artificial Intelligence for Predicting Mortality Due to Sepsis

The LifeGuardAI project is a groundbreaking initiative that aims to utilize artificial intelligence to predict mortality rates associated with sepsis. The project utilizes Multilayer Perceptron (MLP) models and collaborative AI development techniques to provide healthcare professionals with advanced, AI-driven insights for preemptive intervention, ultimately enhancing patient- centered care. The project framework involves a comprehensive approach that begins with defining the problem statement focused on leveraging AI to improve sepsis-related outcomes. The dataset for this project is sourced from the Kaggle Prediction of Sepsis dataset, which contains crucial information related to patient health, such as vital signs, laboratory values, and demographic information

A novel electronic health record-based, machine-learning model to predict severe hypoglycemia leading to hospitalizations in older adults with diabetes: A territory-wide cohort and modeling study.

Older adults with diabetes are at high risk of severe hypoglycemia (SH). Many machine-learning (ML) models predict short-term hypoglycemia are not specific for older adults and show poor precision-recall. We aimed to develop a multidimensional, electronic health record (EHR)-based ML model to predict one-year risk of SH requiring hospitalization in older adults with diabetes. We adopted a case-control design for a retrospective territory-wide cohort of 1,456,618 records from 364,863 unique older adults (age ≥65 years) with diabetes and at least 1 Hong Kong Hospital Authority attendance from 2013 to 2018. We used 258 predictors including demographics, admissions, diagnoses, medications, and routine laboratory tests in a one-year period to predict SH events requiring hospitalization in the following 12 months. The cohort was randomly split into training, testing, and internal validation sets in a 7:2:1 ratio. Six ML algorithms were evaluated including logistic-regression, random forest, gradient boost machine, deep neural network (DNN), XGBoost, and Rulefit. We tested our model in a temporal validation cohort in the Hong Kong Diabetes Register with predictors defined in 2018 and outcome events defined in 2019. Predictive performance was assessed using area under the receiver operating characteristic curve (AUROC), area under the precision-recall curve (AUPRC) statistics, and positive predictive value (PPV). We identified 11,128 SH events requiring hospitalization during the observation periods. The XGBoost model yielded the best performance (AUROC = 0.978 [95% CI 0.972 to 0.984]; AUPRC = 0.670 [95% CI 0.652 to 0.688]; PPV = 0.721 [95% CI 0.703 to 0.739]). This was superior to an 11-variable conventional logistic-regression model comprised of age, sex, history of SH, hypertension, blood glucose, kidney function measurements, and use of oral glucose-lowering drugs (GLDs) (AUROC = 0.906; AUPRC = 0.085; PPV = 0.468). Top impactful predictors included non-use of lipid-regulating drugs, in-patient admission, urgent emergency triage, insulin use, and history of SH. External validation in the HKDR cohort yielded AUROC of 0.856 [95% CI 0.838 to 0.873]. Main limitations of this study included limited transportability of the model and lack of geographically independent validation. Our novel-ML model demonstrated good discrimination and high precision in predicting one-year risk of SH requiring hospitalization. This may be integrated into EHR decision support systems for preemptive intervention in older adults at highest risk.