Mind the GAHP: A novel protocol for improved vascular access in the hypotensive patient.

Background: Obtaining intravenous access in hypotensive patients is challenging and may critically delay resuscitation. The Graduated Vascular Access for Hypotensive Patient (GAHP) protocol leverages intraosseous fluid boluses to specifically dilate proximal veins. This study aims to evaluate the efficacy of GAHP in maximizing venous targets through early distal intraosseous access and a small fluid bolus. Methods: This was a prospective randomized cadaveric pilot study to evaluate extremity venous engorgement during intraosseous infusion. Cadavers (n = 23) had an intraosseous needle inserted into four sites: distal radius, proximal humerus, distal femur, and distal tibia. Intraosseous saline was rapidly infused, venous optimization was measured using real-time ultrasound. Primary outcome was maximum vessel circumference increase with intraosseous infusion. Secondary outcomes were: time to maximum circumference, and infusion volume required. Statistical analyses included Levene's test for equality of variances, Wilcoxon signed-rank test, and generalized estimating equation. Results: There was a significant mean increase of 1.03 cm (95% CI 0.86, 1.20), representing a difference of 102%. We found no significant difference in time to optimize vessel circumference across sites, but volume required significantly differed. Conclusion: GAHP quickly and effectively increased the circumference of anatomically adjacent veins. Anatomical sites did not differ on time to reach maximum enlargement of vessels following intraosseous infusion but did differ in terms of volume required to maximize vessel circumference. Further research is needed using live, hypotensive patients.

Variable importance analysis with interpretable machine learning for fair risk prediction.

Machine learning (ML) methods are increasingly used to assess variable importance, but such black box models lack stability when limited in sample sizes, and do not formally indicate non-important factors. The Shapley variable importance cloud (ShapleyVIC) addresses these limitations by assessing variable importance from an ensemble of regression models, which enhances robustness while maintaining interpretability, and estimates uncertainty of overall importance to formally test its significance. In a clinical study, ShapleyVIC reasonably identified important variables when the random forest and XGBoost failed to, and generally reproduced the findings from smaller subsamples (n = 2500 and 500) when statistical power of the logistic regression became attenuated. Moreover, ShapleyVIC reasonably estimated non-significant importance of race to justify its exclusion from the final prediction model, as opposed to the race-dependent model from the conventional stepwise model building. Hence, ShapleyVIC is robust and interpretable for variable importance assessment, with potential contribution to fairer clinical risk prediction.

Impact and effect of imaging referral guidelines on patients and radiology services: a systematic review.

OBJECTIVES: The objective of this systematic review was to offer a comprehensive overview and explore the associated outcomes from imaging referral guidelines on various key stakeholders, such as patients and radiologists. MATERIALS AND METHODS: An electronic database search was conducted in Medline, Embase and Web of Science to retrieve citations published between 2013 and 2023. The search was constructed using medical subject headings and keywords. Only full-text articles and reviews written in English were included. The quality of the included papers was assessed using the mixed methods appraisal tool. A narrative synthesis was undertaken for the selected articles. RESULTS: The search yielded 4384 records. Following the abstract, full-text screening, and removal of duplication, 31 studies of varying levels of quality were included in the final analysis. Imaging referral guidelines from the American College of Radiology were most commonly used. Clinical decision support systems were the most evaluated mode of intervention, either integrated or standalone. Interventions showed reduced patient radiation doses and waiting times for imaging. There was a general reduction in radiology workload and utilisation of diagnostic imaging. Low-value imaging utilisation decreased with an increase in the appropriateness of imaging referrals and ratings and cost savings. Clinical effectiveness was maintained during the intervention period without notable adverse consequences. CONCLUSION: Using evidence-based imaging referral guidelines improves the quality of healthcare and outcomes while reducing healthcare costs. Imaging referral guidelines are one essential component of improving the value of radiology in the healthcare system. CLINICAL RELEVANCE STATEMENT: There is a need for broader dissemination of imaging referral guidelines to healthcare providers globally in tandem with the harmonisation of the application of these guidelines to improve the overall value of radiology within the healthcare system. KEY POINTS: The application of imaging referral guidelines has an impact and effect on patients, radiologists, and health policymakers. The adoption of imaging referral guidelines in clinical practice can impact healthcare costs and improve healthcare quality and outcomes. Implementing imaging referral guidelines contributes to the attainment of value-based radiology.

Impact of Bystander Cardiopulmonary Resuscitation on Out-of-Hospital Cardiac Arrest Outcome in Vietnam.

Introduction: Patients experiencing an out-of-hospital cardiac arrest (OHCA) frequently do not receive bystander cardiopulmonary resuscitation (CPR), especially in low- and middle-income countries (LMIC). In this study we sought to determine the prevalence of OHCA patients in Vietnam who received bystander CPR and its effects on survival outcomes. Methods: We performed a multicenter, retrospective observational study of patients (≥18 years) presenting with OHCA at three major hospitals in an LMIC from February 2014-December 2018. We collected data on the hospital and patient characteristics, the cardiac arrest events, the emergency medical services (EMS) system, the therapy methods, and the outcomes and compared these data, before and after pairwise 1:1 propensity score matching, between patients who received bystander CPR and those who did not. Upon admission, we assessed factors associated with good neurological survival at hospital discharge in univariable and multivariable logistic models. Results: Of 521 patients, 388 (74.5%) were men, and the mean age was 56.7 years (SD 17.3). Although most cardiac arrests (68.7%, 358/521) occurred at home and 78.8% (410/520) were witnessed, a low proportion (22.1%, 115/521) of these patients received bystander CPR. Only half of the patients were brought by EMS (8.1%, 42/521) or private ambulance (42.8%, 223/521), 50.8% (133/262) of whom had resuscitation attempts. Before matching, there was a significant difference in good neurological survival between patients who received bystander CPR (12.2%, 14/115) and patients who did not (4.7%, 19/406; P < .001). After matching, good neurological survival was absent in all OHCA patients who did not receive CPR from a bystander. The multivariable analysis showed that bystander CPR (adjusted odds ratio: 3.624; 95% confidence interval 1.629-8.063) was an independent predictor of good neurological survival. Conclusion: In our study, only 22.1% of total OHCA patients received bystander CPR, which contributed significantly to a low rate of good neurological survival in Vietnam. To improve the chances of survival with good neurological functions of OHCA patients, more people should be trained to perform bystander CPR and teach others as well. A standard program for emergency first-aid training is necessary for this purpose.

Cost-effectiveness analysis of a 'Termination of Resuscitation' protocol for the management of out-of-hospital cardiac arrest.

BACKGROUND: Historically in Singapore, all out-of-hospital cardiac arrests (OHCA) were transported to hospital for pronouncement of death. A 'Termination of Resuscitation' (TOR) protocol, implemented from 2019 onwards, enables emergency responders to pronounce death at-scene in Singapore. This study aims to evaluate the cost-effectiveness of the TOR protocol for OHCA management. METHODS: Adopting a healthcare provider's perspective, a Markov model was developed to evaluate three competing options: No TOR, Observed TOR reflecting existing practice, and Full TOR if TOR is exercised fully. The model had a cycle duration of 30 days after the initial state of having a cardiac arrest, and was evaluated over a 10-year time horizon. Probabilistic sensitivity analysis was performed to account for uncertainties. The costs per quality adjusted life years (QALY) was calculated. RESULTS: A total of 3,695 OHCA cases eligible for the TOR protocol were analysed; mean age of 73.0 ± 15.5 years. For every 10,000 hypothetical patients, Observed TOR and Full TOR had more deaths by approximately 19 and 31 patients, respectively, compared to No TOR. Full TOR had the least costs and QALYs at $19,633,369 (95% Uncertainty Interval (UI) 19,469,973 to 19,796,764) and 0 QALYs. If TOR is exercised for every eligible case, it could expect to save approximately $400,440 per QALY loss compared to No TOR, and $821,151 per QALY loss compared to Observed TOR. CONCLUSION: The application of the TOR protocol for the management of OHCA was found to be cost-effective within acceptable willingness-to-pay thresholds, providing some justification for sustainable adoption.

Cardiac Arrest and Cardiopulmonary Resuscitation Outcome Reports: 2024 Update of the Utstein Out-of-Hospital Cardiac Arrest Registry Template.

The Utstein Out-of-Hospital Cardiac Arrest Resuscitation Registry Template, introduced in 1991 and updated in 2004 and 2015, standardizes data collection to enable research, evaluation, and comparisons of systems of care. The impetus for the current update stemmed from significant advances in the field and insights from registry development and regional comparisons. This 2024 update involved representatives of the International Liaison Committee on Resuscitation and used a modified Delphi process. Every 2015 Utstein data element was reviewed for relevance, priority (core or supplemental), and improvement. New variables were proposed and refined. All changes were voted on for inclusion. The 2015 domains-system, dispatch, patient, process, and outcomes-were retained. Further clarity is provided for the definitions of out-of-hospital cardiac arrest attended resuscitation and attempted resuscitation. Changes reflect advancements in dispatch, early response systems, and resuscitation care, as well as the importance of prehospital outcomes. Time intervals such as emergency medical service response time now emphasize precise reporting of the times used. New flowcharts aid the reporting of system effectiveness for patients with an attempted resuscitation and system efficacy for the Utstein comparator group. Recognizing the varying capacities of emergency systems globally, the writing group provided a minimal dataset for settings with developing emergency medical systems. Supplementary variables are considered useful for research purposes. These revisions aim to elevate data collection and reporting transparency by registries and researchers and to advance international comparisons and collaborations. The overarching objective remains the improvement of outcomes for patients with out-of-hospital cardiac arrest.

Differential spatial working memory-related functional network reconfiguration in young and older adults.

Functional brain networks have preserved architectures in rest and task; nevertheless, previous work consistently demonstrated task-related brain functional reorganization. Efficient rest-to-task functional network reconfiguration is associated with better cognition in young adults. However, aging and cognitive load effects, as well as contributions of intra- and internetwork reconfiguration, remain unclear. We assessed age-related and load-dependent effects on global and network-specific functional reconfiguration between rest and a spatial working memory (SWM) task in young and older adults, then investigated associations between functional reconfiguration and SWM across loads and age groups. Overall, global and network-level functional reconfiguration between rest and task increased with age and load. Importantly, more efficient functional reconfiguration associated with better performance across age groups. However, older adults relied more on internetwork reconfiguration of higher cognitive and task-relevant networks. These reflect the consistent importance of efficient network updating despite recruitment of additional functional networks to offset reduction in neural resources and a change in brain functional topology in older adults. Our findings generalize the association between efficient functional reconfiguration and cognition to aging and demonstrate distinct brain functional reconfiguration patterns associated with SWM in aging, highlighting the importance of combining rest and task measures to study aging cognition.

Brain networks identified by functional connectivity (FC) have preserved architectures from rest to task and across task demands. Higher similarity, implying more efficient network reconfiguration, was associated with better cognition and task performance in young adults. To examine how it may be influenced by aging, we compared whole-brain and network-level FC similarities between resting-state and spatial working memory fMRI in young and older adults. At whole-brain level and higher order cognitive networks, older adults evidenced less efficient network reconfiguration from rest to task than young adults. Importantly, more efficient reconfiguration was associated with better accuracy. This relationship relied more on internetwork connections in older adults. Despite reduced neural resources compared to young, maintaining efficient network updating still contributes to better cognition at older age.

Cardiac arrest and cardiopulmonary resuscitation outcome reports: 2024 update of the Utstein Out-of-Hospital Cardiac Arrest Registry template.

The Utstein Out-of-Hospital Cardiac Arrest Resuscitation Registry Template, introduced in 1991 and updated in 2004 and 2015, standardizes data collection to enable research, evaluation, and comparisons of systems of care. The impetus for the current update stemmed from significant advances in the field and insights from registry development and regional comparisons. This 2024 update involved representatives of the International Liaison Committee on Resuscitation and used a modified Delphi process. Every 2015 Utstein data element was reviewed for relevance, priority (core or supplemental), and improvement. New variables were proposed and refined. All changes were voted on for inclusion. The 2015 domains-system, dispatch, patient, process, and outcomes-were retained. Further clarity is provided for the definitions of out-of-hospital cardiac arrest attended resuscitation and attempted resuscitation. Changes reflect advancements in dispatch, early response systems, and resuscitation care, as well as the importance of prehospital outcomes. Time intervals such as emergency medical service response time now emphasize precise reporting of the times used. New flowcharts aid the reporting of system effectiveness for patients with an attempted resuscitation and system efficacy for the Utstein comparator group. Recognizing the varying capacities of emergency systems globally, the writing group provided a minimal dataset for settings with developing emergency medical systems. Supplementary variables are considered useful for research purposes. These revisions aim to elevate data collection and reporting transparency by registries and researchers and to advance international comparisons and collaborations. The overarching objective remains the improvement of outcomes for patients with out-of-hospital cardiac arrest.

Validation of the NULL-EASE Score for Predicting Survival in a Multiethnic Asian Cohort of Out-of-Hospital Cardiac Arrest.

BACKGROUND: NULL-PLEASE is a simple and accurate clinical scoring system developed in a Western cohort of patients with out-of-hospital cardiac arrest (OHCA). The need for blood test results limits its use in early stages of care. We adapted and validated the NULL-EASE score (without laboratory tests) in an independent, multiethnic Asian cohort of patients with out-of-hospital cardiac arrest. METHODS AND RESULTS: Using the Singapore OHCA registry, we included consecutive adult patients with out-of-hospital cardiac arrest who survived to hospital admission between April 2010 to December 2020. In-hospital mortality was the primary outcome. Logistic regression analyses were performed with STATA MP v18. Of 3274 patients (median age 64, interquartile range 54-75; 67.9% male) included in the study, 2476 (75.6%) had in-hospital mortality. NULL-EASE score was significantly lower in survivors compared with nonsurvivors (median [inter quartile range] 3 [1-4] versus 6 [4-7]; P<0.001) and strongly predictive of mortality (area under receiver operating characteristic, 0.81 [95% CI, 0.79-0.83]). Patients with a score of ≥3 had higher odds of mortality (adjusted odds ratio, 8.11 [95% CI, 6.57-10.00]) when compared with those with lower scores, after adjusting for sex, residential arrest, diabetes, respiratory disease, and stroke. A cutoff value of ≥3 predicted mortality with 92.2% sensitivity, 84.1% positive predictive value, 46.1% specificity, and 65.5% negative predictive value. NULL-EASE score performed better in younger compared with older patients (area under receiver operating characteristic, 0.82 versus 0.77, P=0.008). CONCLUSIONS: The NULL-EASE score has good discriminative performance (sensitivity and accuracy) in our multiethnic Asian cohort, but the cutoff of ≥3 falls short of the desired level of specificity for therapeutic decision-making.

Towards proactive palliative care in oncology: developing an explainable EHR-based machine learning model for mortality risk prediction.

BACKGROUND: Ex-ante identification of the last year in life facilitates a proactive palliative approach. Machine learning models trained on electronic health records (EHR) demonstrate promising performance in cancer prognostication. However, gaps in literature include incomplete reporting of model performance, inadequate alignment of model formulation with implementation use-case, and insufficient explainability hindering trust and adoption in clinical settings. Hence, we aim to develop an explainable machine learning EHR-based model that prompts palliative care processes by predicting for 365-day mortality risk among patients with advanced cancer within an outpatient setting. METHODS: Our cohort consisted of 5,926 adults diagnosed with Stage 3 or 4 solid organ cancer between July 1, 2017, and June 30, 2020 and receiving ambulatory cancer care within a tertiary center. The classification problem was modelled using Extreme Gradient Boosting (XGBoost) and aligned to our envisioned use-case: "Given a prediction point that corresponds to an outpatient cancer encounter, predict for mortality within 365-days from prediction point, using EHR data up to 365-days prior." The model was trained with 75% of the dataset (n = 39,416 outpatient encounters) and validated on a 25% hold-out dataset (n = 13,122 outpatient encounters). To explain model outputs, we used Shapley Additive Explanations (SHAP) values. Clinical characteristics, laboratory tests and treatment data were used to train the model. Performance was evaluated using area under the receiver operating characteristic curve (AUROC) and area under the precision-recall curve (AUPRC), while model calibration was assessed using the Brier score. RESULTS: In total, 17,149 of the 52,538 prediction points (32.6%) had a mortality event within the 365-day prediction window. The model demonstrated an AUROC of 0.861 (95% CI 0.856-0.867) and AUPRC of 0.771. The Brier score was 0.147, indicating slight overestimations of mortality risk. Explanatory diagrams utilizing SHAP values allowed visualization of feature impacts on predictions at both the global and individual levels. CONCLUSION: Our machine learning model demonstrated good discrimination and precision-recall in predicting 365-day mortality risk among individuals with advanced cancer. It has the potential to provide personalized mortality predictions and facilitate earlier integration of palliative care.

Improved interpretable machine learning emergency department triage tool addressing class imbalance.

Objective: The Score for Emergency Risk Prediction (SERP) is a novel mortality risk prediction score which leverages machine learning in supporting triage decisions. In its derivation study, SERP-2d, SERP-7d and SERP-30d demonstrated good predictive performance for 2-day, 7-day and 30-day mortality. However, the dataset used had significant class imbalance. This study aimed to determine if addressing class imbalance can improve SERP's performance, ultimately improving triage accuracy. Methods: The Singapore General Hospital (SGH) emergency department (ED) dataset was used, which contains 1,833,908 ED records between 2008 and 2020. Records between 2008 and 2017 were randomly split into a training set (80%) and validation set (20%). The 2019 and 2020 records were used as test sets. To address class imbalance, we used random oversampling and random undersampling in the AutoScore-Imbalance framework to develop SERP+-2d, SERP+-7d, and SERP+-30d scores. The performance of SERP+, SERP, and the commonly used triage risk scores was compared. Results: The developed SERP+ scores had five to six variables. The AUC of SERP+ scores (0.874 to 0.905) was higher than that of the corresponding SERP scores (0.859 to 0.894) on both test sets. This superior performance was statistically significant for SERP+-7d (2019: Z = -5.843, p < 0.001, 2020: Z = -4.548, p < 0.001) and SERP+-30d (2019: Z = -3.063, p = 0.002, 2020: Z = -3.256, p = 0.001). SERP+ outperformed SERP marginally on sensitivity, specificity, balanced accuracy, and positive predictive value measures. Negative predictive value was the same for SERP+ and SERP. Additionally, SERP+ showed better performance compared to the commonly used triage risk scores. Conclusions: Accounting for class imbalance during training improved score performance for SERP+. Better stratification of even a small number of patients can be meaningful in the context of the ED triage. Our findings reiterate the potential of machine learning-based scores like SERP+ in supporting accurate, data-driven triage decisions at the ED.

Effect of Cangrelor on Infarct Size in ST-Segment-Elevation Myocardial Infarction Treated by Primary Percutaneous Coronary Intervention: A Randomized Controlled Trial (The PITRI Trial).

BACKGROUND: The administration of intravenous cangrelor at reperfusion achieves faster onset of platelet P2Y12 inhibition than oral ticagrelor and has been shown to reduce myocardial infarction (MI) size in the preclinical setting. We hypothesized that the administration of cangrelor at reperfusion will reduce MI size and prevent microvascular obstruction in patients with ST-segment-elevation MI undergoing primary percutaneous coronary intervention. METHODS: This was a phase 2, multicenter, randomized, double-blind, placebo-controlled clinical trial conducted between November 2017 to November 2021 in 6 cardiac centers in Singapore. Patients were randomized to receive either cangrelor or placebo initiated before the primary percutaneous coronary intervention procedure on top of oral ticagrelor. The key exclusion criteria included presenting <6 hours of symptom onset; previous MI and stroke or transient ischemic attack; on concomitant oral anticoagulants; and a contraindication for cardiovascular magnetic resonance. The primary efficacy end point was acute MI size by cardiovascular magnetic resonance within the first week expressed as percentage of the left ventricle mass (%LVmass). Microvascular obstruction was identified as areas of dark core of hypoenhancement within areas of late gadolinium enhancement. The primary safety end point was Bleeding Academic Research Consortium-defined major bleeding in the first 48 hours. Continuous variables were compared by Mann-Whitney U test (reported as median [first quartile-third quartile]), and categorical variables were compared by Fisher exact test. A 2-sided P<0.05 was considered statistically significant. RESULTS: Of 209 recruited patients, 164 patients (78%) completed the acute cardiovascular magnetic resonance scan. There were no significant differences in acute MI size (placebo, 14.9% [7.3-22.6] %LVmass versus cangrelor, 16.3 [9.9-24.4] %LVmass; P=0.40) or the incidence (placebo, 48% versus cangrelor, 47%; P=0.99) and extent of microvascular obstruction (placebo, 1.63 [0.60-4.65] %LVmass versus cangrelor, 1.18 [0.53-3.37] %LVmass; P=0.46) between placebo and cangrelor despite a 2-fold decrease in platelet reactivity with cangrelor. There were no Bleeding Academic Research Consortium-defined major bleeding events in either group in the first 48 hours. CONCLUSIONS: Cangrelor administered at the time of primary percutaneous coronary intervention did not reduce acute MI size or prevent microvascular obstruction in patients with ST-segment-elevation MI given oral ticagrelor despite a significant reduction of platelet reactivity during the percutaneous coronary intervention procedure. REGISTRATION: URL: https://www.clinicaltrials.gov; Unique identifier: NCT03102723.

Strategies to improve implementation of cascade testing in hereditary cancer syndromes: a systematic review.

Hereditary cancer syndromes constitute approximately 10% of all cancers. Cascade testing involves testing of at-risk relatives to determine if they carry the familial pathogenic variant. Despite growing efforts targeted at improving cascade testing uptake, current literature continues to reflect poor rates of uptake, typically below 30%. This study aims to systematically review current literature on intervention strategies to improve cascade testing, assess the quality of intervention descriptions and evaluate the implementation outcomes of listed interventions. We searched major databases using keywords and subject heading of "cascade testing". Interventions proposed in each study were classified according to the Effective Practice and Organization of Care (EPOC) taxonomy. Quality of intervention description was assessed using the TIDieR checklist, and evaluation of implementation outcomes was performed using Proctor's Implementation Outcomes Framework. Improvements in rates of genetic testing uptake was seen in interventions across the different EPOC taxonomy strategies. The average TIDieR score was 7.3 out of 12. Items least reported include modifications (18.5%), plans to assess fidelity/adherence (7.4%) and actual assessment of fidelity/adherence (7.4%). An average of 2.9 out of 8 aspects of implementation outcomes were examined. The most poorly reported outcomes were cost, fidelity and sustainability, with only 3.7% of studies reporting them. Most interventions have demonstrated success in improving cascade testing uptake. Uptake of cascade testing was highest with delivery arrangement (68%). However, the quality of description of interventions and assessment of implementation outcomes are often suboptimal, hindering their replication and implementation downstream. Therefore, further adoption of standardized guidelines in reporting of interventions and formal assessment of implementation outcomes may help promote translation of these interventions into routine practice.

Prehospital stroke care in Singapore.

ABSTRACT: Due to the narrow window of opportunity for stroke therapeutics to be employed, effectiveness of stroke care systems is predicated on the efficiency of prehospital stroke systems. A robust prehospital stroke system of care that provides a rapid and well-coordinated response maximises favourable poststroke outcomes, but achieving this presents a unique set of challenges dependent on demographic and geographical circumstances. Set in the context of a highly urbanised first-world nation with a rising burden of stroke, Singapore's prehospital stroke system has evolved to reflect the environment in which it operates. This review aims to characterise the current state of prehospital stroke care in Singapore, covering prehospital aspects of the stroke survival chain from symptom onset till arrival at the emergency department. We identify areas for improvement and innovation, as well as provide insights into the possible future of prehospital stroke care in Singapore.

Development and validation of a deep learning system for detection of small bowel pathologies in capsule endoscopy: a pilot study in a Singapore institution.

INTRODUCTION: Deep learning models can assess the quality of images and discriminate among abnormalities in small bowel capsule endoscopy (CE), reducing fatigue and the time needed for diagnosis. They serve as a decision support system, partially automating the diagnosis process by providing probability predictions for abnormalities. METHODS: We demonstrated the use of deep learning models in CE image analysis, specifically by piloting a bowel preparation model (BPM) and an abnormality detection model (ADM) to determine frame-level view quality and the presence of abnormal findings, respectively. We used convolutional neural network-based models pretrained on large-scale open-domain data to extract spatial features of CE images that were then used in a dense feed-forward neural network classifier. We then combined the open-source Kvasir-Capsule dataset (n = 43) and locally collected CE data (n = 29). RESULTS: Model performance was compared using averaged five-fold and two-fold cross-validation for BPMs and ADMs, respectively. The best BPM model based on a pre-trained ResNet50 architecture had an area under the receiver operating characteristic and precision-recall curves of 0.969±0.008 and 0.843±0.041, respectively. The best ADM model, also based on ResNet50, had top-1 and top-2 accuracies of 84.03±0.051 and 94.78±0.028, respectively. The models could process approximately 200-250 images per second and showed good discrimination on time-critical abnormalities such as bleeding. CONCLUSION: Our pilot models showed the potential to improve time to diagnosis in CE workflows. To our knowledge, our approach is unique to the Singapore context. The value of our work can be further evaluated in a pragmatic manner that is sensitive to existing clinician workflow and resource constraints.

Using heart rate variability to develop a predictive model for post-operative cardiovascular complications following major abdominal surgery: A pilot study.

BACKGROUND: Heart Rate Variability (HRV) is a dynamic reflection of heart rhythm regulation by various physiological inputs. HRV deviations have been found to correlate with clinical outcomes in patients under physiological stresses. Perioperative cardiovascular complications occur in up to 5% of adult patients undergoing abdominal surgery and are associated with significantly increased mortality. This pilot study aimed to develop a predictive model for post-operative cardiovascular complications using HRV parameters for early risk stratification and aid post-operative clinical decision-making. METHODS: Adult patients admitted to High Dependency Units after elective major abdominal surgery were recruited. The primary composite outcome was defined as cardiovascular complications within 7 days post-operatively. ECG monitoring for HRV parameters was conducted at three time points (pre-operative, immediately post-operative, and post-operative day 1) and analyzed based on outcome group and time interactions. Candidate HRV predictors were included in a multivariable logistic regression analysis incorporating a stepwise selection algorithm. RESULTS: 89 patients were included in the analysis, with 8 experiencing cardiovascular complications. Three HRV parameters, when measured immediately post-operatively and composited with patient age, provided the basis for a predictive model with AUC of 0.980 (95% CI: 0.953, 1.00). The negative predictive value was 1.00 at a statistically optimal predicted probability cut-off point of 0.16. CONCLUSION: Our model holds potential for accelerating clinical decision-making and aiding in patient triaging post-operatively, using easily acquired HRV parameters. Risk stratification with our model may enable safe early step-down care in patients assessed to have a low risk profile of post-operative cardiovascular complications.

Socioeconomic disadvantage and long-term survival duration in out-of-hospital cardiac arrest patients: A population-based cohort study.

Background: Socioeconomic status (SES) is a well-established determinant of cardiovascular health. However, the relationship between SES and clinical outcomes in long-term out-of-hospital cardiac arrest (OHCA) is less well-understood. The Singapore Housing Index (SHI) is a validated building-level SES indicator. We investigated whether SES as measured by SHI is associated with long-term OHCA survival in Singapore. Methods: We conducted an open cohort study with linked data from the Singapore Pan-Asian Resuscitation Outcomes Study (PAROS), and the Singapore Registry of Births and Deaths (SRBD) from 2010 to 2020. We fitted generalized structural equation models, calculating hazard ratios (HRs) using a Weibull model. We constructed Kaplan-Meier survival curves and calculated the predicted marginal probability for each SHI category. Results: We included 659 cases. In both univariable and multivariable analyses, SHI did not have a significant association with survival. Indirect pathways of SHI mediated through covariates such as Emergency Medical Services (EMS) response time (HR of low-medium, high-medium and high SHI when compared to low SHI: 0.98 (0.88-1.10), 1.01 (0.93-1.11), 1.02 (0.93-1.12) respectively), and age of arrest (HR of low-medium, high-medium and high SHI when compared to low SHI: 1.02 (0.75-1.38), 1.08 (0.84-1.38), 1.18 (0.91-1.54) respectively) had no significant association with OHCA survival. There was no clear trend in the predicted marginal probability of survival among the different SHI categories. Conclusions: We did not find a significant association between SES and OHCA survival outcomes in residential areas in Singapore. Among other reasons, this could be due to affordable healthcare across different socioeconomic classes.

Towards advancing Out-of-Hospital cardiac arrest (OHCA) registries globally: Considerations from the global OHCA registry (GOHCAR) collaborative.

A growing number out-of-hospital cardiac arrest (OHCA) registries have been developed across the globe. A few of these are national, while others cover larger geographical regions. These registries have common objectives; continuous quality improvement, epidemiological research and providing infrastructure for clinical trials. OHCA registries make performance comparison across Emergency Medical Services systems possible for benchmarking, hypothesis generation and further research. Changes in OHCA incidence and outcomes provide insights about the effects of secular trends or health services interventions. These registries, therefore, have become a mainstay of OHCA management and research. However, developing and maintaining these registries is challenging. Coordination of different service providers to support data collection, sustainable resourcing, data quality and data security are the key challenges faced by these registries. Despite all these challenges, noteworthy progress has been made and further standardization and co-ordination across registries can result in great international benefit. In this paper we present a 'why' and 'how to' model for setting up OHCA registries, and suggestions for better international co-ordination through a Global OHCA Registries Collaborative (GOHCAR). We draw together the knowledge of a cohort of international researchers, with experience and expertise in OHCA registry development, management, and data synthesis.

Machine learning prediction of refractory ventricular fibrillation in out-of-hospital cardiac arrest using features available to EMS.

Background: Shock-refractory ventricular fibrillation (VF) or ventricular tachycardia (VT) is a treatment challenge in out-of-hospital cardiac arrest (OHCA). This study aimed to develop and validate machine learning models that could be implemented by emergency medical services (EMS) to predict refractory VF/VT in OHCA patients. Methods: This was a retrospective study examining adult non-traumatic OHCA patients brought into the emergency department by Singapore EMS from the Pan-Asian Resuscitation Outcomes Study (PAROS) registry. Data from April 2010 to March 2020 were extracted for this study. Refractory VF/VT was defined as VF/VT persisting or recurring after at least one shock. Features were selected based on expert clinical opinion and availability to dispatch prior to arrival at scene. Multivariable logistic regression (MVR), LASSO and random forest (RF) models were investigated. Model performance was evaluated using receiver operator characteristic (ROC) area under curve (AUC) analysis and calibration plots. Results: 20,713 patients were included in this study, of which 860 (4.1%) fulfilled the criteria for refractory VF/VT. All models performed comparably and were moderately well-calibrated. ROC-AUC were 0.732 (95% CI, 0.695 - 0.769) for MVR, 0.738 (95% CI, 0.701 - 0.774) for LASSO, and 0.731 (95% CI, 0.690 - 0.773) for RF. The shared important predictors across all models included male gender and public location. Conclusion: The machine learning models developed have potential clinical utility to improve outcomes in cases of refractory VF/VT OHCA. Prediction of refractory VF/VT prior to arrival at patient's side may allow for increased options for intervention both by EMS and tertiary care centres.

Corrigendum to "Development and Asian-wide validation of the Grade for Interpretable Field Triage (GIFT) for predicting mortality in pre-hospital patients using the Pan-Asian Trauma Outcomes Study (PATOS)" [The Lancet Regional Health - Western Pacific 34 (2023) 100733].

[This corrects the article DOI: 10.1016/j.lanwpc.2023.100733.].