Health consumers' ethical concerns towards artificial intelligence in Australian emergency departments.

OBJECTIVES: To investigate health consumers' ethical concerns towards the use of artificial intelligence (AI) in EDs. METHODS: Qualitative semi-structured interviews with health consumers, recruited via health consumer networks and community groups, interviews conducted between January and August 2022. RESULTS: We interviewed 28 health consumers about their perceptions towards the ethical use of AI in EDs. The results discussed in this paper highlight the challenges and barriers for the effective and ethical implementation of AI from the perspective of Australian health consumers. Most health consumers are more likely to support AI health tools in EDs if they continue to be involved in the decision-making process. There is considerably more approval of AI tools that support clinical decision-making, as opposed to replacing it. There is mixed sentiment about the acceptability of AI tools influencing clinical decision-making and judgement. Health consumers are mostly supportive of the use of their data to train and develop AI tools but are concerned with who has access. Addressing bias and discrimination in AI is an important consideration for some health consumers. Robust regulation and governance are critical for health consumers to trust and accept the use of AI. CONCLUSION: Health consumers view AI as an emerging technology that they want to see comprehensively regulated to ensure it functions safely and securely with EDs. Without considerations made for the ethical design, implementation and use of AI technologies, health consumer trust and acceptance in the use of these tools will be limited.

Questioning the Role of Carotid Artery Ultrasound in Assessing Fluid Responsiveness in Critical Illness: A Systematic Review and Meta-Analysis.

Background: A noninvasive and accurate method of identifying fluid responsiveness in hemodynamically unstable patients has long been sought by physicians. Carotid ultrasound (US) is one such modality previously canvassed for this purpose. The aim of this novel systematic review and meta-analysis is to investigate whether critically unwell patients who are requiring intravenous (IV) fluid resuscitation (fluid responders) can be identified accurately with carotid US. Methods: The protocol was registered with PROSPERO on the 30/11/2022 (ID number: CRD42022380284). Studies investigating carotid ultrasound accuracy in assessing fluid responsiveness in hemodynamically unstable patients were included. Studies were identified through searches of six databases, all run on 4 November 2022, Medline, Embase, Emcare, APA PsycInfo, CINAHL, and Cochrane Library. Risk of bias was assessed using the QUADAS-2 and the Grading of Recommendations, Assessment, Development, and Evaluations (GRADE) guidelines. Results were pooled, meta-analysis was conducted where amenable, and hierarchical summary receiver operating characteristic models were established to compare carotid ultrasound measures. Results: Seventeen studies were included (n = 842), with 1048 fluid challenges. 441 (42.1%) were fluid responsive. Four different carotid US measures were investigated, including change in carotid doppler peak velocity (∆CDPV), carotid blood flow (CBF), change in carotid artery velocity time integral (∆CAVTI), and carotid flow time (CFT). Pooled carotid US had a pooled sensitivity, specificity, and AUROC with 95% confidence intervals (CI) of 0.73 (0.66-0.78), 0.82 (0.72-0.90), and 0.81 (0.78-0.85), respectively. ∆CDPV had sensitivity, specificity, and AUROC with 95% CI of 0.72 (0.64-0.80), 0.87 (0.73-0.94), and 0.82 (0.78-0.85), respectively. CBF had sensitivity, specificity, and AUROC with 95% CI of 0.70 (0.56-0.80), 0.80 (0.50-0.94), and 0.77 (0.78-0.85), respectively. Risk of bias and assessment was undertaken using the QUADAS-2 and GRADE tools. The QUADAS-2 found that studies generally had an unclear or high risk of bias but with low applicability concerns. The GRADE assessment showed that ∆CDPV and CBF had low accuracy for sensitivity and specificity. Conclusion: It appears that carotid US has a limited ability to predict fluid responsiveness in critically unwell patients. ∆CDPV demonstrates the greatest accuracy of all measures analyzed. Further high-quality studies using consistent study design would help confirm this.

Improving quantitative susceptibility mapping for the identification of traumatic brain injury neurodegeneration at the individual level.

BACKGROUND: Emerging evidence suggests that traumatic brain injury (TBI) is a major risk factor for developing neurodegenerative disease later in life. Quantitative susceptibility mapping (QSM) has been used by an increasing number of studies in investigations of pathophysiological changes in TBI. However, generating artefact-free quantitative susceptibility maps in brains with large focal lesions, as in the case of moderate-to-severe TBI (ms-TBI), is particularly challenging. To address this issue, we utilized a novel two-pass masking technique and reconstruction procedure (two-pass QSM) to generate quantitative susceptibility maps (QSMxT; Stewart et al., 2022, Magn Reson Med.) in combination with the recently developed virtual brain grafting (VBG) procedure for brain repair (Radwan et al., 2021, NeuroImage) to improve automated delineation of brain areas. We used QSMxT and VBG to generate personalised QSM profiles of individual patients with reference to a sample of healthy controls. METHODS: Chronic ms-TBI patients (N = 8) and healthy controls (N = 12) underwent (multi-echo) GRE, and anatomical MRI (MPRAGE) on a 3T Siemens PRISMA scanner. We reconstructed the magnetic susceptibility maps using two-pass QSM from QSMxT. We then extracted values of magnetic susceptibility in grey matter (GM) regions (following brain repair via VBG) across the whole brain and determined if they deviate from a reference healthy control group [Z-score < -3.43 or > 3.43, relative to the control mean], with the aim of obtaining personalised QSM profiles. RESULTS: Using two-pass QSM, we achieved susceptibility maps with a substantial increase in quality and reduction in artefacts irrespective of the presence of large focal lesions, compared to single-pass QSM. In addition, VBG minimised the loss of GM regions and exclusion of patients due to failures in the region delineation step. Our findings revealed deviations in magnetic susceptibility measures from the HC group that differed across individual TBI patients. These changes included both increases and decreases in magnetic susceptibility values in multiple GM regions across the brain. CONCLUSIONS: We illustrate how to obtain magnetic susceptibility values at the individual level and to build personalised QSM profiles in ms-TBI patients. Our approach opens the door for QSM investigations in more severely injured patients. Such profiles are also critical to overcome the inherent heterogeneity of clinical populations, such as ms-TBI, and to characterize the underlying mechanisms of neurodegeneration at the individual level more precisely. Moreover, this new personalised QSM profiling could in the future assist clinicians in assessing recovery and formulating a neuroscience-guided integrative rehabilitation program tailored to individual TBI patients.

Predicting Fluid Responsiveness Using Carotid Ultrasound in Mechanically Ventilated Patients: A Systematic Review and Meta-Analysis of Diagnostic Test Accuracy Studies.

BACKGROUND: A noninvasive and accurate method of determining fluid responsiveness in ventilated patients would help to mitigate unnecessary fluid administration. Although carotid ultrasound has been previously studied for this purpose, several studies have recently been published. We performed an updated systematic review and meta-analysis to evaluate the accuracy of carotid ultrasound as a tool to predict fluid responsiveness in ventilated patients. METHODS: Studies eligible for review investigated the accuracy of carotid ultrasound parameters in predicting fluid responsiveness in ventilated patients, using sensitivity and specificity as markers of diagnostic accuracy (International Prospective Register of Systematic Reviews [PROSPERO] CRD42022380284). All included studies had to use an independent method of determining cardiac output and exclude spontaneously ventilated patients. Six bibliographic databases and 2 trial registries were searched. Medline, Embase, Emcare, APA PsycInfo, CINAHL, and the Cochrane Library were searched on November 4, 2022. Clinicaltrials.gov and Australian New Zealand Clinical Trials Registry were searched on February 24, 2023. Results were pooled, meta-analysis was conducted where possible, and hierarchical summary receiver operating characteristic models were used to compare carotid ultrasound parameters. Bias and evidence quality were assessed using the Quality Assessment of Diagnostic Accuracy Studies (QUADAS) tool and the Grading of Recommendations, Assessment, Development, and Evaluations (GRADE) guidelines. RESULTS: Thirteen prospective clinical studies were included (n = 648 patients), representing 677 deliveries of volume expansion, with 378 episodes of fluid responsiveness (58.3%). A meta-analysis of change in carotid Doppler peak velocity (∆CDPV) yielded a sensitivity of 0.79 (95% confidence interval [CI], 0.74-0.84) and a specificity of 0.85 (95% CI, 0.76-0.90). Risk of bias relating to recruitment methodology, the independence of index testing to reference standards and exclusionary clinical criteria were evaluated. Overall quality of evidence was low. Study design heterogeneity, including a lack of clear parameter cutoffs, limited the generalizability of our results. CONCLUSIONS: In this meta-analysis, we found that existing literature supports the ability of carotid ultrasound to predict fluid responsiveness in mechanically ventilated adults. ∆CDPV may be an accurate carotid parameter in certain contexts. Further high-quality studies with more homogenous designs are needed to further validate this technology.

Machine learning in clinical practice: Evaluation of an artificial intelligence tool after implementation.

OBJECTIVE: Artificial intelligence (AI) has gradually found its way into healthcare, and its future integration into clinical practice is inevitable. In the present study, we evaluate the accuracy of a novel AI algorithm designed to predict admission based on a triage note after clinical implementation. This is the first of such studies to investigate real-time AI performance in the emergency setting. METHODS: The novel AI algorithm that predicts admission using a triage note was translated into clinical practice and integrated within St Vincent's Hospital Melbourne's electronic emergency patient management system. The data were collected from 1 January 2021 to 17 August 2022 to evaluate the diagnostic accuracy of the AI system after implementation. RESULTS: A total of 77 125 ED presentations were included. The live AI algorithm has a sensitivity of 73.1% (95% confidence interval 72.5-73.8), specificity of 74.3% (73.9-74.7), positive predictive value of 50% (49.6-50.4) and negative predictive value of 88.7% (88.5-89) with a total accuracy of 74% (73.7-74.3). The accuracy of the system was at the lowest for admission to psychiatric units (34%) and at the highest for gastroenterology and medical admission (84% and 80%, respectively). CONCLUSION: Our study showed the diagnostic evaluation of a real-time AI clinical decision-support tool became less accurate than the original. Although real-time sensitivity and specificity of the AI tool was still acceptable as a decision-support tool in the ED, we propose that continuous training and evaluation of AI-enabled clinical support tools in healthcare are conducted to ensure consistent accuracy and performance to prevent inadvertent consequences.

Attitudes towards artificial intelligence in emergency medicine.

OBJECTIVE: To assess Australian and New Zealand emergency clinicians' attitudes towards the use of artificial intelligence (AI) in emergency medicine. METHODS: We undertook a qualitative interview-based study based on grounded theory. Participants were recruited through ED internal mailing lists, the Australasian College for Emergency Medicine Bulletin, and the research teams' personal networks. Interviews were transcribed, coded and themes presented. RESULTS: Twenty-five interviews were conducted between July 2021 and May 2022. Thematic saturation was achieved after 22 interviews. Most participants were from either Western Australia (52%) or Victoria (16%) and were consultants (96%). More participants reported feeling optimistic (10/25) than neutral (6/25), pessimistic (2/25) or mixed (7/25) towards the use of AI in the ED. A minority expressed scepticism regarding the feasibility or value of implementing AI into the ED. Multiple potential risks and ethical issues were discussed by participants including skill loss from overreliance on AI, algorithmic bias, patient privacy and concerns over liability. Participants also discussed perceived inadequacies in existing information technology systems. Participants felt that AI technologies would be used as decision support tools and not replace the roles of emergency clinicians. Participants were not concerned about the impact of AI on their job security. Most (17/25) participants thought that AI would impact emergency medicine within the next 10 years. CONCLUSIONS: Emergency clinicians interviewed were generally optimistic about the use of AI in emergency medicine, so long as it is used as a decision support tool and they maintain the ability to override its recommendations.

Exploring personalized structural connectomics for moderate to severe traumatic brain injury.

Graph theoretical analysis of the structural connectome has been employed successfully to characterize brain network alterations in patients with traumatic brain injury (TBI). However, heterogeneity in neuropathology is a well-known issue in the TBI population, such that group comparisons of patients against controls are confounded by within-group variability. Recently, novel single-subject profiling approaches have been developed to capture inter-patient heterogeneity. We present a personalized connectomics approach that examines structural brain alterations in five chronic patients with moderate to severe TBI who underwent anatomical and diffusion magnetic resonance imaging. We generated individualized profiles of lesion characteristics and network measures (including personalized graph metric GraphMe plots, and nodal and edge-based brain network alterations) and compared them against healthy reference cases (N = 12) to assess brain damage qualitatively and quantitatively at the individual level. Our findings revealed alterations of brain networks with high variability between patients. With validation and comparison to stratified, normative healthy control comparison cohorts, this approach could be used by clinicians to formulate a neuroscience-guided integrative rehabilitation program for TBI patients, and for designing personalized rehabilitation protocols based on their unique lesion load and connectome.

Individualised profiling of white matter organisation in moderate-to-severe traumatic brain injury patients.

BACKGROUND AND PURPOSE: Approximately 65% of moderate-to-severe traumatic brain injury (m-sTBI) patients present with poor long-term behavioural outcomes, which can significantly impair activities of daily living. Numerous diffusion-weighted MRI studies have linked these poor outcomes to decreased white matter integrity of several commissural tracts, association fibres and projection fibres in the brain. However, most studies have focused on group-based analyses, which are unable to deal with the substantial between-patient heterogeneity in m-sTBI. As a result, there is increasing interest and need in conducting individualised neuroimaging analyses. MATERIALS AND METHODS: Here, we generated a detailed subject-specific characterisation of microstructural organisation of white matter tracts in 5 chronic patients with m-sTBI (29 - 49y, 2 females), presented as a proof-of-concept. We developed an imaging analysis framework using fixel-based analysis and TractLearn to determine whether the values of fibre density of white matter tracts at the individual patient level deviate from the healthy control group (n = 12, 8F, Mage = 35.7y, age range 25 - 64y). RESULTS: Our individualised analysis revealed unique white matter profiles, confirming the heterogenous nature of m-sTBI and the need of individualised profiles to properly characterise the extent of injury. Future studies incorporating clinical data, as well as utilising larger reference samples and examining the test-retest reliability of the fixel-wise metrics are warranted. CONCLUSIONS: Individualised profiles may assist clinicians in tracking recovery and planning personalised training programs for chronic m-sTBI patients, which is necessary to achieve optimal behavioural outcomes and improved quality of life.

Intubation Rates following Prehospital Administration of Ketamine for Acute Agitation: A Systematic Review and Meta-Analysis.

BACKGROUND: Ketamine is a fast-acting, dissociative anesthetic with a favorable adverse effect profile that is effective for managing acute agitation as a chemical restraint in the prehospital and emergency department (ED) settings. However, some previously published individual studies have reported high intubation rates when ketamine was administered prehospitally. OBJECTIVE: This systematic review aims to determine the rate and settings in which intubation following prehospital administration of ketamine for agitation is occurring, as well as associated indications and adverse events. METHODS: We searched PubMed, Scopus, Ovid MEDLINE, Embase, CINAHL Plus, PsycINFO, the Cochrane Library, ClinicalTrials.gov, OpenGrey, Open Access Theses and Dissertation, and Google Scholar from the earliest possible date until 13/February/2022. Inclusion criteria required studies to describe agitated patients who received ketamine in the prehospital setting as a first-line drug to control acute agitation. Reference lists of appraised studies were screened for additional relevant articles. Study quality was assessed using the Newcastle-Ottawa quality assessment scale. Synthesis of results was completed via meta-analysis, and the GRADE tool was used for certainty assessment. RESULTS: The search yielded 1466 unique records and abstracts, of which 50 full texts were reviewed, resulting in 18 being included in the analysis. All studies were observational in nature and 15 were from USA. There were 3476 patients in total, and the overall rate of intubation was 16% (95% confidence interval [CI] = 8%-26%). Most intubations occurred in the ED. Within the studies, the prehospital intubation rate ranged from 0% to 7.9% and the ED intubation rate ranged from 0 to 60%. The overall pooled prehospital intubation rate was 1% (95% CI = 0%-2%). The overall pooled ED intubation rate was 19% (95% CI = 11%-30%). The most common indications for intubation were for airway protection and respiratory depression/failure. CONCLUSIONS: There is wide variation in intubation rates between and within studies. The majority of intubations performed following prehospital administration of ketamine for agitation took place in the ED.

Mental health consequences of COVID-19 suppression strategies in Victoria, Australia: a narrative review.

The COVID-19 pandemic has imposed significant mental health burdens upon the general population worldwide, either directly owing to the disease or indirectly through aggressive public health measures to control spread of the virus that causes COVID-19. In this narrative review, we used a systematic approach to summarize the impact of restrictive lockdown measures on the general mental health of people living in Victoria, Australia during 2020 and to identify the groups with an increased risk of adverse mental health outcomes. A systematic database search (Ovid Medline, PsycINFO, Embase) for articles examining the mental health of Victorians in the context of the COVID-19 pandemic during 2020 yielded 88 articles, of which 15 articles were finally included in this review. We found that the general mental health of Victorians was negatively affected by COVID-19 restrictions during 2020. Although studies reported heterogeneous mental health outcomes, we found that the general population consistently used coping strategies and demonstrated mental health help-seeking behaviors in response to the restrictions. Women, children, young people, carers, people who became unemployed owing to the pandemic, and those with pre-existing psychiatric conditions had a higher risk of adverse mental health consequences during the COVID-19 pandemic in 2020.

Effects of the COVID-19 pandemic on healthcare delivery to an immigrant population in the Islamic Republic of Iran.

Background: Healthcare inequity has widely affected marginalized and immigrant communities globally during the COVID-19 pandemic. Aims: This study assessed the effect of COVID-19 pandemic on health care delivery to immigrant populations in Isfahan Province, Islamic Republic of Iran. Methods: All 67 hospitals across Isfahan Province were included in this study conducted from 1 March to 31 May 2020. Data on clinical manifestations, comorbidities, patient management, and outcomes of patients during hospital admission were extracted from medical records and analysed using SPSS for chi-square and odds ratio (OR). Results: One hundred and sixty-eight (3.3%) of 5128 PCR-confirmed COVID-19 cases during the study period were immigrants and were included in the study. There were no differences in sex, clinical presentation, comorbidities, and length of hospital stay between the non-immigrant and immigrant groups. Immigrant patients were significantly younger and had poorer outcomes, including tracheal intubation [OR = 1.9, 95% confidence interval (CI): 1.2-3.1); P = 0.009] and in-hospital mortality (OR = 1.6; 95% CI: 1.1-2.4; P = 0.02). Conclusion: Adverse health outcomes among immigrant communities may be an indication of health inequity and should be addressed by the relevant policymakers.

Most common principal diagnoses assigned to Australian emergency department presentations involving alcohol use: a multi-centre study.

OBJECTIVES: Alcohol is the most widely consumed psychoactive substance in Australia and the consequences of alcohol consumption have enormous personal and social impacts. This study aimed to describe the principal diagnoses of emergency department (ED) presentations involving alcohol use in the previous 12 hours at eight hospitals in Victoria and the Australian Capital Territory, Australia. METHODS: Twelve months' data (1 July 2018 - 30 June 2019) were collected from eight EDs, including demographics, ICD-10 codes, hospital location and self-reported drinking in the preceding 12 hours. The ten most common ICD-10 discharge codes were analysed based on age, sex and hospital geographic area. RESULTS: ICD codes pertaining to mental and behavioural disorders due to alcohol use accounted for the highest proportion in most EDs. Suicide ideation/attempt was in the five highest ICD codes for all but one hospital. It was the second most common alcohol-related presentation for both males and females. CONCLUSIONS: Alcohol plays a major role in a range of presentations, especially in relation to mental health and suicide. IMPLICATIONS FOR PUBLIC HEALTH: The collection of alcohol involvement in ED presentations represents a major step forward in informing the community about the burden of alcohol on their health resources.

Epidemiology, clinical presentations and outcome of patients presenting to the emergency department after a COVID-19 vaccination: An observational study.

OBJECTIVE: The World Health Organization declared the COVID-19 pandemic on 11 March 2020. In 2021, several vaccines were provisionally approved to reduce the risk of transmission and hospitalisation of COVID-19 infection. A surge in COVID-19 vaccination was seen between August and October 2021 in Victoria, Australia. We hypothesised this led to an increase in ED presentations. METHODS: Patients in the present study were adults who presented to the ED within 21 days of receiving a dose of a COVID-19 vaccine between 11 August 2021 and 14 November 2021. All cases underwent chart reviews to extract epidemiological features, clinical presentations, ED assessments, investigations and disposition. RESULTS: Notably, 968 patients were included in the study, comprising 6.1% of all ED presentations during the study period. The median age was 31 years. 82.9% of patients were younger than 45 years. 20.1% of patients arrived by ambulance. Chest pain was the most common presenting complaint (43.6%), followed by headache (10.3%) and palpitations (8.2%). The most common investigations were a full blood examination (73.5%), an ECG (63.8%) and serum troponin (49.1%). 64.8% of patients were directly discharged home and 22.1% were sent home after a short stay admission. Only 2.2% of patients were admitted to the hospital. CONCLUSION: A majority of patients who presented to the ED after their COVID vaccinations were young and discharged home after the initial assessment. These presentations have significantly increased the workload in prehospital settings and EDs, contributing to increased investigation usage, ED treatment space occupancy, and increased costs to the health system.

Attitudes and regard for specific medical conditions among Australian emergency medicine clinicians.

OBJECTIVE: To quantify the attitude ED clinicians hold towards patients presenting with different medical conditions, including a novel pandemic condition. METHODS: A cross-sectional study of emergency doctors and nurses utilising the Medical Condition Regard Scale (MCRS); a validated tool used to capture the bias and emotions of clinicians towards individual medical conditions. The five conditions presented to participants each represent a classical medical, complex medical, psychiatric/substance use, somatoform and a novel medical condition. RESULTS: One hundred and ninety-six clinicians were included in the study including 116 nurses and 80 doctors. Concerning each condition, both medical and nursing staff demonstrated the highest regard for a classical medical condition (58 ± 5 and 57 ± 6, respectively). Significantly different from the classical medical condition, the lowest MCRS scores were for the somatoform condition (36 ± 10) for emergency doctors and the substance use condition (39 ± 11) for emergency nurses. Regard for a novel condition (i.e., COVID-19 infection) was comparably high among both cohorts. CONCLUSION: Emergency doctors and nurses generally hold lower regard for complex medical conditions with behavioural components, including substance use disorders and somatoform conditions.

Emergency medicine patient wait time multivariable prediction models: a multicentre derivation and validation study.

OBJECTIVE: Patients, families and community members would like emergency department wait time visibility. This would improve patient journeys through emergency medicine. The study objective was to derive, internally and externally validate machine learning models to predict emergency patient wait times that are applicable to a wide variety of emergency departments. METHODS: Twelve emergency departments provided 3 years of retrospective administrative data from Australia (2017-2019). Descriptive and exploratory analyses were undertaken on the datasets. Statistical and machine learning models were developed to predict wait times at each site and were internally and externally validated. Model performance was tested on COVID-19 period data (January to June 2020). RESULTS: There were 1 930 609 patient episodes analysed and median site wait times varied from 24 to 54 min. Individual site model prediction median absolute errors varied from±22.6 min (95% CI 22.4 to 22.9) to ±44.0 min (95% CI 43.4 to 44.4). Global model prediction median absolute errors varied from ±33.9 min (95% CI 33.4 to 34.0) to ±43.8 min (95% CI 43.7 to 43.9). Random forest and linear regression models performed the best, rolling average models underestimated wait times. Important variables were triage category, last-k patient average wait time and arrival time. Wait time prediction models are not transferable across hospitals. Models performed well during the COVID-19 lockdown period. CONCLUSIONS: Electronic emergency demographic and flow information can be used to approximate emergency patient wait times. A general model is less accurate if applied without site-specific factors.

Effects of the Response to the COVID-19 Pandemic on Assault-Related Head Injury in Melbourne: A Retrospective Study.

Assault is the leading preventable cause of death, traumatic brain injury (TBI), and associated mental health problems. The COVID-19 pandemic has had a profound impact on patterns of interpersonal violence across the world. In this retrospective cross-sectional study, we analysed medical records of 1232 assault victims (domestic violence: 111, random assault: 900, prison assault: 221) with head injuries who presented to the emergency department (ED) at St Vincent's Hospital in Melbourne, Australia, a city with one of the longest and most severe COVID-19 restrictions worldwide. We examined changes in prevalence in the assault group overall and in domestic violence, random assault, and prison assault victims, comparing data from 19.5 months before and after the first day of COVID-19 restrictions in Melbourne. Moreover, we investigated differences driven by demographic factors (Who: age group, sex, and nationality) and clinical variables (Where: assault location, and When: time of arrival to the ED and time from moment of injury until presentation at ED). Descriptive statistics and chi-square analyses were performed. We found the COVID-19 pandemic significantly affected the Where of assault-related TBI, with a shift in the location of assaults from the street to the home, and the increase at home being driven by random assaults on middle-aged adults. Overall, we observed that 86% of the random assault cases were males, whereas 74% of the domestic assault cases were females. Meanwhile, nearly half (44%) of the random assault victims reported alcohol consumption versus a fifth (20%) of domestic violence victims. These findings will have direct implications for developing screening tools and better preventive and ameliorative interventions to manage the sequelae of assault TBI, particularly in the context of future large-scale health crises or emergencies.

Nature and extent of selection bias resulting from convenience sampling in the emergency department.

BACKGROUND: To compare the clinical and demographic variables of patients who present to the ED at different times of the day in order to determine the nature and extent of potential selection bias inherent in convenience sampling METHODS: We undertook a retrospective, observational study of data routinely collected in five EDs in 2019. Adult patients (aged ≥18 years) who presented with abdominal or chest pain, headache or dyspnoea were enrolled. For each patient group, the discharge diagnoses (primary outcome) of patients who presented during the day (08:00-15:59), evening (16:00-23:59), and night (00:00-07:59) were compared. Demographics, triage category and pain score, and initial vital signs were also compared. RESULTS: 2500 patients were enrolled in each of the four patient groups. For patients with abdominal pain, the diagnoses differed significantly across the time periods (p<0.001) with greater proportions of unspecified/unknown cause diagnoses in the evening (47.4%) compared with the morning (41.7%). For patients with chest pain, heart rate differed (p<0.001) with a mean rate higher in the evening (80 beats/minute) than at night (76). For patients with headache, mean patient age differed (p=0.004) with a greater age in the daytime (46 years) than the evening (41). For patients with dyspnoea, discharge diagnoses differed (p<0.001). Asthma diagnoses were more common at night (12.6%) than during the daytime (7.5%). For patients with dyspnoea, there were also differences in gender distribution (p=0.003), age (p<0.001) and respiratory rates (p=0.003) across the time periods. For each patient group, the departure status differed across the time periods (p<0.001). CONCLUSION: Patients with abdominal or chest pain, headache or dyspnoea differ in a range of clinical and demographic variables depending upon their time of presentation. These differences may potentially introduce selection bias impacting upon the internal validity of a study if convenience sampling of patients is undertaken.

Outcomes for emergency department patients with suspected and confirmed COVID-19: An analysis of the Australian experience in 2020 (COVED-5).

OBJECTIVE: The aim of the present study was to describe the characteristics and outcomes of patients presenting to Australian EDs with suspected and confirmed COVID-19 during 2020, and to determine the predictors of in-hospital death for SARS-CoV-2 positive patients. METHODS: This analysis from the COVED Project presents data from 12 sites across four Australian states for the period from 1 April to 30 November 2020. All adult patients who met local criteria for suspected COVID-19 and underwent testing for SARS-CoV-2 in the ED were eligible for inclusion. Study outcomes were mechanical ventilation and in-hospital mortality. RESULTS: Among 24 405 eligible ED presentations over the whole study period, 423 tested positive for SARS-CoV-2. During the 'second wave' from 1 July to 30 September 2020, 26 (6%) of 406 SARS-CoV-2 patients received invasive mechanical ventilation, compared to 175 (2%) of the 9024 SARS-CoV-2 negative patients (odds ratio [OR] 3.5; 95% confidence interval [CI] 2.3-5.2, P < 0.001), and 41 (10%) SARS-CoV-2 positive patients died in hospital compared to 312 (3%) SARS-CoV-2 negative patients (OR 3.2; 95% CI 2.2-4.4, P = 0.001). For SARS-CoV-2 positive patients, the strongest independent predictors of hospital death were age (OR 1.1; 95% CI 1.1-1.1, P < 0.001), higher triage category (OR 3.5; 95% CI 1.3-9.4, P = 0.012), obesity (OR 4.2; 95% CI 1.2-14.3, P = 0.024) and receiving immunosuppressive treatment (OR 8.2; 95% CI 1.8-36.7, P = 0.006). CONCLUSIONS: ED patients who tested positive for SARS-CoV-2 had higher odds of mechanical ventilation and death in hospital. The strongest predictors of death were age, a higher triage category, obesity and receiving immunosuppressive treatment.

Predicting Ambulance Patient Wait Times: A Multicenter Derivation and Validation Study.

STUDY OBJECTIVE: To derive and internally and externally validate machine-learning models to predict emergency ambulance patient door-to-off-stretcher wait times that are applicable to a wide variety of emergency departments. METHODS: Nine emergency departments provided 3 years (2017 to 2019) of retrospective administrative data from Australia. Descriptive and exploratory analyses were undertaken on the datasets. Statistical and machine-learning models were developed to predict wait times at each site and were internally and externally validated. RESULTS: There were 421,894 episodes analyzed, and median site off-load times varied from 13 (interquartile range [IQR], 9 to 20) to 29 (IQR, 16 to 48) minutes. The global site prediction model median absolute errors were 11.7 minutes (95% confidence interval [CI], 11.7 to 11.8) using linear regression and 12.8 minutes (95% CI, 12.7 to 12.9) using elastic net. The individual site model prediction median absolute errors varied from the most accurate at 6.3 minutes (95% CI, 6.2 to 6.4) to the least accurate at 16.1 minutes (95% CI, 15.8 to 16.3). The model technique performance was the same for linear regression, random forests, elastic net, and rolling average. The important variables were the last k-patient average waits, triage category, and patient age. The global model performed at the lower end of the accuracy range compared with models for the individual sites but was within tolerable limits. CONCLUSION: Electronic emergency demographic and flow information can be used to estimate emergency ambulance patient off-stretcher times. Models can be built with reasonable accuracy for multiple hospitals using a small number of point-of-care variables.

Advanced natural language processing technique to predict patient disposition based on emergency triage notes.

OBJECTIVE: To demonstrate the potential of machine learning and capability of natural language processing (NLP) to predict disposition of patients based on triage notes in the ED. METHODS: A retrospective cohort of ED triage notes from St Vincent's Hospital (Melbourne) was used to develop a deep-learning algorithm that predicts patient disposition. Bidirectional Encoder Representations from Transformers, a recent language representation model developed by Google, was utilised for NLP. Eighty percent of the dataset was used for training the model and 20% was used to test the algorithm performance. Ktrain library, a wrapper for TensorFlow Keras, was employed to develop the model. RESULTS: The accuracy of the algorithm was 83% and the area under the curve was 0.88. Sensitivity, specificity, precision and F1-score of the algorithm were 72%, 86%, 56% and 63%, respectively. CONCLUSION: Machine learning and NLP can be together applied to the ED triage note to predict patient disposition with a high level of accuracy. The algorithm can potentially assist ED clinicians in early identification of patients requiring admission by mitigating the cognitive load, thus optimises resource allocation in EDs.