Convolutional Neural Network to Classify Infrared Thermal Images of Fractured Wrists in Pediatrics.

Convolutional neural network (CNN) models were devised and evaluated to classify infrared thermal (IRT) images of pediatric wrist fractures. The images were recorded from 19 participants with a wrist fracture and 21 without a fracture (sprain). The injury diagnosis was by X-ray radiography. For each participant, 299 IRT images of their wrists were recorded. These generated 11,960 images (40 participants × 299 images). For each image, the wrist region of interest (ROI) was selected and fast Fourier transformed (FFT) to obtain a magnitude frequency spectrum. The spectrum was resized to 100 × 100 pixels from its center as this region represented the main frequency components. Image augmentations of rotation, translation and shearing were applied to the 11,960 magnitude frequency spectra to assist with the CNN generalization during training. The CNN had 34 layers associated with convolution, batch normalization, rectified linear unit, maximum pooling and SoftMax and classification. The ratio of images for the training and test was 70:30, respectively. The effects of augmentation and dropout on CNN performance were explored. Wrist fracture identification sensitivity and accuracy of 88% and 76%, respectively, were achieved. The CNN model was able to identify wrist fractures; however, a larger sample size would improve accuracy.

Building artificial intelligence and machine learning models : a primer for emergency physicians.

There has been a rise in the number of studies relating to the role of artificial intelligence (AI) in healthcare. Its potential in Emergency Medicine (EM) has been explored in recent years with operational, predictive, diagnostic and prognostic emergency department (ED) implementations being developed. For EM researchers building models de novo, collaborative working with data scientists is invaluable throughout the process. Synergism and understanding between domain (EM) and data experts increases the likelihood of realising a successful real-world model. Our linked manuscript provided a conceptual framework (including a glossary of AI terms) to support clinicians in interpreting AI research. The aim of this paper is to supplement that framework by exploring the key issues for clinicians and researchers to consider in the process of developing an AI model.

Comparison of qSOFA and Hospital Early Warning Scores for prognosis in suspected sepsis in emergency department patients: a systematic review.

BACKGROUND: Sepsis is a major cause of morbidity and mortality and many tools exist to facilitate early recognition. This review compares two tools: the quick Sequential Organ Failure Assessment (qSOFA) and Early Warning Scores (National/Modified Early Warning Scores (NEWS/MEWS)) for predicting intensive care unit (ICU) admission and mortality when applied in the emergency department. METHODS: A literature search was conducted using Medline, CINAHL, Embase and Cochrane Library, handsearching of references and a grey literature search with no language or date restrictions. Two authors selected studies and quality assessment completed using QUADAS-2. Area under the receiver operating characteristic curve (AUROC), sensitivities and specificities were compared. RESULTS: 13 studies were included, totalling 403 865 patients. All reported mortality and six reported ICU admission.The ranges for AUROC estimates varied from little better than chance to good prediction of mortality (NEWS: 0.59-0.88; qSOFA: 0.57-0.79; MEWS 0.56-0.75), however, individual papers generally reported higher AUROC values for NEWS than qSOFA. NEWS values demonstrated a tendency towards better sensitivity for ICU admission (NEWS ≥5, 46%-91%; qSOFA ≥2, 12%-53%) and mortality (NEWS ≥5, 51%-97%; qSOFA ≥2, 14%-71%) but lower specificity (ICU: NEWS ≥5, 25%-91%; qSOFA ≥2, 67%-99%; mortality: NEWS ≥5, 22%-91%; qSOFA ≥2, 58%-99%). CONCLUSION: The wide range of AUROC estimates and high heterogeneity limit our conclusions. Allowing for this, the NEWS AUROC was consistently higher than qSOFA within individual papers. Both scores allow threshold setting, determined by the preferred compromise between sensitivity and specificity. At established thresholds NEWS tended to higher sensitivity while qSOFA tended to a higher specificity. PROSPERO REGISTRATION NUMBER: CRD42019131414.

A case report of blunt cardiac rupture.

INTRODUCTION: Cardiac rupture is a full thickness laceration of the myocardium that occurs after blunt chest trauma. They are notoriously fatal, with only a handful of patients documented to have survived. These injuries are not commonly associated with low energy chest trauma and may be overlooked as a differential in trauma cases if health care providers are not aware of their possibility. We now report the case of a patient who survived this injury. We believe this is the 16th reported survivor of blunt cardiac rupture. PRESENTATION OF CASE: A 46-year-old construction worker was brought to the emergency department following blunt chest trauma. On arrival he was hypotensive and tachycardic. There was a transient response to intravenous fluid resuscitation. He was found to have a contusion to the left anterior chest wall and left haemothorax on imaging with normal mediastinum. Emergency left anterolateral thoracotomy revealed a pericardial and left ventricular laceration which was repaired. He was weaned off ventilatory support on day 4 post exploration and had an uneventful recovery. DISCUSSION: Survival after blunt cardiac rupture is extremely low. In the past 60 years, only 15 cases have been described where patients survived this injury. They are usually immediately fatal and are caused by high velocity injuries. Our case was interesting because, this injury was due to a low velocity injury resulting in ventricular laceration due to a displaced rib fracture. We believe this is the 16th reported survivor of blunt cardiac rupture. CONCLUSION: It is important, therefore, for first responders to recognize that blunt cardiac rupture can also result from seemingly innocuous, low velocity injuries so that the diagnosis can at least be entertained early. Survival depends on early diagnosis and prompt thoracotomy.

Infrared thermal imaging as a screening tool for paediatric wrist fractures.

Wrist injuries are common in paediatric trauma; however, only half of children evaluated with an x-ray for possible fractures will have one. Thermal imaging offers a possible non-ionising method of screening for fractures and thus reducing negative x-ray rates. One hundred five children attending the Emergency Department for wrist injuries were recruited. Two 30-s thermal videos were recorded from injured and uninjured wrists-in flat and 45° elevated positions. A region of interest (ROI) was defined on each wrist. Cases in which the ROI was covered or had ice applied were excluded, leaving 40 patients for analysis. Comparisons of ROI included (i) injured and uninjured wrists-flat and elevated positions; (ii) as in (i) with a reference region on the proximal forearm subtracted; (iii) injured wrist ROI-flat and elevated positions. Fractures and sprains increased the mean skin surface temperature by 1.519% (p = 0.008) and 0.971% (p = 0.055) respectively compared with the uninjured wrist. The mean temperature difference between flat and elevated positions for fractures was 0.268% and - 0.1291% for sprains. This difference was statistically significant for fracture (p = 0.004) but not sprain (p = 0.500). The temperature differences recorded by thermal imaging between fractured and sprained wrists may assist in differentiation of these injuries. Graphical abstract Operational stages involved from thermal video recording to generation of results.

Why do parents present to the Paediatric Emergency Department with conditions suitable for management in less acute settings? A qualitative study.

OBJECTIVE: To determine the factors influencing attendance at a Paediatric Emergency Department with conditions suitable for management in less acute settings. METHODS: Semi-structured interviews exploring parental decision-making processes surrounding care-seeking behaviours in urgent situations were undertaken with parents of children allocated to the two lowest triage categories at a type-1 urban Paediatric Emergency Department over two 2-week periods. RESULTS: All parents expressed a desire to access care from the most appropriate service for their child in unscheduled situations. A number of factors influenced parental perceptions of appropriateness: (1) perceptions concerning the urgency or severity of a condition; (2) uncertainty regarding their assessment of the severity or urgency of a problem; (3) the possibility that a condition may require resources available solely at the Paediatric Emergency Department; and (4) awareness of alternative services. Parents reported that previous utilisation of alternative services, which lead to them being referred onwards to the Paediatric Emergency Department, influenced their decision to attend the Paediatric Emergency Department on this occasion. CONCLUSION: Parental uncertainty and a low tolerance of risk among parents and healthcare professionals increases the likelihood of emergency department attendance. Parents are likely to continue to utilise the Paediatric Emergency Department for minor conditions. Future strategies should focus on improving outcomes for children with non-urgent problems who access care at the emergency department rather than discouraging utilisation.

Clinical emergency care research in low-income and middle-income countries: opportunities and challenges.

Disease processes that frequently require emergency care constitute approximately 50% of the total disease burden in low-income and middle-income countries (LMICs). Many LMICs continue to deal with emergencies caused by communicable disease states such as pneumonia, diarrhoea, malaria and meningitis, while also experiencing a marked increase in non-communicable diseases, such as cardiovascular diseases, diabetes mellitus and trauma. For many of these states, emergency care interventions have been developed through research in high-income countries (HICs) and advances in care have been achieved. However, in LMICs, clinical research, especially interventional trials, in emergency care are rare. Furthermore, there exists minimal research on the emergency management of diseases, which are rarely encountered in HICs but impact the majority of LMIC populations. This paper explores challenges in conducting clinical research in patients with emergency conditions in LMICs, identifies examples of successful clinical research and highlights the system, individual and study design characteristics that made such research possible in LMICs. Derived from the available literature, a focused list of high impact research considerations are put forth.

Implementing the National Institute for Health and Clinical Excellence Head Injury 2014 Guidelines in a major children's hospital emergency department.

OBJECTIVES AND BACKGROUND: Head injury is a common paediatric emergency department presentation. The National Institute for Health and Clinical Excellence updated its guidance in January 2014 regarding imaging required for adults and children following a head injury (CG176). This study looked at the rates of computed tomography (CT) head scans performed and adherence rates to CG176. PATIENTS AND METHODS: A single-centre audit was carried out, examining imaging practice in children with head injuries. CG176 was implemented formally in August 2014 to the new trainee doctors. The primary outcome was adherence to CG176. As the data were binary, 95% confidence intervals were used for comparison. RESULTS: In all, 1797 patients were identified as having a head injury. Implementation at the Sheffield's Children NHS Foundation Trust resulted in a statistically significant increase in guideline adherence from 79.2% [95% confidence interval (CI): 76.4-81.9%] to 85.1% (95% CI: 82.9-87.4%). The greatest impact in adherence was found in CT head scans, from 95.8% (95% CI: 94.5-97.2%) to 97.7% (95% CI: 96.7-98.6%). CONCLUSION: The implementation at the Sheffield's Children NHS Foundation Trust was successful in satisfying the aim of CG176 by increasing adherence and decreasing CT head scans. This success could be explained by the formal implementation to the new cohort of doctors and better physician agreement with the guidelines. The increase in adherence is contrary to the previous studies.

Which observed behaviours may reassure physicians that a child is not septic? An international Delphi study.

OBJECTIVE: In an attempt to improve the diagnosis of sepsis in children, diagnostic aids have concentrated on clinical features that suggest that sepsis is present. Clinicians need to be able to clinically rule out sepsis as well as rule it in. Little is known about which features are consistent with wellness and/or absence of sepsis. Guidelines are therefore likely to improve sensitivity without preserving specificity. We aimed to gather expert opinion on which (if any) features would make clinicians consider a child to be unlikely to have sepsis. DESIGN: We undertook a modified two-round international Delphi study, where clinicians were asked for features they believed were indicators of wellness in an ill child. PARTICIPANTS: One hundred and ninety-five clinicians (predominantly physicians) who routinely assessed unwell children and had been doing so for most of their careers. RESULTS: Over 90% of respondents rated age-appropriate verbalisation, playing, smiling and activity as reassuring that a child was unlikely to have sepsis. Eating, spontaneous interaction and normal movement were also agreed to be reassuring by over 70% of participants. Consolability and showing fear of the clinician were not felt to be adequately reassuring. There was wide range of opinion on how reassuring the use of an electronic device was thought to be. CONCLUSIONS: This study confirms that physicians are reassured by specific behaviours in ill children, and provides a framework which may be used to help guide the assessment of the unwell child. Validation of individual features could lead to improved specificity of diagnostic aids for diagnosing sepsis.

Primary care services co-located with Emergency Departments across a UK region: early views on their development.

BACKGROUND: Co-location of primary care services with Emergency Departments (ED) is one initiative aiming to reduce the burden on EDs of patients attending with non-urgent problems. However, the extent to which these services are operating within or alongside EDs is not currently known.This study aimed to create a typology of co-located primary care services in operation across Yorkshire and Humber (Y&H) as well as identify early barriers and facilitators to their implementation and sustainability. METHODS: A self-report survey was sent to the lead consultant or other key contact at 17 EDs in the Y&H region to establish the extent and configuration of co-located primary care services. Semi-structured interviews were then conducted with urgent and unscheduled care stakeholders across five hospital sites to explore the barriers and facilitators to the formation and sustainability of these services. RESULTS: Thirteen EDs completed the survey and interviews were carried out with four ED consultants, one ED nurse and three general practitioners (GPs). Three distinct models were identified: 'Primary Care Services Embedded within the ED' (seven sites), 'Co-located Urgent Care Centre' (two sites) and 'GP out-of-hours' (nine sites). Qualitative data were analysed using framework analysis. Four interview themes emerged (justification for the service, level of integration, referral processes and sustainability) highlighting some of the challenges in implementing these co-located primary care services. CONCLUSION: Creating a service within or alongside the ED in which GPs can use their distinct skills and therefore add value to the existing skill mix of ED staff is an important consideration when setting up these systems. Effective triage arrangements should also be established to ensure appropriate patients are referred to GPs. Further research is required to identify the full range of models nationally and to carry out a rigorous assessment of their impact.

AHEAD Study: an observational study of the management of anticoagulated patients who suffer head injury.

OBJECTIVES: Management of anticoagulated patients after head injury is unclear due to lack of robust evidence. This study aimed to determine the adverse outcome rate in these patients and identify risk factors associated with poor outcome. DESIGN: Multicentre, observational study using routine patient records. SETTING: 33 emergency departments in England and Scotland. PARTICIPANTS: 3566 adults (aged ≥16 years) who had suffered blunt head injury and were currently taking warfarin. MAIN OUTCOME MEASURES: Primary outcome measure was rate of adverse outcome defined as death or neurosurgery following initial injury, clinically significant CT scan finding or reattendance with related complication within 10 weeks of initial hospital attendance. Secondary objectives included identifying risk factors for adverse outcome using univariable and multivariable analyses. RESULTS: Clinical data available for 3534/3566 patients (99.1%), median age 79 years; mean initial international normalised ratio (INR) 2.67 (SD 1.34); 81.2% Glasgow Coma Scale (GCS) 15: 59.8% received a CT scan with significant head injury-related finding in 5.4% (n=208); 0.5% underwent neurosurgery; 1.2% patients suffered a head injury-related death. Overall adverse outcome rate was 5.9% (95% CI 5.2% to 6.7%). Patients with GCS=15 and no associated symptoms had lowest risk of adverse outcome (risk 2.7%; 95% CI 2.1 to 3.6). Patients with GCS=15 multivariable analysis (using imputation) found risk of adverse outcome to increase when reporting at least one associated symptom: vomiting (relative risk (RR) 1.8; 95% CI 1.0 to 3.4), amnesia (RR 3.5; 95% CI 2.1 to 5.7), headache (RR 1.3; 95% CI 0.8 to 2.2), loss of consciousness (RR 1.75; 95% CI 1.0 to 3.0). INR measurement did not predict adverse outcome in patients with GCS=15 (RR 1.1; 95% CI 1.0 to 1.2). CONCLUSIONS: In alert warfarinised patients following head injury, the presence of symptoms is associated with greater risk of adverse outcome. Those with GCS=15 and no symptoms are a substantial group and have a low risk of adverse outcome. TRIAL REGISTRATION NUMBER: NCT02461498.

Infrared thermography in paediatrics: a narrative review of clinical use.

BACKGROUND: Infrared thermography (IRT) has been used in adult medicine for decades, but recent improvements in quality of imaging and increasing computer processing power have allowed for a diversification of clinical applications. The specific usage of IRT in a paediatric population has not been widely explored, so this article aims to summarise the available literature in this area. IRT involves the non-contact, accurate measurement of skin surface temperature to identify temperature changes suggesting disease. IRT could well have unique applications in paediatric medicine. METHODS: Electronic searches were performed independently by two authors, using the databases of MEDLINE (via Web of Science), the Cochrane Library, CINAHL (EBSCO) and Scopus, including articles published from 1990 to July 2016. The search strategy that was used aimed to include articles that covered the topics of IRT and children, including studies with participants 18 years old or younger. Articles were screened by title and abstract by two authors. Meta-analysis was not performed due to the marked heterogeneity in applications, study design and outcomes: this is a narrative summary of the available literature. RESULTS: IRT has been shown to be an effective additional diagnostic tool in a number of different paediatric specialties, namely in fracture screening, burns assessment and neonatal monitoring. Small measurable skin temperature changes can effectively add to the clinical picture, while computer-tracking systems can be reliably used to focus investigations on particular areas of the body. CONCLUSION: Throughout this review of the available literature, there has been a general consensus that this non-invasive, non-irradiating and relatively inexpensive technology may well have a place in the management of paediatric patients in the future.

Ipsilateral Plastic Deformation Monteggia and Galeazzi-Type Fracture in a Child: A Case Report.

A 7-year-old boy attended the emergency department after falling from a climbing frame onto his outstretched left wrist. On examination, there was mild swelling to the left elbow and tenderness to the antecubital fossa. There was also tenderness diffusely to the distal ulnar and radius. There was no neurovascular deficit. Radiographs revealed a plastic deformation fracture of the left radius and ulna, with dislocations of the ipsilateral radiocapitellar joint and distal radioulnar joint. A diagnosis of combined Monteggia and Galeazzi-type fractures of the left forearm was made. It is rare to find cases of combined Monteggia and Galeazzi fractures to the same forearm. Furthermore, to our knowledge, ipsilateral plastic deformation Monteggia and Galeazzi-type fractures in children have not been reported in the literature.

Should all anticoagulated patients with head injury receive a CT scan? Decision-analysis modelling of an observational cohort.

OBJECTIVES: It is not currently clear whether all anticoagulated patients with a head injury should receive CT scanning or only those with evidence of traumatic brain injury (eg, loss of consciousness or amnesia). We aimed to determine the cost-effectiveness of CT for all compared with selective CT use for anticoagulated patients with a head injury. DESIGN: Decision-analysis modelling of data from a multicentre observational study. SETTING: 33 emergency departments in England and Scotland. PARTICIPANTS: 3566 adults (aged ≥16 years) who had suffered blunt head injury, were taking warfarin and underwent selective CT scanning. MAIN OUTCOME MEASURES: Estimated expected benefits in terms of quality-adjusted life years (QALYs) were the entire cohort to receive a CT scan; estimated increased costs of CT and also the potential cost implications associated with patient survival and improved health. These values were used to estimate the cost per QALY of implementing a strategy of CT for all patients compared with observed practice based on guidelines recommending selective CT use. RESULTS: Of the 1420 of 3534 patients (40%) who did not receive a CT scan, 7 (0.5%) suffered a potentially avoidable head injury-related adverse outcome. If CT scanning had been performed in all patients, appropriate treatment could have gained 3.41 additional QALYs but would have incurred £193 149 additional treatment costs and £130 683 additional CT costs. The incremental cost-effectiveness ratio of £94 895/QALY gained for unselective compared with selective CT use is markedly above the threshold of £20-30 000/QALY used by the UK National Institute for Care Excellence to determine cost-effectiveness. CONCLUSIONS: CT scanning for all anticoagulated patients with head injury is not cost-effective compared with selective use of CT scanning based on guidelines recommending scanning only for those with evidence of traumatic brain injury. TRIAL REGISTRATION NUMBER: NCT 02461498.

Primary care services located with EDs: a review of effectiveness.

BACKGROUND: Primary care focused unscheduled care centres (UCC) co-located with major EDs have been proposed as a solution to the rise in ED attendances. They aim to reduce the burden of primary care patients attending the ED, hence reducing crowding, waits and cost.This review analysed available literature in the context of the impact of general practitioner (GP) delivered, hospital-based (adjacent or within the ED) unscheduled care services on process outcomes, cost-effectiveness and patient satisfaction. METHODS: A narrative literature review of studies published between 1980 and 2015 was undertaken. All study types were reviewed and included if they reported a service model using hospital-based primary care clinicians with a control consisting of standard ED clinician-delivered care. RESULTS: The electronic searches yielded 7544 citations, with 20 records used in the review. These were grouped into three main themes: process outcomes, cost-effectiveness and satisfaction. A paradoxical increase in attendances has been described, which is likely to be attributable to provider-induced demand, and the evidence for improved throughput is poor. Marginal savings may be realised per patient, but this is likely to be overshadowed by the overall cost of introducing a new service. CONCLUSIONS: There is little evidence to support the implementation of co-located UCC models. A robust evaluation of proposed models is needed to inform future policy.

The safety of emergency medicine.

The patient safety movement has been active for over a decade, but the issue of patient safety in emergency care and the emergency department (ED) has only recently been brought into the forefront. The ED environment has traditionally been considered unsafe, but there is little data to support this assertion. This paper reviews the literature on patient safety and highlights the challenges associated with using the current evidence base to inform practice due to the variability in methods of measuring safety. Studies looking at safety in the ED report low rates for adverse events ranging from 3.6 to 32.6 events per 1000 attendances. The wide variation in reported rates on adverse events reflects the significant differences in methods of reporting and classifying safety incidents and harm between departments; standardisation in the ED context is urgently required to allow comparisons to be made between departments and to quantify the impact of specific interventions. We outline the key factors in emergency care which may hinder the provision of safer care and consider solutions which have evolved or been proposed to identify and mitigate against harm. Interventions such as team training, telephone follow-up, ED pharmacist interventions and rounding, all show some evidence of improving safety in the ED. We further highlight the need for a collaborative whole system approach as almost half of safety incidents in the ED are attributable to external factors, particularly those related to information flow, crowding, demand and boarding.