Explore the vast range of titles available.

BackgroundHome high flow therapy (HFT) has emerged as a potentially useful respiratory support modality. NICE MedTech appraisal and NHS Scotland procurement framework provides support for prescription of home HFT in selected circumstances. Implementation experience with home HFT is required to determine feasibility and approach for further clinical trials and scale-up of service provision. Since 2018, we have offered treatment trials with home HFT on a case-by-case basis to patients with severe COPD; typically where there is severe dyspnoea, and/or frequent admissions, and/or evolving respiratory failure despite optimised management, with no indication for an alternative home respiratory support modalities.Methods/ResultsA retrospective database and electronic health record review identified 27 patients with COPD who have trialled home HFT. Home HFT setup with either a myAirvo2 (F&P) or LumisHFT (ResMed) device typically required a 60-minute daycase review or clinician demonstration during an inpatient episode. Flow and temperature settings were directed by patient comfort; most patients opted for flow rates between 25–30L/min. Reported utilisation patterns varied between patient from regular overnight use to intermittent as required use during the day. Improvements were noted in the admission profiles of patient with COPD following initiation of home NHF (table 1). The majority of patients noted a positive experience, with ease of use of the therapy and improvement in at least one symptom domain. Five patients discontinued treatment because of comfort issues and perceived lack of benefit.16/27 patients with COPD commenced home HFT while using the LenusCOPD patient app, which captures daily patient-reported outcomes (PROs) and other detailed aggregated data, for future evaluations.Abstract P60 Table 1Improvement in hospital admission profiles of patients with COPD in year following initiation of home HFT Year prior to setup of home HFT Year following setup of home HFT Mean annual hospital admissions per patient 1.7 1.1 Mean annual occupied bed days per patient 13.7 9.7 ConclusionProvision of home HFT within a breathing support service is feasible, with positive patient user experience and a reduction in hospital admissions noted in this retrospective cohort data review. Expanded use of remote-monitoring systems capturing PROs, connected therapy, physiology and event data will facilitate the required continued implementation-effectiveness evaluations of home HFT.

IntroductionCOPD is a common, progressive, preventable and treatable respiratory disorder effecting approximately 1.2 million people in the UK. It is estimated to become the third leading cause of death worldwide by 2030. Accurately identifying high-risk patients in a live clinical setting is essential for proactive care re-orientation and prioritisation.Aims and ObjectivesMachine Learning (ML) models were developed using clinical data to predict risk in COPD patients with the aim to optimise care and improve patient outcomes. This AI is being operationalised in live patient care as part of a feasibility study.MethodsWe used de-identified demographics, hospital admissions, diagnosis, prescribing and labs data from 60,000 patients to develop risk prediction models. We focused on risk of mortality, respiratory-related hospital readmission, and exacerbation prediction. 15% of all patients were held out for a final test set. All sets were checked to ensure they were drawn from a sample representative of the full population. An 80:20 split of the remaining 85% of patients, and cross validation methods, were used for model training and validation. The 12-month mortality prediction model was tested on the hold-out test dataset. Patients who were recruited to our RECEIVER COPD digital service trial (support.nhscopd.scot) had been omitted from model training and validation. We were therefore able to undertake a further retrospective evaluation on this trial data, running synthetic AI-MDTs with the model applied at patient onboarding and at monthly intervals for the following 6 months, with comparison of model predictions to events over the subsequent 12 months.ResultsThe model performed well achieving an averaged ROC-AUC of 0.83 (0.76–0.92) and an averaged PR-AUC of 0.57 (0.38–0.76) averaged over the 7 runs. Local explainability for each patient prediction was also calculated using SHAP. The feasibility study using live AI models within our COPD MDT is ongoing.ConclusionsOur 12-month mortality prediction model performed well and is ready for live adoption in the AI insights app and propspective evaluation in the DYNAMIC-AI trial. The other two models are at advanced development stage. The ongoing feasibility study will demonstrate how AI can be used as part of live patient care.

IntroductionAccurate predictions of hospital length of stay (LOS) at the time of admission allows clinicians to direct patients to the most appropriate medical services, prevent overcrowding in emergency departments via improved patient flow, and better manage hospital resources.ObjectivesTo develop, evaluate and explain machine learning classifiers that predict prolonged LOS (≥2 days) using information that is known at the time of acute admission, does not change during the patient’s hospital stay, and would be easy to input to a model deployed in a clinical setting.MethodsA SafeHaven dataset of de-identified electronic health records for acute admissions of patients with COPD to four Scottish hospitals between January 2010 and March 2019 was prepared. Using XGBoost algorithms and a binary classifier (admission <48 hours or >48 hours) we developed a set of machine-learning models that predict whether a patient will have a prolonged LOS and investigated which variables contribute the most to prediction performance. We produced separate models for: 1) all acute admissions in the study period (n=75387); 2) COPD related admissions (n=12137); 3) admissions relating to COPD or a broader set of respiratory conditions (n=20134). We evaluated model performance on an unseen test data set based on Receiver Operating Characteristic and Precision Recall Curves, and the precision, recall and F1 scores. Further, we compared models to two established clinical scores to predict emergency department disposition: the Glasgow Admission Prediction Score (GAPS) and the Ambulatory Score (Ambs). We used SHapley Additive exPlanations to explain why specific model predictions are made for individual patients.ResultsOur models highlighted several key factors that contribute to prolonged LOS in COPD patients. Some relate to patient clinical history, such as certain existing comorbidities, previous diagnoses on discharge and LOS for previous hospital visits, which is rarely considered in LOS prediction models.ConclusionsWe have identified several factors relating to clinical and admission history that influence COPD patients’ likelihood of prolonged acute admissions and are able to explain the rationale behind individual predictions. Since these factors would be known at admission time, they could be passed to a deployed LOS predictive model to aid clinical decision making.

BackgroundDigital solutions offer the opportunity to improve accessibility and uptake of strategies that improve clinical outcomes for people with COPD. LenusCOPD has been co-designed to enable digital transformation of COPD services for proactive preventative care. A patient-facing progressive web application, clinician dashboard and support website integrates patient-reported outcomes (PROs), self-management resources, structured clinical summary, wearable and home NIV data with asynchronous patient-clinician messaging. We commenced the implementation-effectiveness observational cohort RECEIVER trial in September 2019 to explore the feasibility and utility of this application alongside routine care (NCT04240353). The primary endpoint was the sustained patient usage and secondary endpoints including admissions, mortality, exacerbations, service workload and quality of life. We paused recruitment in March 2021and provided LenusCOPD as routine care in the ‘DYNAMIC-SCOT’ COVID-19 response service scale-up.Methods83 RECEIVER trial participants and 142 DYNAMIC-SCOT participants had completed minimum 1 year follow-up when we censored data on 31st August 2021. We established a control cohort with 5 patients matched per RECEIVER participant from de-identified contemporary routine clinical data held with NHS Greater Glasgow & Clyde Safe Haven.FindingsSustained patient app utilisation was noted in both cohorts, with an average of 3.5 PRO submissions per person per week. Median time to admission or death was 43 days in control, 338 days in RECEIVER and 400 days in the sub-cohort of DYNAMIC-SCOT participants who had had a respiratory-related admission in the preceding year (figure 1). The 12-month risk of admission or death was 74% in control patients, 53% in RECEIVER and 47% in the DYNAMIC-SCOT sub-cohort participants. LenusCOPD service users had a greater reduction in annual admission and occupied bed day rates compared to control patients, with a median of 2.5 community-managed COPD exacerbations per patient per year and stable quality of life scores during follow up. Patient-clinician messaging workload was manageable.Abstract P199 Figure 1Kaplan-Meier survival plots of time to readmission or death from index/onboarding date until 31st August 2021 in control, RECEIVER and DYNAMIC-SCOT cohorts (S1, subdivided by occurrence or absence of a respiratory-related admission in the year prior to onboarding to the service)InterpretationA high proportion of people continued to use the co-designed LenusCOPD application during extended follow-up. Outcome data supports scale-up of this digital service transformation. Qualitative evaluations into participant’s perceived benefits of the app are ongoing, along with formation of risk prediction models using AI-based algorithms.Please refer to page A215 for declarations of interest related to this abstract.

IntroductionDiagnosing and treating lung cancer in early stages is essential for survival outcomes. The chest X-ray (CXR) remains the primary screening tool to identify lung cancers in the UK; however, there is a shortfall of radiologists, while demand continues to increase. Image analysis by machine-learning software has the potential to support radiology workflows with a focus on immediate triage of suspicious X-rays. The RADICAL study will evaluate Qure.ai’s ‘qXR’ software in reducing reporting time for suspicious X-rays in NHS Greater Glasgow & Clyde.Methods and analysisThis is a stepped-wedge cluster-randomised study consisting of a retrospective technical evaluation and prospective clinical effectiveness study alongside the assessment of acceptability via qualitative work and evaluation of cost-effectiveness via a cost utility analysis. The primary objective is to assess the clinical effectiveness of qXR to prioritise patients suspected with lung cancer on CXR for follow-up CT. Secondary objectives will look at the utility, safety, technical performance, health economics and acceptability of the intervention. The study period is 24 months, consisting of an initial 12 month data collection period and a 12 month follow-up period. All the standard care CXRs from outpatient and primary care requests will be securely transmitted to Qure.ai software ‘qXR’ for interpretation. Images with features of cancer will be flagged as ‘Urgent Suspicion of Cancer’ and be prioritised for radiologist review within the existing reporting workflow.Ethics and disseminationThe study will follow the principles of Good Clinical Practice. The protocol was granted REC approval in August 2023 from North West—Greater Manchester West Research Ethics Committee (REC 23/NW/0211). This study was registered on clinicaltrials.gov (NCT06044454). An interim report will be produced for use by the Scottish Government. The results from this study will be presented at artificial intelligence, radiology and respiratory meetings and published in peer-reviewed journals.Trial registration numberNCT06044454.

IntroductionDiagnostic imaging is vital in emergency departments (EDs). Accessibility and reporting impacts ED workflow and patient care. With radiology workforce shortages, reporting capacity is limited, leading to image interpretation delays. Turnaround times for image reporting are an ED bottleneck. Artificial intelligence (AI) algorithms can improve productivity, efficiency and accuracy in diagnostic radiology, contingent on their clinical efficacy. This includes positively impacting patient care and improving clinical workflow. The ACCEPT-AI study will evaluate Qure.ai’s qER software in identifying and prioritising patients with critical findings from AI analysis of non-contrast head CT (NCCT) scans.Methods and analysisThis is a multicentre trial, spanning four diverse sites, over 13 months. It will include all individuals above the age of 18 years who present to the ED, referred for an NCCT. The project will be divided into three consecutive phases (pre-implementation, implementation and post-implementation of the qER solution) in a stepped-wedge design to control for adoption bias and adjust for time-based changes in the background patient characteristics. Pre-implementation involves baseline data for standard care to support the primary and secondary outcomes. The implementation phase includes staff training and qER solution threshold adjustments in detecting target abnormalities adjusted, if necessary. The post-implementation phase will introduce a notification (prioritised flag) in the radiology information system. The radiologist can choose to agree with the qER findings or ignore it according to their clinical judgement before writing and signing off the report. Non-qER processed scans will be handled as per standard care.Ethics and disseminationThe study will be conducted in accordance with the principles of Good Clinical Practice. The protocol was approved by the Research Ethics Committee of East Midlands (Leicester Central), in May 2023 (REC (Research Ethics Committee) 23/EM/0108). Results will be published in peer-reviewed journals and disseminated in scientific findings (ClinicalTrials.gov: NCT06027411)Trial registration number NCT06027411.

Digital tools may improve chronic obstructive pulmonary disease (COPD) management, but further evidence of significant, persisting benefits are required. The RECEIVER trial was devised to evaluate the Lenus COPD support service by determining if people with severe COPD would continue to utilize the co-designed patient web application throughout study follow-up and to explore the impact of this digital service on clinical outcomes with its adoption alongside routine care. The prospective observational cohort hybrid implementation-effectiveness study began in September 2019 and included 83 participants. Recruitment stopped in March 2020 due to COVID-19, but follow-up continued as planned. A contemporary matched control cohort was identified to compare participant clinical outcomes with and minimize biases associated with wider COVID-19 impacts. Utilization was determined by daily COPD assessment test (CAT) completion through the application. Survival metrics and post-index date changes in annual hospitalizations were compared between the RECEIVER and control cohorts. Longitudinal quality of life and symptom burden data and community-managed exacerbation events were also captured through the application. High and sustained application utilization was noted across the RECEIVER cohort with a mean follow-up of 78 weeks (64/83 participants completed at least one CAT entry on ≥50% of possible follow-up weeks). Subgroup analysis of participants resident in more socioeconomically deprived postcode areas revealed equivalent utilization. Median time to death or a COPD or respiratory-related admission was higher in the RECEIVER cohort compared to control (335 days vs 155 days). Mean reduction in annual occupied bed days was 8.12 days vs 3.38 days in the control cohort. Quality of life and symptom burden remained stable despite the progressive nature of COPD. The sustained utilization of the co-designed patient application and improvements in participant outcomes observed in the RECEIVER trial support scale-up implementation with continued evaluation of this digital service.

IntroductionObjective risk stratification based is recommended in all patients with a new diagnosis of stable ischaemic heart disease. However, more than half of patients with chronic coronary syndromes who have a future myocardial infarct do not have obstructive coronary disease on coronary imaging, or the ischaemic substrate to enable effective risk stratification with functional testing. There is need for an effective risk stratification tool that can be applied to all patients with chronic coronary syndrome to help guide management decisions.PurposeTo evaluate the role of cardiac troponin testing in the risk stratification of patients with chronic coronary syndrome.Method: Consecutive patients attending a tertiary cardiac centre for investigation of chronic coronary syndrome with coronary angiography were eligible for enrolment into this prospective observational study. High-sensitivity cardiac troponin I was measured in all patients immediately prior to angiography with clinicians blinded to the results. Troponin concentrations were log transformed and evaluated as a continuous variable in adjusted Cox regression models, and categorised as low (<5 ng/L), intermediate (5 ng/L - 99th centile), or high (>99th centile). The primary outcome was a composite of myocardial infarction or cardiovascular death over a median follow-up of 2.5 years.ResultsIn total, 4,344 consecutive patients were enrolled (median age 66 years (IQR 59 - 73), 32.4% female). The majority had obstructive coronary disease on angiography (62.4%, 2,712/4,344), with fewer having non-obstructive disease (27.4%, 1,193/4,344) or angiographically normal coronary arteries (10.2%, 442/4,344). Patients with obstructive disease had higher troponin levels (median 4.0 ng/L, IQR 2.1 - 8.6) than those with non-obstructive disease (2.7 ng/L, IQR 1.4 - 5.1; P<0.001). Patients with the highest troponin concentration were most likely to have a primary outcome (62.8 events per 1,000 patient-years) as compared to those with intermediate (45.5 per 1,000 patient-years) or low troponin levels (15.1 per 1,000 patient-years). In patients with obstructive disease, the incidence of the primary outcome was highest in those with the highest troponin (64.5 per 1,000 patient-years) as compared to those with obstructive disease and either intermediate or low troponin levels (53.2 and 21.2 per 1,000 patient-years, respectively). After adjusting for coronary disease severity, troponin remained an important independent predictor of the primary outcome (aHR 3.1 95%CI 2.4–3.9).Abstract 167 Figure 1Cumulative incidence of the primary outcome (myocardial infarction or cardiovascular death) in patients with obstructive coronary disease, stratified by cardiac troponin concentration [low (green, <5 ng/), intermediate (orange, 5 ng/L - 99th centile), or high (red, >99th centile)ConclusionIn patients with chronic coronary syndrome, cardiac troponin can reliably identify individuals at the highest risk of myocardial infarction or cardiovascular death. Combined with angiographic indices of disease severity, troponin testing in the chronic coronary syndrome could augment current risk stratification strategies and may inform optimised treatment decisions.Conflict of InterestNone

IntroductionFractures of the odontoid process frequently result from low impact falls in frail or older adults. These are increasing in incidence and importance as the population ages. In the UK, odontoid fractures in older adults are usually managed in hard collars to immobilise the fracture and promote bony healing. However, bony healing does not always occur in older adults, and bony healing is not associated with quality of life, functional, or pain outcomes. Further, hard collars can cause complications such as skin pressure ulcers, swallowing difficulties and difficulties with personal care. We hypothesise that management with no immobilisation may be superior to management in a hard collar for older or frail adults with odontoid fractures.Methods and analysesThis is the protocol for the Duration of External Neck Stabilisation (DENS) trial—a non-blinded randomised controlled trial comparing management in a hard collar with management without a collar for older (≥65 years) or frail (Rockwood Clinical Frailty Scale ≥5) adults with a new odontoid fracture. 887 neurologically intact participants with any odontoid process fracture type will be randomised to continuing with a hard collar (standard care) or removal of the collar (intervention). The primary outcome is quality of life measured using the EQ-5D-5L at 12 weeks. Secondary outcomes include pain scores, neck disability index, health and social care use and costs, and mortality.Ethics and disseminationInformed consent for participation will be sought from those able to provide it. We will also include those who lack capacity to ensure representativeness of frail and acutely unwell older adults. Results will be disseminated via scientific publication, lay summary, and visual abstract. The DENS trial received a favourable ethical opinion from the Scotland A Research Ethics Committee (21/SS/0036) and the Leeds West Research Ethics Committee (21/YH/0141).Trial registration numberNCT04895644.

Background Delivery of major trauma care is complex and often fast paced. Clear and comprehensive documentation is paramount to support effective communication during complex clinical care episodes, and to allow collection of data for audit, research and continuous improvement. Clinical events are typically recorded on paper-based records that are developed for individual centres or systems. As one of the priorities laid out by the Scottish Trauma Network project was to develop an electronic data collection system, the TraumaApp was created as a data collection tool for major trauma that could be adopted worldwide. Methods The study was performed as a service evaluation based at the Queen Elizabeth University Hospital Emergency Department. Fifty staff members were recruited in pairs and listened to five paired major trauma standby and handover recordings. Participants were randomised to input data to the TraumaApp and one into the existing paper proforma. The time taken to input data add into was measured, along with time for clarifications and any errors made. Those using the app completed a System Usability Score. Results No statistically significant difference was demonstrated between times taken for data entry for the digital and paper documentation, apart from the Case 5 Handover ( p  < 0.05). Case 1 showed a significantly higher time for clarifications and number of errors with digital data collection ( p  = 0.01 and p  = 1.79E-05 respectively). There were no other differences between data for the app and the proforma. The mean System Usability score for this cohort was 75 out of 100, with a standard deviation of 17 (rounded to nearest integer). Conclusion Digital real-time recording of clinical events using a tool such as the TraumaApp is comparable to completion of paper proforma. The System Usability Score for the TraumaApp was above the internationally validated standard of acceptable usability. There was no evidence of improvement in use over time or familiarity, most likely due to the brevity of the assessments and the refined user interface. This would benefit from further research, exploring data completeness and a potential mixed methods approach to explore training requirements for use of the TraumaApp.