Urgent and Emergency Care Publications

Research papers published by our team

2022

Authors - J. Magallanes, T. Stone, P. D. Morris, S. Mason, S. Wood and M. -C. Villa-Uriol

https://ieeexplore.ieee.org/document/9557226

Building a visual overview of temporal event sequences with an optimal level-of-detail (i.e. simplified but informative) is an ongoing challenge - expecting the user to zoom into every important aspect of the overview can lead to missing insights. We propose a technique to build a multilevel overview of event sequences, whose granularity can be transformed across sequence clusters (vertical level-of-detail) or longitudinally (horizontal level-of-detail), using hierarchical aggregation and a novel cluster data representation Align-Score-Simplify. By default, the overview shows an optimal number of sequence clusters obtained through the average silhouette width metric - then users are able to explore alternative optimal sequence clusterings. The vertical level-of-detail of the overview changes along with the number of clusters, whilst the horizontal level-of-detail refers to the level of summarization applied to each cluster representation. The proposed technique has been implemented into a visualization system called Sequence Cluster Explorer (Sequen-C) that allows multilevel and detail-on-demand exploration through three coordinated views, and the inspection of data attributes at cluster, unique sequence, and individual sequence level. We present two case studies using real-world datasets in the healthcare domain: CUREd and MIMIC-III; which demonstrate how the technique can aid users to obtain a summary of common and deviating pathways, and explore data attributes for selected patterns.

DOI: https://doi.org/10.1109/TVCG.2021.3114868

2021

Authors - Christopher Burton, Tony Stone,  Phillip Oliver, Jon M Dickson, Jen Lewis,  Suzanne Mason

https://emj.bmj.com/content/39/1/3

Abstract

Objective

Frequent attendance at the ED is a worldwide problem. We hypothesised that frequent attendance could be understood as a feature of a complex system comprising patients, healthcare and society. Complex systems have characteristic statistical properties, with stable patterns at the level of the system emerging from unstable patterns at the level of individuals who make up the system.

Methods

Analysis of a linked dataset of routinely collected health records from all 13 hospital trusts providing ED care in the Yorkshire and Humber region of the UK (population 5.5 million). We analysed the distribution of attendances per person in each of 3 years and measured the transition of individual patients between frequent, infrequent and non-attendance. We fitted data to power law distributions typically seen in complex systems using maximum likelihood estimation.

Results

The data included 3.6 million attendances at EDs in 13 hospital trusts. 29/39 (74.3%) analyses showed a statistical fit to a power law; 2 (5.1%) fitted an alternative distribution. All trusts’ data fitted a power law in at least 1 year. Differences over time and between hospital trusts were small and partly explained by demographics. In contrast, individual patients’ frequent attendance was unstable between years.

Conclusions

ED attendance patterns are stable at the level of the system, but unstable at the level of individual frequent attenders. Attendances follow a power law distribution typical of complex systems. Interventions to address ED frequent attendance need to consider the whole system and not just the individual frequent attenders.

Authors - Simpson R, O'Keeffe C, Stone T, Jacques R, Mason S

https://emj.bmj.com/content/39/1/17

Abstract

Introduction

A significant proportion of ED attendances in children may be non-urgent attendances (NUAs), which could be better managed elsewhere. This study aimed to quantify NUAs and urgent attendances (UAs) in children to ED and determine which children present in this way and when.

Methods

Dataset extracted from the CUREd research database containing linked data on the provision of care in Yorkshire and Humber. Analysis focused on children’s ED attendances (April 2014–March 2017). Summary statistics and odds ratios (OR) comparing NUAs and UAs were examined by: age, mode and time of arrival and deprivation alongside comparing summary statistics for waiting, treatment and total department times.

Results

NUAs were more likely in younger children: OR for NUA in children aged 1–4 years, 0.82 (95% CI: 0.80 to 0.83), age 15 years, 0.39 (95% CI: 0.38 to 0.40), when compared with those under 1 year. NUAs were more likely to arrive out of hours (OOHs) compared with in hours: OR 1.19 (95% CI 1.18 to 1.20), and OOHs arrivals were less common in older children compared with those under 1 year: age 1–4 years, 0.87 (95% CI: 0.84 to 0.89) age 15 years, 0.66 (95% CI: 0.63 to 0.69). NUAs also spent less total time in the ED, with a median (IQR) of 98 min (60–147) compared with 127 min (80–185) for UAs.

Conclusion

A substantial proportion of ED attendances in children are NUAs. Our data suggest there are particular groups of children for whom targeted interventions would be most beneficial. Children under 5 years would be such a group, particularly in providing accessible, timely care outside of usual community care opening hours.

Authors - J. Magallanes, T. Stone, P. D. Morris, S. Mason, S. Wood and M. -C. Villa-Uriol

https://pubmed.ncbi.nlm.nih.gov/34596549/

Abstract

Building a visual overview of temporal event sequences with an optimal level-of-detail (i.e. simplified but informative) is an ongoing challenge - expecting the user to zoom into every important aspect of the overview can lead to missing insights. We propose a technique to build a multilevel overview of event sequences, whose granularity can be transformed across sequence clusters (vertical level-of-detail) or longitudinally (horizontal level-of-detail), using hierarchical aggregation and a novel cluster data representation Align-Score-Simplify. By default, the overview shows an optimal number of sequence clusters obtained through the average silhouette width metric - then users are able to explore alternative optimal sequence clusterings. The vertical level-of-detail of the overview changes along with the number of clusters, whilst the horizontal level-of-detail refers to the level of summarization applied to each cluster representation. The proposed technique has been implemented into a visualization system called Sequence Cluster Explorer (Sequen-C) that allows multilevel and detail-on-demand exploration through three coordinated views, and the inspection of data attributes at cluster, unique sequence, and individual sequence level. We present two case studies using real-world datasets in the healthcare domain: CUREd and MIMIC-III; which demonstrate how the technique can aid users to obtain a summary of common and deviating pathways, and explore data attributes for selected patterns.

Authors - Christopher Burton, Tony Stone,  Phillip Oliver, Jon M Dickson, Jen Lewis, Suzanne Mason

https://emj.bmj.com/content/39/1/3

Abstract

Objective

Frequent attendance at the ED is a worldwide problem. We hypothesised that frequent attendance could be understood as a feature of a complex system comprising patients, healthcare and society. Complex systems have characteristic statistical properties, with stable patterns at the level of the system emerging from unstable patterns at the level of individuals who make up the system.

Methods

Analysis of a linked dataset of routinely collected health records from all 13 hospital trusts providing ED care in the Yorkshire and Humber region of the UK (population 5.5 million). We analysed the distribution of attendances per person in each of 3 years and measured the transition of individual patients between frequent, infrequent and non-attendance. We fitted data to power law distributions typically seen in complex systems using maximum likelihood estimation.

Results

The data included 3.6 million attendances at EDs in 13 hospital trusts. 29/39 (74.3%) analyses showed a statistical fit to a power law; 2 (5.1%) fitted an alternative distribution. All trusts’ data fitted a power law in at least 1 year. Differences over time and between hospital trusts were small and partly explained by demographics. In contrast, individual patients’ frequent attendance was unstable between years.

Conclusions

ED attendance patterns are stable at the level of the system, but unstable at the level of individual frequent attenders. Attendances follow a power law distribution typical of complex systems. Interventions to address ED frequent attendance need to consider the whole system and not just the individual frequent attenders.

Authors - J. Magallanes, T. Stone, P. D. Morris, S. Mason, S. Wood and M. -C. Villa-Uriol

https://www.thelancet.com/journals/lanhl/article/PIIS2666-7568(21)00004-0/fulltext

Abstract

Building a visual overview of temporal event sequences with an optimal level-of-detail (i.e. simplified but informative) is an ongoing challenge - expecting the user to zoom into every important aspect of the overview can lead to missing insights. We propose a technique to build a multilevel overview of event sequences, whose granularity can be transformed across sequence clusters (vertical level-of-detail) or longitudinally (horizontal level-of-detail), using hierarchical aggregation and a novel cluster data representation Align-Score-Simplify. By default, the overview shows an optimal number of sequence clusters obtained through the average silhouette width metric - then users are able to explore alternative optimal sequence clusterings. The vertical level-of-detail of the overview changes along with the number of clusters, whilst the horizontal level-of-detail refers to the level of summarization applied to each cluster representation. The proposed technique has been implemented into a visualization system called Sequence Cluster Explorer (Sequen-C) that allows multilevel and detail-on-demand exploration through three coordinated views, and the inspection of data attributes at cluster, unique sequence, and individual sequence level. We present two case studies using real-world datasets in the healthcare domain: CUREd and MIMIC-III; which demonstrate how the technique can aid users to obtain a summary of common and deviating pathways, and explore data attributes for selected patterns.

Authors - Michael Poon, Suzanne Mason & 4C

https://www.thelancet.com/journals/lanhl/article/PIIS2666-7568(21)00004-0/fulltext

Abstract

Objective

To monitor hospital activity for presentation, diagnosis and treatment of cardiovascular diseases during the COVID-19) pandemic to inform on indirect effects.

Methods

Retrospective serial cross-sectional study in nine UK hospitals using hospital activity data from 28 October 2019 (pre-COVID-19) to 10 May 2020 (pre-easing of lockdown) and for the same weeks during 2018–2019. We analysed aggregate data for selected cardiovascular diseases before and during the epidemic. We produced an online visualisation tool to enable near real-time monitoring of trends.

Results

Across nine hospitals, total admissions and emergency department (ED) attendances decreased after lockdown (23 March 2020) by 57.9% (57.1%–58.6%) and 52.9% (52.2%–53.5%), respectively, compared with the previous year. Activity for cardiac, cerebrovascular and other vascular conditions started to decline 1–2 weeks before lockdown and fell by 31%–88% after lockdown, with the greatest reductions observed for coronary artery bypass grafts, carotid endarterectomy, aortic aneurysm repair and peripheral arterial disease procedures. Compared with before the first UK COVID-19 (31 January 2020), activity declined across diseases and specialties between the first case and lockdown (total ED attendances relative reduction (RR) 0.94, 0.93–0.95; total hospital admissions RR 0.96, 0.95–0.97) and after lockdown (attendances RR 0.63, 0.62–0.64; admissions RR 0.59, 0.57–0.60). There was limited recovery towards usual levels of some activities from mid-April 2020.

Conclusions

Substantial reductions in total and cardiovascular activities are likely to contribute to a major burden of indirect effects of the pandemic, suggesting they should be monitored and mitigated urgently.

2020

Authors - Michael Poon, Suzanne Mason & 4C

https://heart.bmj.com/content/106/24/1890.abstract

Abstract

Objective

To monitor hospital activity for presentation, diagnosis and treatment of cardiovascular diseases during the COVID-19) pandemic to inform on indirect effects.

Methods

Retrospective serial cross-sectional study in nine UK hospitals using hospital activity data from 28 October 2019 (pre-COVID-19) to 10 May 2020 (pre-easing of lockdown) and for the same weeks during 2018–2019. We analysed aggregate data for selected cardiovascular diseases before and during the epidemic. We produced an online visualisation tool to enable near real-time monitoring of trends.

Results

Across nine hospitals, total admissions and emergency department (ED) attendances decreased after lockdown (23 March 2020) by 57.9% (57.1%–58.6%) and 52.9% (52.2%–53.5%), respectively, compared with the previous year. Activity for cardiac, cerebrovascular and other vascular conditions started to decline 1–2 weeks before lockdown and fell by 31%–88% after lockdown, with the greatest reductions observed for coronary artery bypass grafts, carotid endarterectomy, aortic aneurysm repair and peripheral arterial disease procedures. Compared with before the first UK COVID-19 (31 January 2020), activity declined across diseases and specialties between the first case and lockdown (total ED attendances relative reduction (RR) 0.94, 0.93–0.95; total hospital admissions RR 0.96, 0.95–0.97) and after lockdown (attendances RR 0.63, 0.62–0.64; admissions RR 0.59, 0.57–0.60). There was limited recovery towards usual levels of some activities from mid-April 2020.

Conclusions

Substantial reductions in total and cardiovascular activities are likely to contribute to a major burden of indirect effects of the pandemic, suggesting they should be monitored and mitigated urgently.

Authors - Miles, J, Turner, J, Jacques, R, Mason S

https://diagnprognres.biomedcentral.com/articles/10.1186/s41512-020-00084-1

Abstract

Background

The primary objective of this review is to assess the accuracy of machine learning methods in their application of triaging the acuity of patients presenting in the Emergency Care System (ECS). The population are patients that have contacted the ambulance service or turned up at the Emergency Department. The index test is a machine-learning algorithm that aims to stratify the acuity of incoming patients at initial triage. This is in comparison to either an existing decision support tool, clinical opinion or in the absence of these, no comparator. The outcome of this review is the calibration, discrimination and classification statistics.

Methods

Only derivation studies (with or without internal validation) were included. MEDLINE, CINAHL, PubMed and the grey literature were searched on the 14th December 2019. Risk of bias was assessed using the PROBAST tool and data was extracted using the CHARMS checklist. Discrimination (C-statistic) was a commonly reported model performance measure and therefore these statistics were represented as a range within each machine learning method. The majority of studies had poorly reported outcomes and thus a narrative synthesis of results was performed.

Results

There was a total of 92 models (from 25 studies) included in the review. There were two main triage outcomes: hospitalisation (56 models), and critical care need (25 models). For hospitalisation, neural networks and tree-based methods both had a median C-statistic of 0.81 (IQR 0.80-0.84, 0.79-0.82). Logistic regression had a median C-statistic of 0.80 (0.74-0.83). For critical care need, neural networks had a median C-statistic of 0.89 (0.86-0.91), tree based 0.85 (0.84-0.88), and logistic regression 0.83 (0.79-0.84).

Conclusions

Machine-learning methods appear accurate in triaging undifferentiated patients entering the Emergency Care System. There was no clear benefit of using one technique over another; however, models derived by logistic regression were more transparent in reporting model performance. Future studies should adhere to reporting guidelines and use these at the protocol design stage.

Authors - Hughes-Gooding T, Dickson JM, O'Keeffe C

https://emj.bmj.com/content/emermed/early/2020/06/15/emermed-2019-208820.full.pdf

ABSTRACT

Introduction

The urgent and emergency care (UEC) system is struggling with increased demand, some of which is clinically unnecessary. Patients suffering suspected seizures commonly present to EDs, but most seizures are self-limiting and have low risk of short-term adverse outcomes. We aimed to investigate the flow of suspected seizure patients through the UEC system using data linkage to facilitate the development of new models of care.

Methods

We used a two-stage process of deterministic linking to perform a cross-sectional analysis of data from adults in a large region in England (population 5.4million) during 2014. The core dataset comprised a total of 739 436 ambulance emergency incidents, 1 033 778 ED attendances and 362 358 admissions.

Results

A high proportion of cases were successfully linked (86.9% ED-inpatient, 77.7% ED-ambulance). Suspected seizures represented 2.8% of all ambulance service incidents. 61.7% of these incidents led to dispatch of a rapid-response ambulance (8min) and 72.1% were conveyed to hospital. 37 patients died before being conveyed to hospital and 24 died in the ED (total 61; 0.3%). The inpatient death rate was 0.4%. Suspected seizures represented 0.71% of ED attendances, 89.8% of these arrived by emergency ambulance, 45.4% were admitted and 44.5% of these admissions lasted under 48 hours.

Conclusions

This study confirms previously published data from smaller unlinked datasets, validating the linkage method, and provides new data for suspected seizures. There are significant barriers to realising the full potential of data linkage. Collaborative action is needed to create facilitative governance frameworks and improve data quality and analytical capacity.

Authors - Ablard S, Kuczawski M, Sampson FC

https://emj.bmj.com/content/37/4/200

Abstract

Background Policies aimed at diverting care from EDs to alternative services have not been successful in reducing ED attendances and have contributed to confusion for service users when making care-seeking decisions. It is important that service users are at the heart of decision making to ensure new services meet the needs of those who will be accessing them. In this study, service users were encouraged to think freely about the desirable qualities of an ideal urgent and emergency care (UEC) system.

Methods From September to February 2019, an open inductive methodology was used to conduct focus groups with service users who had used UK UEC services within the previous year. Service users that had contact with NHS111, ambulance service, General Practice out-of-hours, minor injuries unit, walk-in centre or ED were purposively sampled and stratified into the following groups: (1) 18–45 years; (2)≥75 years; (3) adults with young children; (4) adults with long-term conditions. Focus groups were structured around experiences of accessing UEC services and perspectives of an ‘ideal’ UEC system.

Results 30 service users took part in the study, across four focus groups. The ideal UEC system centred around three themes: a simplified UEC system (easier to understand and a single-point of access); more ‘joined-up’ UEC services and better communication between health staff and patients.

Conclusion Desirable qualities of an ideal UEC system from a service user perspective related to simplifying access for example, through a single point of access system where health professionals decide the appropriate service required and improving continuity of care through better integration of UEC services. Service users value reassurance and communication from health professionals about care pathways and care choices, and this helps service users feel more in control of their healthcare journey.

Authors - Credé S ,Mason S , Such E, Jacques R.

https://www.medrxiv.org/content/10.1101/2020.10.08.20209296v1

Abstract

Background

Globally, international migration is increasing. Population growth, along with other demographic changes, may be expected to put new pressures on healthcare systems. Some studies across Europe suggest that emergency departments (EDs) are used more, and differently, by migrants compared to non-migrant populations, which may be a result of unfamiliarity with the healthcare systems and difficulties accessing primary healthcare. However, little evidence exists to understand how migrant parents, who are typically young and of childbearing age, utilise EDs for their children. This study aimed to examine the association between paediatric ED utilisation in the first 5 years of life and maternal migration status in the Born in Bradford (BiB) cohort study.

Methods and findings

We analysed linked data from the BiB study—an ongoing, multi-ethnic prospective birth cohort study in Bradford. Bradford is a large, ethnically diverse city in the north of England. In 2017, more than a third of births in Bradford were to mothers who were born outside the UK. Between March 2007 and December 2010, pregnant women were recruited to BiB during routine antenatal care, and the children born to these mothers have been, and continue to be, followed over time to assess how social, genetic, environmental, and behavioural factors impact on health from childhood to adulthood. Data analysed in this study included baseline questionnaire data from BiB mothers, and Bradford Royal Infirmary ED episode data for their children. Main outcomes were likelihood of paediatric ED use (no visits versus at least 1 ED visit in the first 5 years of life) and ED utilisation rates (number and frequency of ED visits) for children who have accessed the ED. The main explanatory variable was mother’s migrant status (foreign-born versus UK/Irish-born). Multivariable analyses (logistic and zero-truncated negative binomial regression) were conducted adjusting for socio-demographic and socio-economic factors. The final dataset included 10,168 children born between April 2007 and June 2011, of whom 35.6% were born to migrant mothers. Foreign-born mothers originated from South Asia (28.6%), Europe/Central Asia (3.2%), Africa (2.1%), East Asia/Pacific (1.1%), and the Middle East (0.6%). At recruitment the mothers ranged in age from 15 to 49 years old. Overall, 3,104 (30.5%) children had at least 1 ED visit in the first 5 years of life, with the highest proportion of visits being in the first year of life (36.7%). The proportion of children who visited the ED at least once was lower for children of migrant mothers as compared to children of non-migrant mothers (29.4% versus 31.2%). Children of migrant mothers were found to be less likely to visit the ED (odds ratio 0.88 [95% CI 0.80 to 0.97], p = 0.012). However, among children who visited the ED, the utilisation rate was significantly higher for children of migrant mothers (incidence rate ratio [IRR] 1.19 [95% CI 1.01 to 1.40], p = 0.040). Utilisation rates were higher for children born to mothers from Europe (IRR 1.71 [95% CI 1.07 to 2.71], p = 0.024) and established migrants (≥5 years living in UK) (IRR 1.24 [95% CI 1.02 to 1.51], p = 0.032) compared to UK/Irish-born mothers. Important limitations include being unable to measure children’s underlying health status and the urgency of ED attendance, as well as the analysis being limited by missing data.

Conclusions

In this study we observed that there is no higher likelihood of first paediatric ED attendance in the first 5 years of life for children in the BiB cohort for migrant mothers. However, among ED users, children of migrant mothers attend the service more frequently than children of UK/Irish-born mothers. Our findings show that patterns of ED utilisation differ by mother’s region of origin and time since arrival in the UK.

Authors - Christopher Burton, Tony Stone,  Phillip Oliver, Jon M Dickson, Jen Lewis,  Suzanne Mason

https://www.medrxiv.org/content/10.1101/2020.10.08.20209296v1

ABSTRACT

Objective High use of the ED is a worldwide problem. We hypothesised that high use of the ED could be understood as a feature of a complex system comprising patients, healthcare and society. Complex systems have characteristic statistical properties, with stable patterns at the level of the system emerging from unstable patterns at the level of individuals who make up the system.

Methods

Analysis of a linked dataset of routinely collected health records from all 13 hospital trusts providing ED care in the Yorkshire and Humber region of the UK (population 5.5 million). We analysed the distribution of attendances per person in each of three years and measured the transition of individual patients between high, low and non-attendance. We fitted data to power law distributions typically seen in complex systems using maximum likelihood estimation.

Results

The data included 3.6 million attendances at EDs in 13 hospital trusts. 29/39 (74.3%) analyses showed a statistical fit to a power law; 2 (5.1%) fitted an alternative distribution.. All trusts’ data fitted a power law in at least one year. Differences over time and between hospital trusts were small and partly explained by demographics. In contrast, individual patients’ high use was unstable between years.

Conclusions

ED attendance patterns are stable at the level of the system, but unstable at the level of individual high users. Attendances follow a power law distribution typical of complex systems. Interventions to address ED high use need to consider the whole system and not just the individual high users.