Imperial College London

DrVasaCurcin

Faculty of MedicineSchool of Public Health

Honorary Lecturer
 
 
 
//

Contact

 

+44 (0)20 7594 0716vasa.curcin Website

 
 
//

Location

 

320Reynolds BuildingCharing Cross Campus

//

Summary

 

Publications

Publication Type
Year
to

140 results found

Carr E, Bendayan R, Bean D, Stammers M, Wang W, Zhang H, Searle T, Kraljevic Z, Shek A, Phan HTT, Muruet W, Gupta RK, Shinton AJ, Wyatt M, Shi T, Zhang X, Pickles A, Stahl D, Zakeri R, Noursadeghi M, OGallagher K, Rogers M, Folarin A, Karwath A, Wickstrøm KE, Köhn-Luque A, Slater L, Cardoso VR, Bourdeaux C, Holten AR, Ball S, McWilliams C, Roguski L, Borca F, Batchelor J, Amundsen EK, Wu X, Gkoutos GV, Sun J, Pinto A, Guthrie B, Breen C, Douiri A, Wu H, Curcin V, Teo JT, Shah AM, Dobson RJBet al., 2021, Evaluation and improvement of the National Early Warning Score (NEWS2) for COVID-19: a multi-hospital study, BMC Medicine, Vol: 19

Background: The National Early Warning Score (NEWS2) is currently recommended in the UK for the risk stratification of COVID-19 patients, but little is known about its ability to detect severe cases. We aimed to evaluate NEWS2 for the prediction of severe COVID-19 outcome and identify and validate a set of blood and physiological parameters routinely collected at hospital admission to improve upon the use of NEWS2 alone for medium-term risk stratification. Methods: Training cohorts comprised 1276 patients admitted to King’s College Hospital National Health Service (NHS) Foundation Trust with COVID-19 disease from 1 March to 30 April 2020. External validation cohorts included 6237 patients from five UK NHS Trusts (Guy’s and St Thomas’ Hospitals, University Hospitals Southampton, University Hospitals Bristol and Weston NHS Foundation Trust, University College London Hospitals, University Hospitals Birmingham), one hospital in Norway (Oslo University Hospital), and two hospitals in Wuhan, China (Wuhan Sixth Hospital and Taikang Tongji Hospital). The outcome was severe COVID-19 disease (transfer to intensive care unit (ICU) or death) at 14 days after hospital admission. Age, physiological measures, blood biomarkers, sex, ethnicity, and comorbidities (hypertension, diabetes, cardiovascular, respiratory and kidney diseases) measured at hospital admission were considered in the models. Results: A baseline model of ‘NEWS2 + age’ had poor-to-moderate discrimination for severe COVID-19 infection at 14 days (area under receiver operating characteristic curve (AUC) in training cohort = 0.700, 95% confidence interval (CI) 0.680, 0.722; Brier score = 0.192, 95% CI 0.186, 0.197). A supplemented model adding eight routinely collected blood and physiological parameters (supplemental oxygen flow rate, urea, age, oxygen saturation, C-reactive protein, estimated glomerular filtration rate, neutrophil count, neutrophil/lymphocyte ratio) improved discriminat

Journal article

Chapman M, Mumtaz S, Rasmussen LV, Karwath A, Gkoutos GV, Gao C, Thayer D, Pacheco JA, Parkinson H, Richesson RL, Jefferson E, Denaxas S, Curcin Vet al., 2021, Desiderata for the development of next-generation electronic health record phenotype libraries., Gigascience, Vol: 10

BACKGROUND: High-quality phenotype definitions are desirable to enable the extraction of patient cohorts from large electronic health record repositories and are characterized by properties such as portability, reproducibility, and validity. Phenotype libraries, where definitions are stored, have the potential to contribute significantly to the quality of the definitions they host. In this work, we present a set of desiderata for the design of a next-generation phenotype library that is able to ensure the quality of hosted definitions by combining the functionality currently offered by disparate tooling. METHODS: A group of researchers examined work to date on phenotype models, implementation, and validation, as well as contemporary phenotype libraries developed as a part of their own phenomics communities. Existing phenotype frameworks were also examined. This work was translated and refined by all the authors into a set of best practices. RESULTS: We present 14 library desiderata that promote high-quality phenotype definitions, in the areas of modelling, logging, validation, and sharing and warehousing. CONCLUSIONS: There are a number of choices to be made when constructing phenotype libraries. Our considerations distil the best practices in the field and include pointers towards their further development to support portable, reproducible, and clinically valid phenotype design. The provision of high-quality phenotype definitions enables electronic health record data to be more effectively used in medical domains.

Journal article

Wongkoblap A, Vadillo MA, Curcin V, 2021, Deep Learning With Anaphora Resolution for the Detection of Tweeters With Depression: Algorithm Development and Validation Study., JMIR Ment Health, Vol: 8, ISSN: 2368-7959

BACKGROUND: Mental health problems are widely recognized as a major public health challenge worldwide. This concern highlights the need to develop effective tools for detecting mental health disorders in the population. Social networks are a promising source of data wherein patients publish rich personal information that can be mined to extract valuable psychological cues; however, these data come with their own set of challenges, such as the need to disambiguate between statements about oneself and third parties. Traditionally, natural language processing techniques for social media have looked at text classifiers and user classification models separately, hence presenting a challenge for researchers who want to combine text sentiment and user sentiment analysis. OBJECTIVE: The objective of this study is to develop a predictive model that can detect users with depression from Twitter posts and instantly identify textual content associated with mental health topics. The model can also address the problem of anaphoric resolution and highlight anaphoric interpretations. METHODS: We retrieved the data set from Twitter by using a regular expression or stream of real-time tweets comprising 3682 users, of which 1983 self-declared their depression and 1699 declared no depression. Two multiple instance learning models were developed-one with and one without an anaphoric resolution encoder-to identify users with depression and highlight posts related to the mental health of the author. Several previously published models were applied to our data set, and their performance was compared with that of our models. RESULTS: The maximum accuracy, F1 score, and area under the curve of our anaphoric resolution model were 92%, 92%, and 90%, respectively. The model outperformed alternative predictive models, which ranged from classical machine learning models to deep learning models. CONCLUSIONS: Our model with anaphoric resolution shows promising results when compared with other predi

Journal article

Cabral C, Curtis K, Curcin V, Dominguez J, Prasad V, Schilder A, Turner N, Wilkes S, Taylor J, Gallagher S, Little P, Delaney B, Moore M, Hay AD, Horwood Jet al., 2021, Challenges to implementing electronic trial data collection in primary care: a qualitative study, BMC FAMILY PRACTICE, Vol: 22

Journal article

Ford E, Edelman N, Somers L, Shrewsbury D, Lopez Levy M, van Marwijk H, Curcin V, Porat Tet al., 2021, Barriers and facilitators to the adoption of electronic clinical decision support systems: a qualitative interview study with UK general practitioners, BMC Medical Informatics and Decision Making, Vol: 21, ISSN: 1472-6947

BACKGROUND: Well-established electronic data capture in UK general practice means that algorithms, developed on patient data, can be used for automated clinical decision support systems (CDSSs). These can predict patient risk, help with prescribing safety, improve diagnosis and prompt clinicians to record extra data. However, there is persistent evidence of low uptake of CDSSs in the clinic. We interviewed UK General Practitioners (GPs) to understand what features of CDSSs, and the contexts of their use, facilitate or present barriers to their use. METHODS: We interviewed 11 practicing GPs in London and South England using a semi-structured interview schedule and discussed a hypothetical CDSS that could detect early signs of dementia. We applied thematic analysis to the anonymised interview transcripts. RESULTS: We identified three overarching themes: trust in individual CDSSs; usability of individual CDSSs; and usability of CDSSs in the broader practice context, to which nine subthemes contributed. Trust was affected by CDSS provenance, perceived threat to autonomy and clear management guidance. Usability was influenced by sensitivity to the patient context, CDSS flexibility, ease of control, and non-intrusiveness. CDSSs were more likely to be used by GPs if they did not contribute to alert proliferation and subsequent fatigue, or if GPs were provided with training in their use. CONCLUSIONS: Building on these findings we make a number of recommendations for CDSS developers to consider when bringing a new CDSS into GP patient records systems. These include co-producing CDSS with GPs to improve fit within clinic workflow and wider practice systems, ensuring a high level of accuracy and a clear clinical pathway, and providing CDSS training for practice staff. These recommendations may reduce the proliferation of unhelpful alerts that can result in important decision-support being ignored.

Journal article

Chapman M, Domínguez J, Fairweather E, Delaney BC, Curcin Vet al., 2021, Using Computable Phenotypes in Point-of-Care Clinical Trial Recruitment., Stud Health Technol Inform, Vol: 281, Pages: 560-564

A key challenge in point-of-care clinical trial recruitment is to autonomously identify eligible patients on presentation. Similarly, the aim of computable phenotyping is to identify those individuals within a population that exhibit a certain condition. This synergy creates an opportunity to leverage phenotypes in identifying eligible patients for clinical trials. To investigate the feasibility of this approach, we use the Transform clinical trial platform and replace its archetype-based eligibility criteria mechanism with a computable phenotype execution microservice. Utilising a phenotype for acute otitis media with discharge (AOMd) created with the Phenoflow platform, we compare the performance of Transform with and without the use of phenotype-based eligibility criteria when recruiting AOMd patients. The parameters of the trial simulated are based on those of the REST clinical trial, conducted in UK primary care.

Journal article

Tapuria A, Kordowicz M, Ashworth M, Ferlie E, Curcin V, Koleva-Kolarova R, Fox-Rushby J, Edwards S, Crilly T, Wolfe Cet al., 2021, IT Evaluation of Foundation Healthcare Group Vanguard Project., Stud Health Technol Inform, Vol: 281, Pages: 625-629

The aim of the Foundation Healthcare Group (FHG) Vanguard model was to develop a sustainable local hospital model between two National Health Service (NHS) Trusts (a London Teaching Hospital Trust and a District General Hospital Trust) that makes best use of scarce resources and can be replicated across the NHS, UK. The aim of this study was to evaluate the provision, use and implementation of the IT infrastructure; based on qualitative interviews and focused mainly on the perspectives of the IT staff and the clinicians' perspectives. In total 24 interview transcripts, along with 'Acute Care Collaboration' questionnaire responses, were analysed using a thematic framework for IT infrastructure, sharing themes across the vascular, paediatric and cardiovascular strands of the FHG programme. Findings indicated that Skype for Business had been an innovative and helpful development widely available to be used between the two Trusts. Clinicians initially reported lack of IT support and infrastructure expected at the outset for a national Vanguard project, but later appreciated that remote access to most clinical applications between the two Trusts became operational. The Local Care Record (LCR), an IT project was perceived to have been delivered successfully in South London. Shared technology reduced patient travelling time by providing locally based shared care. Spreading and scaling-up innovations from the Vanguard sites was the aspiration and challenge for system leaders.

Journal article

Tapuria A, Porat T, Kalra D, Dsouza G, Xiaohui S, Curcin Vet al., 2021, Impact of patient access to their electronic health record: systematic review., Informatics for Health and Social Care, Vol: 46, Pages: 194-204, ISSN: 0959-8316

Patient access to their own electronic health records (EHRs) is likely to become an integral part of healthcare systems worldwide. It has the potential to decrease the healthcare provision costs, improve access to healthcare data, self-care, quality of care, and health and patient-centered outcomes. This systematic literature review is aimed at identifying the impact in terms of benefits and issues that have so far been demonstrated by providing patients access to their own EHRs, via providers' secure patient portals from primary healthcare centers and hospitals. Searches were conducted in PubMed, MEDLINE, CINHAL, and Google scholar. Over 2000 papers were screened and were filtered based on duplicates, then by reading the titles and finally based on their abstracts or full text. In total, 74 papers were retained, analyzed, and summarized. Papers were included if providing patient access to their own EHRs was the primary intervention used in the study and its impact or outcome was evaluated. The search technique used to identify relevant literature for this paper involved input from five experts. While findings from 54 of the 74 papers showed positive outcome or benefits of patient access to their EHRs via patient portals, 10 papers have highlighted concerns, 8 papers have highlighted both and 2 have highlighted absence of negative outcomes. The benefits range from re-assurance, reduced anxiety, positive impact on consultations, better doctor-patient relationship, increased awareness and adherence to medication, and improved patient outcomes (e.g., improving blood pressure and glycemic control in a range of study populations). In addition, patient access to their health information was found to improve self-reported levels of engagement or activation related to self-management, enhanced knowledge, and improve recovery scores, and organizational efficiencies in a tertiary level mental health care facility. However, three studies did not find any statistically signific

Journal article

Chapman M, Fairweather E, Khan A, Curcin Vet al., 2021, COVID-19 Analytics in Jupyter: Intuitive Provenance Integration Using ProvIt, Pages: 256-262, ISSN: 0302-9743

Whilst the need to record and understand the evolution of data, together with the processes and users associated with those changes, is now widely appreciated, the uptake of solutions to these issues remains slow. Data provenance techniques have the potential to provide such an understanding, but their use is often considered a specialist activity, requiring detailed knowledge of standards such as W3C PROV. In this work, we introduce ProvIt, a suite of tools designed to lower the barriers to entry for the use of provenance technology. We demonstrate the utility of ProvIt by using it to add provenance capabilities to the Jupyter IDE, in order to provide insight into the tools used by a group of researchers analysing a COVID-19 dataset.

Conference paper

Fairweather E, Chapman M, Curcin V, 2021, A Delayed Instantiation Approach to Template-Driven Provenance for Electronic Health Record Phenotyping, Pages: 3-19, ISSN: 0302-9743

Provenance templates are an established methodology for the capture of provenance data. Each template defines the provenance of a domain-specific action in abstract form, which may then be instantiated as required by a single call to a given service interface. This approach, whilst simplifying the process of recording provenance for the user, introduces computational and storage demands on the capture process, particularly when used by clients with write-intensive provenance requirements such as other service-based software. To address these issues, we adopt a new approach based upon delayed instantiation and present a revised, two-part paradigm for template-driven provenance, in which we separate capture and query functionality to improve the overall efficiency of the model. A dedicated capture service is first employed to record template service requests in a relational database in the form of a meta-level description detailing the construction of each document. These low-overhead records are then accessed by an independent query service to construct views of concrete provenance documents for specific time frames as and when required by the user. These views may subsequently be analysed using query templates, a new technique defined here whereby templates can also be used to search for any matching subgraphs within a document and return the respective instantiating substitutions. We evaluate the performance gains of our new system in the context of Phenoflow, an electronic health record (EHR) phenotyping platform.

Conference paper

Fairweather E, Wittner R, Chapman M, Holub P, Curcin Vet al., 2021, Non-repudiable Provenance for Clinical Decision Support Systems, Pages: 165-182, ISSN: 0302-9743

Provenance templates are now a recognised methodology for the construction of data provenance records. Each template defines the provenance of a domain-specific action in abstract form, which may then be instantiated as required by a single call to the provenance template service. As data reliability and trustworthiness becomes a critical issue in an increasing number of domains, there is a corresponding need to ensure that the provenance of that data is non-repudiable. In this paper we contribute two new, complementary modules to our template model and implementation to produce non-repudiable data provenance. The first, a module that traces the operation of the provenance template service itself, and records a provenance trace of the construction of an object-level document, at the level of individual service calls. The second, a non-repudiation module that generates evidence for the data recorded about each call, annotates the service trace accordingly, and submits a representation of that evidence to a provider-agnostic notary service. We evaluate the applicability of our approach in the context of a clinical decision support system. We first define a policy to ensure the non-repudiation of evidence with respect to a security threat analysis in order to demonstrate the suitability of our solution. We then select three use cases from within a particular system, Consult, with contrasting data provenance recording requirements and analyse the subsequent performance of our prototype implementation against three different notary providers.

Conference paper

Zakeri R, Bendayan R, Ashworth M, Bean DM, Dodhia H, Durbaba S, O'Gallagher K, Palmer C, Curcin V, Aitken E, Bernal W, Barker RD, Norton S, Gulliford M, Teo JTH, Galloway J, Dobson RJB, Shah AMet al., 2020, A case-control and cohort study to determine the relationship between ethnic background and severe COVID-19, EClinicalMedicine, Vol: 28

Background: People of minority ethnic backgrounds may be disproportionately affected by severe COVID-19. Whether this relates to increased infection risk, more severe disease progression, or worse in-hospital survival is unknown. The contribution of comorbidities or socioeconomic deprivation to ethnic patterning of outcomes is also unclear. Methods: We conducted a case-control and a cohort study in an inner city primary and secondary care setting to examine whether ethnic background affects the risk of hospital admission with severe COVID-19 and/or in-hospital mortality. Inner city adult residents admitted to hospital with confirmed COVID-19 (n = 872 cases) were compared with 3,488 matched controls randomly sampled from a primary healthcare database comprising 344,083 people residing in the same region. For the cohort study, we studied 1827 adults consecutively admitted with COVID-19. The primary exposure variable was self-defined ethnicity. Analyses were adjusted for socio-demographic and clinical variables. Findings: The 872 cases comprised 48.1% Black, 33.7% White, 12.6% Mixed/Other and 5.6% Asian patients. In conditional logistic regression analyses, Black and Mixed/Other ethnicity were associated with higher admission risk than white (OR 3.12 [95% CI 2.63–3.71] and 2.97 [2.30–3.85] respectively). Adjustment for comorbidities and deprivation modestly attenuated the association (OR 2.24 [1.83–2.74] for Black, 2.70 [2.03–3.59] for Mixed/Other). Asian ethnicity was not associated with higher admission risk (adjusted OR 1.01 [0.70–1.46]). In the cohort study of 1827 patients, 455 (28.9%) died over a median (IQR) of 8 (4–16) days. Age and male sex, but not Black (adjusted HR 1.06 [0.82–1.37]) or Mixed/Other ethnicity (adjusted HR 0.72 [0.47–1.10]), were associated with in-hospital mortality. Asian ethnicity was associated with higher in-hospital mortality but with a large confidence interval (adjusted HR 1.71 [1.15&ndas

Journal article

Balatsoukas P, Sassoon I, Chapman M, Kokciyan N, Drake A, Modgil S, Ashworth M, Curcin V, Sklar E, Parsons Set al., 2020, In the wild pilot usability assessment of a connected health system for stroke self management

This paper reports on the findings of a pilot study for the formative "in the wild"assessment of the usability of CONSULT, a research-led connected health system for stroke self-management and prevention. CONSULT integrates data from commercial wellness sensors, electronic health records and clinical guidelines and enables users to monitor their vital signs to support self-monitoring and provision of tailored advice. The CONSULT system includes a dashboard and a chatbot. To assess the usability of our system, six volunteers were recruited to interact with CONSULT over a period of seven days. System logs confirmed that participants interacted with the CONSULT system throughout. CONSULT's ability to integrate data from different sensors was an aspect of this systems that all participants liked and kept them motivated to track their vital signs. The study also revealed several usability issues that designers of this type of systems should consider. Some of the most prevalent issues were: information overload, data misinterpretation, need for more anthropomorphic conversational capabilities for the chatbot; lack of visibility of the data transmission status. This paper concludes with reflections on the importance of these findings when assessing the usability of connected health systems, like CONSULT.

Conference paper

Missier P, Bryans J, Gamble C, Curcin Vet al., 2020, Abstracting PROV provenance graphs: A validity-preserving approach, Future Generation Computer Systems, Vol: 111, Pages: 352-367, ISSN: 0167-739X

Data provenance is a structured form of metadata designed to record the activities and datasets involved in data production, as well as their dependency relationships. The PROV data model, released by the W3C in 2013, defines a schema and constraints that together provide a structural and semantic foundation for provenance. This enables the interoperable exchange of provenance between data producers and consumers. When the provenance content is sensitive and subject to disclosure restrictions, however, a way of hiding parts of the provenance in a principled way before communicating it to certain parties is required. In this paper we present a provenance abstraction operator that achieves this goal. It maps a graphical representation of a PROV document PG1 to a new abstract version PG2, ensuring that (i) PG2 is a valid PROV graph, and (ii) the dependencies that appear in PG2 are justified by those that appear in PG1. These two properties ensure that further abstraction of abstract PROV graphs is possible. A guiding principle of the work is that of minimum damage: the resultant graph is altered as little as possible, while ensuring that the two properties are maintained. The operator developed is implemented as part of a user tool, described in a separate paper, that lets owners of sensitive provenance information control the abstraction by specifying an abstraction policy.

Journal article

Curcin V, 2020, Why does human phenomics matter today?, LEARNING HEALTH SYSTEMS, Vol: 4, ISSN: 2379-6146

Journal article

Sassoon I, Kokciyan N, Chapman M, Sklar E, Curcin V, Modgil S, Parsons Set al., 2020, Implementing argument and explanation schemes in dialogue, Pages: 471-472, ISSN: 0922-6389

Conference paper

Gulliford MC, Charlton J, Winter JR, Sun X, Rezel-Potts E, Bunce C, Fox R, Little P, Hay AD, Moore MV, Ashworth M, SafeAB Study Groupet al., 2020, Probability of sepsis after infection consultations in primary care in the United Kingdom in 2002-2017: Population-based cohort study and decision analytic model., PLoS Med, Vol: 17

BACKGROUND: Efforts to reduce unnecessary antibiotic prescribing have coincided with increasing awareness of sepsis. We aimed to estimate the probability of sepsis following infection consultations in primary care when antibiotics were or were not prescribed. METHODS AND FINDINGS: We conducted a cohort study including all registered patients at 706 general practices in the United Kingdom Clinical Practice Research Datalink, with 66.2 million person-years of follow-up from 2002 to 2017. There were 35,244 first episodes of sepsis (17,886, 51%, female; median age 71 years, interquartile range 57-82 years). Consultations for respiratory tract infection (RTI), skin or urinary tract infection (UTI), and antibiotic prescriptions were exposures. A Bayesian decision tree was used to estimate the probability (95% uncertainty intervals [UIs]) of sepsis following an infection consultation. Age, gender, and frailty were evaluated as association modifiers. The probability of sepsis was lower if an antibiotic was prescribed, but the number of antibiotic prescriptions required to prevent one episode of sepsis (number needed to treat [NNT]) decreased with age. At 0-4 years old, the NNT was 29,773 (95% UI 18,458-71,091) in boys and 27,014 (16,739-65,709) in girls; over 85 years old, NNT was 262 (236-293) in men and 385 (352-421) in women. Frailty was associated with greater risk of sepsis and lower NNT. For severely frail patients aged 55-64 years, the NNT was 247 (156-459) in men and 343 (234-556) in women. At all ages, the probability of sepsis was greatest for UTI, followed by skin infection, followed by RTI. At 65-74 years, the NNT following RTI was 1,257 (1,112-1,434) in men and 2,278 (1,966-2,686) in women; the NNT following skin infection was 503 (398-646) in men and 784 (602-1,051) in women; following UTI, the NNT was 121 (102-145) in men and 284 (241-342) in women. NNT values were generally smaller for the period from 2014 to 2017, when sepsis was diagnosed more frequently.

Journal article

Tapuria A, Kalra D, Curcin V, 2020, Feasibility of Using EN 13606 Clinical Archetypes for Defining Computable Phenotypes., Pages: 228-232

INTRODUCTION: Computable phenotypes are gaining importance as structured and reproducible method of using electronic health data to identify people with certain clinical conditions. A formal standard is not available for defining and formally representing phenotyping algorithms. In this paper, we have tried to build a formal representation of such phenotyping algorithm. METHODS: We built EN 13606 EHR standard for building clinical archetypes to represent the computable phenotyping algorithm for 'diagnosis of cardiac failure'. As part of this work, we created a set of new clinical archetypes for defining 'cardiac failure diagnosis'. The EN13606 editor called Object Dictionary Client was used which was in-house developed by University College London. We evaluated the ability of EN 13606 to provide clinical archetypes to define EHR phenotyping algorithms using the predefined desiderata for the purpose [Mo et al]. RESULTS: EN 13606 archetypes could represent phenotype components grouped and nested based on their logical meaning. It was possible to build the EHR phenotyping algorithm with the clinical elements and their interrelationships along with hierarchical structure and temporal criteria. But the specific mathematical calculation and temporal relations involved in the algorithm was difficult to incorporate. These will need to be coded and integrated within the clinical information system. These archetypes can be mapped for comparison with the openEHR models. Binding to external clinical terminology is fully supported. However, it does not satisfy all the desiderata defined by Mo et al. A possible way could be an approach using phenotype ontologies and its architectural representation integrated with ISO interoperability. CONCLUSION: The EN13606 archetypes can be used to define the phenotype algorithm that basically identifies patients by a set of clinical characteristics in their records. Phenotype representations defined in EN 13606 do not satisfy all the desidera

Conference paper

Curtis K, Moore M, Cabral C, Curcin V, Horwood J, Morris R, Prasad V, Schilder A, Turner N, Wilkes S, Hay AD, Taylor Jet al., 2020, A multi-centre, pragmatic, three-arm, individually randomised, non-inferiority, open trial to compare immediate orally administered, immediate topically administered or delayed orally administered antibiotics for acute otitis media with discharge in children: The Runny Ear Study (REST): study protocol., Trials, Vol: 21

BACKGROUND: Acute otitis media (AOM) is a common painful infection in children, with around 2.8 million cases presenting to primary care in England and Wales annually. Nearly all children who present to their general practitioner (GP) with AOM or AOM with discharge (AOMd) are treated with orally administered antibiotics. These can cause side effects; contribute to the growing problem of antimicrobial resistance, and more rarely, allergic reactions. Alternative treatments, such as an antibiotic eardrops, or 'delayed' orally administered antibiotics, could be at least as effective and safe as immediate orally administered antibiotics for children with AOMd. METHODS/DESIGN: REST is a pragmatic, three-arm, individually randomised, non-inferiority trial being conducted in 175 GP practices across the United Kingdom (UK). The study aims to recruit 399 children aged (≥ 12 months and < 16 years) presenting to their GP with AOMd. Children will be randomised to one of three arms: immediate ciprofloxacin 0.3% eardrops; delayed orally administered amoxicillin (clarithromycin if penicillin allergic) or immediate orally administered amoxicillin (clarithromycin). Recruitment, including eligibility screening, randomisation and data collection, are conducted using the innovative, TRANSFoRm electronic trial management platform. Integrated within the primary care electronic medical records it provides automatic eligibility checking, part-filling of e-CRFs, study workflow management and routine NHS follow-up data collection. The primary outcome is time to resolution of all significant symptoms and will be collected by the parent using a Symptom Recovery Questionnaire (SRQ). Secondary outcomes, including cost-effectiveness, duration of moderately bad or worse symptoms and repeat AOMd episodes, will be collected at day-14 and at 3 months. DISCUSSION: It is unclear whether prescribing orally administered antibiotics to children with AOMd resu

Journal article

Wang W, Kiik M, Peek N, Curcin V, Marshall IJ, Rudd AG, Wang Y, Douiri A, Wolfe CD, Bray Bet al., 2020, A systematic review of machine learning models for predicting outcomes of stroke with structured data., PLoS One, Vol: 15

BACKGROUND AND PURPOSE: Machine learning (ML) has attracted much attention with the hope that it could make use of large, routinely collected datasets and deliver accurate personalised prognosis. The aim of this systematic review is to identify and critically appraise the reporting and developing of ML models for predicting outcomes after stroke. METHODS: We searched PubMed and Web of Science from 1990 to March 2019, using previously published search filters for stroke, ML, and prediction models. We focused on structured clinical data, excluding image and text analysis. This review was registered with PROSPERO (CRD42019127154). RESULTS: Eighteen studies were eligible for inclusion. Most studies reported less than half of the terms in the reporting quality checklist. The most frequently predicted stroke outcomes were mortality (7 studies) and functional outcome (5 studies). The most commonly used ML methods were random forests (9 studies), support vector machines (8 studies), decision trees (6 studies), and neural networks (6 studies). The median sample size was 475 (range 70-3184), with a median of 22 predictors (range 4-152) considered. All studies evaluated discrimination with thirteen using area under the ROC curve whilst calibration was assessed in three. Two studies performed external validation. None described the final model sufficiently well to reproduce it. CONCLUSIONS: The use of ML for predicting stroke outcomes is increasing. However, few met basic reporting standards for clinical prediction tools and none made their models available in a way which could be used or evaluated. Major improvements in ML study conduct and reporting are needed before it can meaningfully be considered for practice.

Journal article

Rezel-Potts E, Gulliford MC, Safe AB Study Group, 2020, Sepsis recording in primary care electronic health records, linked hospital episodes and mortality records: Population-based cohort study in England., PLoS One, Vol: 15

BACKGROUND: Sepsis is a growing concern for health systems, but the epidemiology of sepsis is poorly characterised. We evaluated sepsis recording across primary care electronic records, hospital episodes and mortality registrations. METHODS AND FINDINGS: Cohort study including 378 general practices in England from Clinical Practice Research Datalink (CPRD) GOLD database from 2002-2017 with 36,209,676 patient-years of follow-up with linked Hospital Episode Statistics (HES) and Office for National Statistics (ONS) mortality registrations. Incident sepsis episodes were identified for each source. Concurrent records from different sources were identified and age-standardised and age-specific incidence rates compared. Logistic regression analysis evaluated associations of gender, age-group, fifth of deprivation and period of diagnosis with concurrent sepsis recording. There were 20,206 first episodes of sepsis from primary care, 20,278 from HES and 13,972 from ONS. There were 4,117 (20%) first HES sepsis events and 2,438 (17%) mortality records concurrent with incident primary care sepsis records within 30 days. Concurrent HES and primary care records of sepsis within 30 days before or after first diagnosis were higher at younger or older ages and for patients with the most recent period of diagnosis. Those diagnosed during 2007:2011 were less likely to have a concurrent HES record given CPRD compared to those diagnosed during 2012-2017 (odd ratio 0.65, 95% confidence interval 0.60-0.70). At age 85 and older, primary care incidence was 5.22 per 1,000 patient years (95% CI 1.75-11.97) in men and 3.55 (0.87-9.58) in women which increased to 10.09 (4.86-18.51) for men and 7.22 (2.96-14.72) for women after inclusion of all three sources. CONCLUSION: Explicit recording of 'sepsis' is inconsistent across healthcare sectors with a high proportion of non-concurrent records. Incidence estimates are higher when linked data are analysed.

Journal article

Ford E, Curlewis K, Wongkoblap A, Curcin Vet al., 2019, Public Opinions on Using Social Media Content to Identify Users With Depression and Target Mental Health Care Advertising: Mixed Methods Survey, JMIR MENTAL HEALTH, Vol: 6, ISSN: 2368-7959

Journal article

Balatsoukas P, Porat T, Sassoon IK, Essers K, Kokciyan N, Chapman MD, Drake A, Modgil S, Ashworth M, Sklar EI, Parsons SD, Curcin Vet al., 2019, User involvement in the design of a data-driven self-management decision support tool for stroke survivors, 18th IEEE International Conference on Smart Technologies, EUROCON 2019, Publisher: IEEE

Many chronic conditions can be better managed by patients themselves with the use of decision support tools. This becomes even more necessary in the case of multimorbidity (i.e. presence of multiple chronic diseases) or in conditions where several underlying risk factors need to be managed and monitored in order to avoid relapse or the reoccurrence of an event, like in the case of stroke. However, despite the fact that these decision support systems are becoming prevalent, little is known about the best practices in designing for end-users - patients and their carers. The aim of the present paper is to report on the process of involving users to inform the design of a novel data-driven self-management mobile decision support tool for stroke survivors, called CONSULT. User involvement was facilitated through the use of a two-phase participatory design approach. During both phases a total of 44 stakeholders participated, including stroke survivors, carers, healthcare professionals and researchers. The paper documents the findings of the participatory design process, in the form of design recommendations, and describes their implications for user interface design.

Conference paper

Kökciyan N, Chapman M, Balatsoukas P, Sassoon I, Essers K, Ashworth M, Curcin V, Modgil S, Parsons S, Sklar EIet al., 2019, A Collaborative Decision Support Tool for Managing Chronic Conditions., Pages: 644-648

This paper describes work to assess the feasibility of using a decision support tool to help patients with chronic conditions, specifically stroke, manage their condition in collaboration with their carers and the health care professionals who are looking after them. The system contains several novel elements: the integration of data from commercial wellness sensors, electronic health records and clinical guidelines; the use of computational argumentation to track the source of data and to resolve conflicts and make recommendations; and argumentation-based dialogue to support interaction with patients. The proposed approach is implemented as an application that can run on smart devices (e.g. tablets). The users have personalised dashboards where they can visualise their health data and interact with a conversational chatbot that provides further explanations about their overall well-being.

Conference paper

Porat T, Marshall I, Sadler E, Vadillo MA, McKevitt C, Wolfe CDA, Curcin Vet al., 2019, Collaborative design of a decision aid for stroke survivors with multimorbidity: a qualitative study in the UK engaging key stakeholders, BMJ Open, Vol: 9, ISSN: 2044-6055

Objectives: Effective secondary stroke prevention strategies are sub-optimally used. Noveldevelopment of interventions to enable healthcare professionals and stroke survivors to manage riskfactors for stroke recurrence are required. We sought to engage key stakeholders in the design andevaluation of an intervention informed by a Learning Health System approach, to improve risk factormanagement and secondary prevention for stroke survivors with multimorbidity.Design: Qualitative, including focus groups, semi-structured interviews and usability evaluations. Datawas audio-recorded, transcribed and coded thematically.Participants: Stroke survivors, carers, health and social care professionals, commissioners, policymakers and researchers.Setting: Stroke survivors were recruited from the South London Stroke Register; health and social careprofessionals through South London general practices and King’s College London (KCL) networks;carers, commissioners, policy-makers and researchers through KCL networks.Results: 53 stakeholders in total participated in focus groups, interviews and usability evaluations.Thirty-seven participated in focus groups and interviews, including stroke survivors and carers (N=11),health and social care professionals (N=16), commissioners and policy-makers (N=6) and researchers(N=4). Sixteen participated in usability evaluations, including stroke survivors (N=8) and generalpractitioners (GPs; N=8). Eight themes informed the collaborative design of DOTT (Deciding onTreatments Together), a decision aid integrated with the electronic health record system, to be usedin primary care during clinical consultations between the healthcare professional and stroke survivor.DOTT aims to facilitate shared decision making on personalised treatments leading to improvedtreatment adherence and risk control. DOTT was found acceptable and usable among stroke survivorsand GPs during a series of evaluations.Conclusions: Adopting a user-centred data-driven design a

Journal article

Ford E, Boyd A, Bowles JKF, Havard A, Aldridge RW, Curcin V, Greiver M, Harron K, Katikireddi V, Rodgers SE, Sperrin Met al., 2019, Our data, our society, our health: A vision for inclusive and transparent health data science in the United Kingdom and beyond, LEARNING HEALTH SYSTEMS, Vol: 3, ISSN: 2379-6146

Journal article

Wongkoblap A, Vadillo MA, Curcin V, 2019, Predicting Social Network Users with Depression from Simulated Temporal Data

Mental health issues are widely accepted as one of the most prominent health challenges in the world, with over 300 million people currently suffering from depression alone. With massive volumes of user-generated data on social networking platforms, researchers are increasingly using machine learning to determine whether this content can be used to detect mental health problems in users. This study aims to investigate whether training a predictive model with multiple instance learning (MIL) via Long Short-Term Memory (LSTM) and gated recurrent unit (GRU) can improve the performance of a predictive model to detect social network users with depression. The power of MIL is to learn from user-level labels to identify post-level labels. By combining every possibility of posts label category, it can generate temporal posting profiles which can then be used to classify users with depression. This study highlights that training a MIL model via LSTM and GRU can improve the accuracy of a MIL model trained with convolutional neural networks.

Conference paper

Chapman M, Curcin V, 2019, A Microservice Architecture for the Design of Computer-Interpretable Guideline Processing Tools

Several tools exist that are designed to process computer interpretable guidelines (CIGs), each with a distinct purpose, such as detecting interactions or patient personalisation. While it is desirable to use these tools as part of larger decision support systems (DSSs) doing so is often not straightforward, as their design does not often support external interoperability or account for the fact that other CIG tools may be running in parallel, a situation that will become increasingly more prevalent with the increased adoption of CIGs in different parts of the health system. This results in an integration overhead, system redundancy and a lack of flexibility in how these tools can be combined. To address these issues, we define a blueprint architecture to be used in the design of guideline processing tools, based on the conceptualisation of key components as RESTful microservices. In addition, we define the types of data endpoints that each component should expose, for both the communication between internal components and communication with external components that exist as a part of a DSS. To demonstrate the utility of our architecture, we show how an example guideline processing tool can be restructured according to these principles, in order to enable it to be flexibly integrated into the DSS used in the CONSULT project.

Conference paper

Chapman M, Balatsoukas P, Ashworth M, Curcin V, Kökciyan N, Essers K, Sassoon I, Modgil S, Parsons S, Sklar EIet al., 2019, Computational argumentation-based clinical decision support: Demonstration, Pages: 2345-2347, ISSN: 1548-8403

This demonstration highlights the design of the Consult system, a modular decision-support system (DSS) intended to help patients suffering from chronic conditions self-manage their treatments. The system takes input from multiple sources, including commercial wellness sensors and a patient's electronic health record, to inform a computational argumentation engine that constructs weighted opinions using these inputs and knowledge about their sources, and uses an interaction agent driven by argumentation-based dialogue to respond to user queries.

Conference paper

This data is extracted from the Web of Science and reproduced under a licence from Thomson Reuters. You may not copy or re-distribute this data in whole or in part without the written consent of the Science business of Thomson Reuters.

Request URL: http://wlsprd.imperial.ac.uk:80/respub/WEB-INF/jsp/search-html.jsp Request URI: /respub/WEB-INF/jsp/search-html.jsp Query String: respub-action=search.html&id=00321770&limit=30&person=true