Linked Data Sources Research Articles

Comparative risk of cerebral venous sinus thrombosis (CVST) following COVID-19 vaccination or infection: A national cohort study using linked electronic health records

ABSTRACT To inform the public and policy makers, we investigated and compared the risk of cerebral venous sinus thrombosis (CVST) after SARS-Cov-2 vaccination or infection using a national cohort of 2,643,699 individuals aged 17 y and above, alive, and resident in Wales on 1 January 2020 followed up through multiple linked data sources until 28 March 2021. Exposures were first dose of Oxford-ChAdOx1 or Pfizer-BioNTech vaccine or polymerase chain reaction (PCR)-confirmed SARS-Cov-2 infection. The outcome was an incident record of CVST. Hazard ratios (HR) were calculated using multivariable Cox regression, adjusted for confounders. HR from SARS-Cov-2 infection was compared with that for SARS-Cov-2 vaccination. We identified 910,556 (34.4%) records of first SARS-Cov-2 vaccination and 165,862 (6.3%) of SARS-Cov-2 infection. A total of 1,372 CVST events were recorded during the study period, of which 52 (3.8%) and 48 (3.5%) occurred within 28 d after vaccination and infection, respectively. We observed slight non-significant risk of CVST within 28 d of vaccination [aHR: 1.34, 95% CI: 0.95-1.90], which remained after stratifying by vaccine [BNT162b2, aHR: 1.18 (95% CI: 0.63-2.21); ChAdOx1, aHR: 1.40 (95% CI: 0.95-2.05)]. Three times the number of CVST events is observed within 28 d of a positive SARS-Cov-2 test [aHR: 3.02 (95% CI: 2.17-4.21)]. The risk of CVST following SARS-Cov-2 infection is 2.3 times that following SARS-Cov-2 vaccine. This is important information both for those designing COVID-19 vaccination programs and for individuals making their own informed decisions on the risk-benefit of vaccination. This record-linkage approach will be useful in monitoring the safety of future vaccine programs.

Use of linked registry claims data for long term surveillance of devices after endovascular abdominal aortic aneurysm repair: observational surveillance study

ObjectiveTo evaluate long term outcomes (reintervention and late rupture of abdominal aortic aneurysm) of aortic endografts in real world practice using linked registry claims data.DesignObservational surveillance study.Setting282 centers in the...

A hybrid recommender system based on description/dialetheic logic and linked data

AbstractIn this research, we developed a hybrid recommender system based on description/dialetheic logic, which provides an innovative architecture for the recommendation based on ambiguous reasoning. In addition, our approach allows enriching the knowledge during the reasoning using the linked data paradigm. Thus, the architecture allows the integration of linked data, in order to be exploited by the recommender. On the other hand, the reasoning mechanisms of dialetheic logic allow handling situations of contradiction or inconsistency, to determine the information to offer to users. According to the reviewed literature, our proposal is the first that implements a dialetheic engine in a Recommender System. In general, there are a lot of works about the utilization of linked open data (LOD) to improve the Recommender Systems, for example, to enrich the information to recommend or to solve the cold start problem. However, there are no proposals to deal with the problems of ambiguity, incoherence or incompleteness of the information in these environments extended with LOD. In this way, this paper proposes a hybrid reasoning engine that uses the linked data paradigm for the knowledge extraction, and dialetheic logic for the management of inconsistency/ambiguity information, in order to obtain recommendations. Particularly, the information extracted with Linked Data is processed with the Dialetheic Logic reasoner to solve ambiguous cases. Thus, each recommendation is enriched with related content extracted from the linked data sources, which has been disambiguated. The results are very promising since our hybrid reasoning mechanism allows obtaining more precise recommendations, considering the different classical states of ambiguity in dialetheic logic (contingent statements about the future, failure of a presupposition, vagueness, counterfactual reasoning), according to various metrics of quality used to evaluate the recommendations achieved.

Stroke in Older Adults Living in Care Homes: Results From a National Data Linkage Study in Wales

To determine the proportion of older people moving to care homes with a recent stroke, incidence of stroke after moving to a care home, mortality following stroke, and secondary stroke prevention management in older care home residents. Retrospective cohort study using population-scale individual-level linked data sources between 2003 and 2018 in the Secure Anonymized Information Linkage (SAIL) Databank. People aged ≥65years residing in long-term care homes in Wales. Competing risk models and logistic regression models were used to examine the association between prior stroke, incident stroke, and mortality following stroke. Of 86,602 individuals, 7.0% (n= 6055) experienced a stroke in the 12months prior to care home entry. The incidence of stroke within 12months after entry to a care home was 26.2 per 1000 person-years [95% confidence interval (CI) 25.0, 27.5]. Previous stroke was associated with higher risk of incident stroke after moving to a care home (subdistribution hazard ratio 1.83, 95% CI 1.57, 2.13) and 30-day mortality following stroke (odds ratio 2.18, 95% CI 1.59, 2.98). Severe frailty was not significantly associated with risk of stroke or 30-day mortality following stroke. Secondary stroke prevention included statins (51.0%), antiplatelets (61.2%), anticoagulants (52.4% of those with atrial fibrillation), and antihypertensives (92.1% of those with hypertension). At the time of care home entry, individuals with history of stroke in the previous 12months are at a higher risk of incident stroke and mortality following an incident stroke. These individuals are frequently not prescribed medications for secondary stroke prevention. Further evidence is needed to determine the optimal care pathways for older people living in long-term care homes with history of stroke.

Birth Outcomes of Infants Born to Mothers With Disability: A Population at Risk.

Birth Outcomes of Infants Born to Mothers With Disability: A Population at Risk.

Using Linked Data Sources to Enhance Catalog Discovery

Our research explores how linked data sources and non-library metadata can support open-ended discovery of library resources. We also consider which experimental methods are best suited to the improvement of library catalog systems. We provide an overview of the questions driving our discovery experiments with linked data, a summary of some of our usability findings, as well as our design and implementation approach. In addition, we situate the discussion of our work within the larger framework of library cataloging and curation practices.

DEEP LEARNING ALGORITHM FOR CONNECTING SCIENTIFIC RECORDS AND SOCIAL PLATFORM

In the healthcare industry, professionals develop big amounts of disorganized data. The complexity of this data and the loss of computational capability lead to delays in the investigation. Nevertheless, with the advent of Deep Learning algorithms and connection to computing power such as Graphic Processor Units (GPUs) and Tensor Processing Units (TPUs), text and image processing has become usable. Deep Learning (DL) data bring about a big outcome in Natural Language Processing (NLP) and computer perception. The main purpose of this study is to build an undivided approach that can relate social platforms, literature, and scientific records to develop an approach to medicinal education for the public and experts.
 This study focuses on NLP in the healthcare industry and compiles data by Electronic Medical Records (EMR), medical literature, and social platforms. The framework proposed in this study is one for connecting social platforms, medical literature, and Electronic Medical Records scientific records using Deep Learning algorithms. Linking data sources requires defining the relationships between them, and finding concepts in medical texts. The National Library of Medicine (NLM) introduced the Unified Medical Language System (UMLS) and uses this system as the basis for the proposed system. The dynamic nature of a social platform can be recognized and supervised methodologies can be applied under supervision to develop conception. Named entity Recognition (NER) enables the active eradication of data or individuals by the pharmaceutical literature.

Scalable resource description framework clustering: A distributed approach for analyzing knowledge graphs using minHash locality sensitive hashing

AbstractWeb is becoming rich in data. Some of the sources from where these data are originating includes Blogs, YouTube, Twitter, Emails, E‐commerce, Banking, sensors, and the Internet of Things. But these data are structured in a very poor fashion. The content of the web is becoming heterogeneous in nature both in terms of compendium and structure. It can be said that these data are human‐readable data but the main motive is to draw inferences from these data which is only possible if it can be made machine‐accessible. Clustering is considered an important task to organize these data and draw meaningful inferences from these data. In this paper, a clustering approach is proposed that can be applied to knowledge graphs and the possibility of applying Locality Sensitive Hashing is explored. Given the size of linked data, it is observed that this approach can be effective and scalable in comparison to other clustering approaches such as Hierarchical clustering, K‐Means clustering, and K‐Medoid clustering in discovering different communities that are defined by the link structure of the graph. The experimental results on different types of Linked Data sources justify the efficacy of the proposed model in terms of scalability and efficiency.

The Impact of Care Intensity and Work on the Mental Health of Family Caregivers: Losses and Gains.

We estimate the causal impact of intensive caregiving, defined as providing at least 80 h of care per month, and work on the mental health of caregivers while considering possible sources of endogeneity in these relationships. We use 2 linked data sources from the United States by matching caregivers in the National Study of Caregiving with corresponding care recipients in the National Health and Aging Trends Study for years 2011-2017. We address possible sources of endogeneity in the relationships between caregiving, work, and mental health by using instrumental variables methodology, instrumenting for both caregiving and work behavior. We examine 2 measures used to screen for depression (PHQ-2, psychodiagnostic test) and anxiety (GAD-2, generalized anxiety disorders screening instrument), a composite measure that combines these measures (PHQ-4), and positive well-being variables to ascertain possible gains from caregiving. Providing at least 80 h of care per month to a parent compared to less intensive caregiving increases the PHQ-4 scale for anxiety and depression disorders. This is driven by the screening score for anxiety and not psychodiagnostic test scores for depression. Relationship quality decreases substantially for intensive caregivers, and intensive caregiving leads to less satisfaction that the care recipient is well-cared for. We do not find offsetting mental health gains for intensive caregivers compared to nonintensive caregivers. Work does not independently affect the mental health of caregivers. Caregiver interventions that reduce objective demands or support intensive caregivers could reduce or prevent well-being losses and improve the caregiver's relationship with the recipient.

Methodological guidelines to estimate population-based health indicators using linked data and/or machine learning techniques

BackgroundThe capacity to use data linkage and artificial intelligence to estimate and predict health indicators varies across European countries. However, the estimation of health indicators from linked administrative data is challenging due to several reasons such as variability in data sources and data collection methods resulting in reduced interoperability at various levels and timeliness, availability of a large number of variables, lack of skills and capacity to link and analyze big data. The main objective of this study is to develop the methodological guidelines calculating population-based health indicators to guide European countries using linked data and/or machine learning (ML) techniques with new methods.MethodWe have performed the following step-wise approach systematically to develop the methodological guidelines: i. Scientific literature review, ii. Identification of inspiring examples from European countries, and iii. Developing the checklist of guidelines contents.ResultsWe have developed the methodological guidelines, which provide a systematic approach for studies using linked data and/or ML-techniques to produce population-based health indicators. These guidelines include a detailed checklist of the following items: rationale and objective of the study (i.e., research question), study design, linked data sources, study population/sample size, study outcomes, data preparation, data analysis (i.e., statistical techniques, sensitivity analysis and potential issues during data analysis) and study limitations.ConclusionsThis is the first study to develop the methodological guidelines for studies focused on population health using linked data and/or machine learning techniques. These guidelines would support researchers to adopt and develop a systematic approach for high-quality research methods. There is a need for high-quality research methodologies using more linked data and ML-techniques to develop a structured cross-disciplinary approach for improving the population health information and thereby the population health.

Generating Real-World Evidence on the Quality Use, Benefits and Safety of Medicines in Australia: History, Challenges and a Roadmap for the Future.

Australia spends more than $20 billion annually on medicines, delivering significant health benefits for the population. However, inappropriate prescribing and medicine use also result in harm to individuals and populations, and waste of precious health resources. Medication data linked with other routine collections enable evidence generation in pharmacoepidemiology; the science of quantifying the use, effectiveness and safety of medicines in real-world clinical practice. This review details the history of medicines policy and data access in Australia, the strengths of existing data sources, and the infrastructure and governance enabling and impeding evidence generation in the field. Currently, substantial gaps persist with respect to cohesive, contemporary linked data sources supporting quality use of medicines, effectiveness and safety research; exemplified by Australia’s limited capacity to contribute to the global effort in real-world studies of vaccine and disease-modifying treatments for COVID-19. We propose a roadmap to bolster the discipline, and population health more broadly, underpinned by a distinct capability governing and streamlining access to linked data assets for accredited researchers. Robust real-world evidence generation requires current data roadblocks to be remedied as a matter of urgency to deliver efficient and equitable health care and improve the health and well-being of all Australians.

An introduction of the Accelerated Development of VAccine beNefit-risk Collaboration in Europe (ADVANCE) Project

The vaccine benefit-risk assessment entails comprehensive and systematic evaluation of the major benefits and risks of vaccine based on information after a marketing authorization. In 2009, the influenza A (H1N1) pandemic led to the licensing of new H1N1 vaccines in many countries including Europe. The post-marketing evaluation and monitoring of H1N1 vaccines was carried out based on a project previously conducted in Europe. In 2013, the Accelerated Development of VAccine beNefit-risk Collaboration in Europe (ADVANCE) project was launched with the support of Innovative Medicines Initiative (IMI) to address the limitations reported in the previous project. ADVANCE consisted of seven Work Packages (WPs), each indicating the vaccine benefitrisk monitoring readiness across four areas: Governance, Data source, Methods, and Study. First, a model for vaccine surveillance and governance with various functions such as decision making, quality management, execution, and financial administration was established. Also, the project developed a tool to identify, collect, and link the data sources available for vaccine surveillance. In addition, various methodologies were reviewed to assess the infectious disease burden, vaccine coverage, and benefit-risk based on linked data sources in Europe. Guidelines were proposed for real-world vaccine surveillance studies. Accordingly, a variety of tools and programs were developed, including a code of conduct related to the benefit-risk assessment of vaccines, ADVANCE International Research Readiness (AIRR), CodeMapper, and VaccO. Since the end of the ADVANCE project in 2019, the Vaccine Monitoring Collaboration for Europe (VAC4EU) has been conducting effective and sustainable vaccine benefit-risk monitoring programs based on the vision and ideology of ADVANCE. This study explains the contents and findings of ADVANCE with emphasis on WPs and proposes directions for establishing a vaccine benefit-risk assessment system in Korea.

Case-ascertainment of acute myocardial infarction hospitalisations across two national data sources for an English cardio-oncology cohort

Abstract Background Cardio-oncology is an emerging field aimed at minimising cardiovascular morbidity and mortality in survivors of cancer. Though cardio-oncology patients make up a significant proportion of hospital admissions, they are typically excluded from clinical trials. Electronic health record-based research is a good alternative to understand the risk of cardiovascular disease in these patients. However, completeness and validity of cardiovascular outcome ascertainment may vary by data source. Purpose 1) Determine ascertainment of acute myocardial infarction (AMI) hospitalisations across two national data sources for English cancer patients, and 2) understand factors that affect ascertainment in this population. Methods English cancer patients were obtained from the National Cancer Registration and Analysis Service. AMI hospital admissions occurring during 2010–2017 were identified from two sources: the Myocardial Ischaemia National Audit Project (MINAP) registry and Hospital Episode Statistics (HES). Admissions were ascertainable from both MINAP and HES, or “matched”, if dated within 30 days. Results 112,502 hospital admissions for 95,509 patients were identified of which 57,265 (50.9%) admissions were captured in MINAP and HES, 26,104 (23.2%) in MINAP only, and 29,133 (25.9%) in HES only. Case-ascertainment percentages were similar across cancer site, stage and grade of disease. Cases ascertained from MINAP only were younger, had fewer comorbidities, and were more likely to have ST-elevated MI and better survival than those ascertained from HES only. The percentage of admissions that could be ascertained from both MINAP and HES has improved from over time. Conclusions MINAP and HES define different populations of cancer patients with AMI. Only half of AMI hospital admissions were captured in both MINAP and hospital records. Therefore, these results support the use of multiple linked data sources for observational cardio-oncological research into AMI. Funding Acknowledgement Type of funding sources: Foundation. Main funding source(s): British Heart Foundation and Cancer Research UK

Named Entity Disambiguation for Archival Collections: Metadata, Wikidata, and Linked Data

AbstractRepresenting archival metadata as linked data can increase the findability and usability of items, and linked data sources such as Wikidata can be used to further enrich existing collection metadata. However, a central challenge to this process is the named entity disambiguation or entity linking that is required to ensure that the named entities in a collection are being properly matched to Wikidata entities so that any additional metadata is applied correctly. This paper details our experimentation with one entity linking system called OpenTapioca, which was chosen for its use of Wikidata and its accessibility to librarians and archivists with minimal technical intervention. We discuss the results of using OpenTapioca for named entity disambiguation on the Belfer Cylinders Collection from the Special Collections Research Center at Syracuse University, highlighting the successes and limitations of the system and of using Wikidata as a knowledge base.

Disparities in Utilization of Psychiatry Services Among Home Care Clients: The Tale of Two Canadian Jurisdictions.

Publicly funded home care in Canada supports older adults in the community to delay institutional care, which results in complex care populations with multimorbidity that includes mental health problems. The purpose of this study is to examine prevalence of psychiatric diagnoses and other mental health symptoms among older clients in two publicly funded Home Care (HC) Programs and their psychiatry service utilization (psychiatrist visits) after being admitted to home care. This retrospective cohort study examines clients age 60 years and older in the two Canadian provinces of Manitoba (MB), specifically the Winnipeg Regional Health Authority (WRHA) (n = 5,278), and Nova Scotia (NS) (n = 5,323). Clients were admitted between 2011 and 2013 and followed up to 4 years. Linked data sources include the InterRAI Resident Assessment Instrument for Home Care (RAI-HC), physician visit/billing data and hospital admission data. Both regions had similar proportions (53%) of home care clients with one or more psychiatric diagnoses. However, we observed over 10 times the volume of psychiatry visits in the WRHA cohort (8,246 visits vs. 792 visits in NS); this translated into a 4-fold increased likelihood of receiving psychiatry visits (17.2% of WRHA clients vs. 4.2% of NS clients) and 2.5 times more visits on average per client (9.1 avg. visits in MB vs. 3.6 avg. visits in NS). The location of psychiatry services varied, with a greater number of psychiatry visits occurring while in hospital for WRHA HC clients compared to more visits in the community for NS HC clients. Younger age, psychotropic medication use, depressive symptoms, dementia, and having an unstable health condition were significantly associated with receipt of psychiatry visits in both cohorts. Access to psychiatric care differed between the cohorts despite little to no difference in need. We conclude that many home care clients who could have benefitted from psychiatrist visits did not receive them. This is particularly true for rural areas of NS. By linking the RAI-HC with other health data, our study raises important questions about differential access to psychiatry services by site of care (hospital vs. community), by geographical location (MB vs. NS and urban vs. rural) and by age. This has implications for staff training and mental health resources in home care to properly support the mental health needs of clients in care. Study results suggest the need for a mental health strategy within public home care services.

Linking Data and Descriptions on Moths Using the Wikimedia Ecosystem

Linking Data and Descriptions on Moths Using the Wikimedia Ecosystem

749The global prevalence of Rheumatoid Arthritis: A meta-analysis

Abstract Background To determine the global prevalence of rheumatoid arthritis (RA) based on published studies and reveal factors which influence the RA prevalence estimates. Methods Four electronic databases- ProQuest Central, MEDLINE, Web of Science, and EMBASE were searched for publications from 1980 and 2019, reporting prevalence estimates of RA. A random-effect meta-analysis model was used to produce the pooled prevalence estimates. The potential sources of between-study heterogeneity were identified using sensitivity analysis, sub-group and meta-regression analyses. Results A total of 67 studies consisting of 212,335,171 patients were included in the meta-analysis. The global prevalence of RA was estimated 0.46% (95% CI: 0.39-0.54; I2=99.9%) with a 95% prediction interval (0.06–1.27). The point-prevalence of RA was 0.45% (95% CI: 0.38- 0.53%), while the pooled period-prevalence was 0.46% (0.36% and 0.57%). The highest RA pooled prevalence was estimated at 0.69% (95% CI: 0.47–0.95) derived from linked data sources studies. Based on subgroup analyses, the pooled prevalence of RA was influenced by geographical location, the risk bias of studies, period-prevalence method and urban population setting over the stratified periods. Conclusion The global prevalence of RA was 460 per 100,000 population from 1980–2018, with a 95% prediction interval (0.06– 1.27%). RA prevalence estimates were influenced by geographical location, the risk bias assessment of studies, period-prevalence method and urban population setting over time. Key messages The global prevalence of RA was 460 per 100,000 population from 1980–2018, with a 95% prediction interval (0.06– 1.27%).

Within and across country variations in treatment of patients with heart failure and diabetes.

To compare within-country variation of health care utilization and spending of patients with chronic heart failure (CHF) and diabetes across countries. Patient-level linked data sources compiled by the International Collaborative on Costs, Outcomes, and Needs in Care across nine countries: Australia, Canada, England, France, Germany, New Zealand, Spain, Switzerland, and the United States. Patients were identified in routine hospital data with a primary diagnosis of CHF and a secondary diagnosis of diabetes in 2015/2016. We calculated the care consumption of patients after a hospital admission over a year across the care pathway-ranging from primary care to home health nursing care. To compare the distribution of care consumption in each country, we use Gini coefficients, Lorenz curves, and female-male ratios for eight utilization and spending measures. In all countries, rehabilitation and home nursing care were highly concentrated in the top decile of patients, while the number of drug prescriptions were more uniformly distributed. On average, the Gini coefficient for drug consumption is about 0.30 (95% confidence interval (CI): 0.27-0.36), while it is, 0.50 (0.45-0.56) for primary care visits, and more than 0.75 (0.81-0.92) for rehabilitation use and nurse visits at home (0.78; 0.62-0.9). Variations in spending were more pronounced than in utilization. Compared to men, women spend more days at initial hospital admission (+5%, 1.01-1.06), have a higher number of prescriptions (+7%, 1.05-1.09), and substantially more rehabilitation and home care (+20% to 35%, 0.79-1.6, 0.99-1.64), but have fewer visits to specialists (-10%; 0.84-0.97). Distribution of health care consumption in different settings varies within countries, but there are also some common treatment patterns across all countries. Clinicians and policy makers need to look into these differences in care utilization by sex and care setting to determine whether they are justified or indicate suboptimal care.

Health Query Expansion based on Graph Matching between DBpedia and UMLS

<p class="0abstract">Information Retrieval (IR) in the medical domain is considered as a challenging task for many reasons. Short health queries tend to lack information on user's intent, and the target corpus may not have sufficient information for Relevance Feedbacks. And even, if the user obtains relevant documents to his/her queries, it is difficult for him/her to understand the technical terms. In contrast, in this paper, we propose an approach for health queries reformulation based on graph matching between two external linked data sources: DBpedia and Unified Medical Language System (UMLS). DBpedia has a broad coverage of topics and less noise compared to Wikipedia articles, and UMLS is specific to the medical domain. We also introduced the degree centrality to measure the graph connectivity and to select the most efficient candidate terms for query expansion. Experimental results on MEDLINE collection using Okapi BM25 as a retrieval model showed that our approach outperformed related methods, and the two sources achieved very good retrieval results. They helped in the diversification of the retrieved documents and the improvement of the recall.</p>

Case-ascertainment of acute myocardial infarction hospitalizations in cancer patients: a cohort study using English linked electronic health data.

AimsTo assess the recording and accuracy of acute myocardial infarction (AMI) hospital admissions between two electronic health record databases within an English cancer population over time and understand the factors that affect case-ascertainment.Methods and resultsWe identified 112 502 hospital admissions for AMI in England 2010–2017 from the Myocardial Ischaemia National Audit Project (MINAP) disease registry and hospital episode statistics (HES) for 95 509 patients with a previous cancer diagnosis up to 15 years prior to admission. Cancer diagnoses were identified from the National Cancer Registration Dataset (NCRD). We calculated the percentage of AMI admissions captured by each source and examined patient characteristics associated with source of ascertainment. Survival analysis assessed whether differences in survival between case-ascertainment sources could be explained by patient characteristics. A total of 57 265 (50.9%) AMI admissions in patients with a prior diagnosis of cancer were captured in both MINAP and HES. Patients captured in both sources were younger, more likely to have ST-segment elevation myocardial infarction and had better prognosis, with lower mortality rates up to 9 years after AMI admission compared with patients captured in only one source. The percentage of admissions captured in both data sources improved over time. Cancer characteristics (site, stage, and grade) had little effect on how AMI was captured.ConclusionMINAP and HES define different populations of patients with AMI. However, cancer characteristics do not substantially impact on case-ascertainment. These findings support a strategy of using multiple linked data sources for observational cardio-oncological research into AMI.