Catálogo de publicaciones - revistas

<jats:p>ObjectivesChildhood adversity (CA) carries an increased risk of developing later mental health (MH) problems and suicidal behaviour. This study aimed to summarise lifetime hospital attendances for CA and MH in young people who later died by suicide.&#x0D; ApproachThis study is a retrospective longitudinal case control study. Lifetime Scottish inpatient acute and psychiatric records were linked to death records and summarised for individuals born since 1981 who died by suicide in the period 1991-2017 (cases), and controls (1:10) matched on sex, age, and postcode. Relevant records were coded MH (including self-harm) and/or CA. Descriptive statistics and odds ratios (OR) were computed.&#x0D; ResultsData for 2,477 and 24,777 ‘cases’ and ‘controls’ were extracted, of whom 2,106 cases (85%) and 13,589 controls (55%) had lifespan hospital records. Mean age at death for cases was 23.7 (SD=4.9) and 75.9% were male. Psychiatric records represented 11.6% and 1.4% of records for cases and controls, respectively.&#x0D; For the age range 10-18, Maltreatment &amp; violence-related codes were recorded for 160 (7.6%) cases and 371 (2.7%) controls, corresponding to OR=2.9 (95%CI: 2.4-3.6). This was compared with MH at 458 (21.7%) cases and 560 (4.1%) controls and OR=6.5 (95%CI: 5.7-7.4). The highest adjusted ORs were for self-harm episodes recorded in general hospital with aORmale=6.56 (95%CI: 4.96-8.68) and aORfemale=6.87 (95%CI: 4.99-9.48).&#x0D; ConclusionAll CA and MH presentations in inpatient hospital records were associated with greater risk of subsequent suicide, with the strongest association for self-harm.</jats:p>

<jats:p>ObjectivesUnder section 31 (s.31) of the UK Children Act 1989, public law care proceedings can be issued if there is concern a child is subject to, or at risk of significant harm. We examined health vulnerabilities of parents involved in public law care proceedings in the two-year period prior to involvement.&#x0D; ApproachOur study created an anonymised individual-level population-based cohort, with a matched comparison group of parents in Wales who were not subject to care proceedings, matched on age, sex and deprivation. Family court data provided by Cafcass Cymru were linked to population-level healthcare records held within the Secure Anonymised Information Linkage (SAIL) Databank. Demographic characteristics, overall health service use and health profiles of parents of children subject to s.31 care proceedings between 2011 and 2019 were examined.&#x0D; ResultsData were available for 8,821 parents involved in care proceedings between 2011 and 2019, with a comparison group of 32,006 parents. Nearly half (47.6%) of cohort parents resided in the most deprived quintile. Higher levels of healthcare use were found for cohort mothers and fathers compared to the comparison group across multiple healthcare settings, with the most pronounced differences for emergency department attendances (59.3% vs 37.0%). Health conditions with the largest variation between groups were related to mental health (43.6% vs 16.0%), substance use (19.4% vs 1.6%) and injuries (41.5% vs 23.6%).&#x0D; ConclusionThis study highlights the heightened socioeconomic and health vulnerabilities of parents who experience care proceedings concerning a child. Better understanding of the needs and vulnerabilities of this population may provide opportunities to improve a range of support and preventative interventions that respond to crises in the community.</jats:p>

<jats:p>ObjectiveThe objective was to examine the relationship between maternal intimate partner violence (IPV) victimization and children’s developmental health using linked population-wide administrative datasets. We examined developmental vulnerability (DV) at kindergarten of children exposed to maternal IPV victimization aged 0 to 5 using provincial prosecution records compared to unexposed counterparts.&#x0D; ApproachThis retrospective cohort study linked administrative datasets (legal, health, education, social services) from the Population Research Data Repository at the Manitoba Centre for Health Policy. Exposed mother-child pairs with 1+ prosecution records of maternal IPV victimization during early childhood (child aged 0 to 5) between 2003-2018 in Manitoba (n = 5,728) were matched to unexposed pairs (1:3) based on sex/birthdate of child and neighbourhood income. DV at kindergarten was measured across 5 domains (physical, social, emotional, language/cognitive [LC], communication/general knowledge) using the Early Developmental Instrument (EDI). Children without eligible EDI scores were excluded. Multiple logistic regression models were conducted.&#x0D; ResultsThe cohort included 5321 children (exposed n=1365, unexposed n=3956). 32.98% of the cohort was developmentally vulnerable in one or more domains (1/+) and 19.60% was developmentally vulnerable in two or more domains (2/+). Unadjusted relationships between maternal IPV victimization from age 0 to 5 and developmental vulnerability at kindergarten were statistically significant across all 5 domains (e.g., LC OR=2.76 [2.36, 3.23]) and in 1/+ (OR=2.72 [2.39, 3.09]) as well as 2/+ (OR=2.89 [2.51, 3.34]) domains. After adjusting for covariates, children who were exposed to maternal IPV victimization from ages 0 to 5 had increased odds of being developmentally vulnerable in social competence (aOR=1.33 [1.07, 1.66]) and emotional maturity (aOR=1.29 [1.03, 1.62]), also in 2/+ domains (aOR=1.42 [1.15, 1.73]) at kindergarten, compared to unexposed counterparts.&#x0D; ConclusionThe study provided Canadian population-wide evidence of the association between maternal IPV victimization and early childhood development, specifically later socio-emotional vulnerability. Interventions and support systems for this population of families should be developed and implemented, with an emphasis on mitigating long-term socio-emotional developmental risks in children exposed to IPV.</jats:p>

<jats:p>ObjectiveThe aim of this project is to analyse temporal patterns of repeat violent victimisation (RVV) in Scotland using administrative data and explore the extent to which these patterns are associated with underlying health vulnerabilities (related to mental health and problematic alcohol and/or drug use) and vary by individual socio-demographic characteristics.&#x0D; ApproachViolence in Scotland, and especially RVV, is a ‘key public health priority’. However, very little is known about the link between violence and other health vulnerabilities at an individual level, which may be useful in informing prevention approaches. This paper uses data on violence-related ambulance call-outs linked to hospital records containing information on wider health vulnerabilities. We use survival mixture models to identify and characterise population sub-groups, based on time to repeated violent episodes and explore how different health-related factors are associated with RVV and how this varies across sub-populations.&#x0D; ResultsWe identify two classes of RVV with two distinct survival trajectories: a ‘low vulnerability’ class and a ‘high vulnerability’ class. People in the high vulnerability class experienced RVV more quickly and were more likely to have underlying mental health conditions and/or problematic drug or alcohol use than those in the low vulnerability class.  The association between socio-demographic characteristics and RVV differed across the two classes, with significant inequality in RVV for high-risk individuals between those in the most and the least affluent communities. The findings suggest that the first 6 months after the index event is a critical period of RVV for people in the high-risk group, but there may be different exposure mechanisms for different types of victims.&#x0D; ConclusionBetter use of administrative data from health sources could help improve our understanding and identification of repeat victims of violence. The ability to identify whether victims are likely to be ‘high-risk’ or ‘low-risk’ could inform the design of focused strategic approaches and interventions aimed at reducing repeat violent victimization.</jats:p>

<jats:p>ObjectivesThis research seeks to understand viewpoints on the use of data containing personally identifiable information by a range of organizations for divergent purposes, focusing on the sharing and linkage of administrative data for the purposes of conducting research and evaluating public policies in Brazil.&#x0D; ApproachAn exploratory approach was employed to perceive how the concepts relevant to this subject are understood, experienced and expressed by our interlocutors. A semi-structured interview technique was chosen to establish a base of scripted questions capable of building on and adapting to discussions with the participants, as well as addressing the questions that emerged during interviews. The interviewees were divided into three groups: data subjects (patients and beneficiaries of social welfare programs), researchers, and managers with experience in public policy in the areas of health and social protection. Specific scripts were elaborated for each group of respondents.&#x0D; ResultsGroups and individuals are constantly balancing risks and benefits with regard to exposing and sharing their data; risks weigh more heavily depending on the individual’s socioeconomic context, which is permeated by intersectionality. The processing of personal data by the government raises more fears than actions taken by large technology companies. Individuals and social groups want to receive feedback on research carried out in their communities, and also desire to participate in the design of scientific research and the analysis of evidence used to guide public policies which directly affect them. Data governance is indispensable, requiring not only data management but also specific conditions on data sharing and linkage, principally regarding the sharing of administrative data aligned with legitimate public interests.&#x0D; ConclusionRaising awareness and providing information on individual and collective rights on personal and sensitive data collection, as well as informing the public about the purposes of data sharing and linkage, is of utmost importance for responsible data management.  Administrative data governance should be planned and implemented to foster trust and transparency among all involved and interested parties.</jats:p>

<jats:p>ObjectivesAll children have a right to education, but research indicates that those receiving children’s social care (CSC) services are at increased risk of non-enrolment in school, including through off-rolling (illegal exclusion). We aimed to use administrative data to estimate the association between CSC history and non-enrolment in secondary school.&#x0D; ApproachUsing the National Pupil Database (data on all English state school enrolments), we identified a cohort of 1,059,781 pupils aged 11 in 2011 and 2012. Children were categorised as having a history of being children in need (CiN), on child protection plans (CPPs) or looked after (CLA) using linked data from CSC services. We estimated the proportion of children not enrolled across ages 12 to 16 by CSC history. We then assessed with regression modelling the association between CSC history and non-enrolment in years 10/11. We also examined variation in overall non-enrolment rates between local authorities and regions.&#x0D; ResultsOf children without CSC history, 3.8% had 1 or more non-enrolments across ages 12 to 16. This proportion was higher among children with a history of being CiN (8.1%), on a CPP (9.4%) or being CLA (10.4%). The odds of non-enrolment in years 10/11 were higher among those with CLA history vs non-exposed peers (OR 4.76, 95% CI 4.49-5.05) as well as in those with CPP (3.60, 3.39-3.81) and CiN history (2.53, 2.49-2.58). History of special educational needs and disability (SEND) further increased non-enrolment odds. These associations and interactions persisted after adjusting for confounders. Non-enrolment rates were highest in the London region and varied significantly between local authorities.&#x0D; ConclusionOur findings show that children with CSC history (especially those with SEND) are more likely to be non-enrolled in secondary school than other children. Work is needed to understand the non-enrolment mechanisms, which may include illegal off-rolling and other exclusionary practices, to improve the education of children with CSC history.</jats:p>

<jats:p>ObjectivesShielding was introduced as part of the UK government’s response to the SARS-CoV-2 pandemic to protect Clinically Extremely Vulnerable (CEV) people from infection and serious illness. Various research questions emerged in relation to non-clinical vulnerabilities of those shielding which could be addressed by utilising available health and administrative data.&#x0D; ApproachThe Shielded Patient List (SPL) was linked with various datasets on the UK Secure Electronic Research Platform (UKSERP) including the Pupil Level Annual School Census (PLASC), National Survey and Ordnance Survey data for Wales. Some of these were anonymised datasets contained in the Secure Anonymised Information Linkage (SAIL) databank. Algorithms were applied to determine household composition and whether private outdoor space was available for the shielding group. Results were then extracted for Wales broken down by local authority.&#x0D; ResultsResults from the various strands of research related to shielding will be presented covering provision of outdoor space, household characteristics and composition.&#x0D; ConclusionThese analyses demonstrate how population-level data resources can be leveraged quickly to answer newly-emerging policy questions as part of the response to the SARS-CoV-2 pandemic.</jats:p>

<jats:p>Although machine learning has permeated many disciplines, the convergence of causal methods and machine learning remains sparse in the existing literature. Our aim was to formulate a marginal structural model in which we envisioned hypothetical (i.e. counterfactual) dynamic treatment regimes using a combination of drug therapies to manage diabetes: metformin, sulfonylurea and SGLT-2. We were interested in estimating “diabetes care provision” in next calendar year using a composite measure of chronic disease prevention and screening elements. We demonstrated the application of dynamic treatment regimes using the National Diabetes Action Canada Repository in which we applied a collection of mainstream statistical learning algorithms. We generated an ensemble of statistical learning algorithms using the SuperLearner based on the following base learners: (i) least absolute shrinkage and selection operator, (ii) ridge regression, (iii) elastic net, (iv) random forest, (v) gradient boosting machines, (vi) neural network. Each statistical learning algorithm was fitted using the pseudo-population with respect to the marginalization of the time-dependent confounding process. The covariate balance was assessed using the longitudinal (i.e. cumulative-time product) stabilized weights with calibrated restrictions. Our results indicated that the treatment drop-in cohorts (with respect to metformin, sulfonylurea and SGLT-2) may improve diabetes care provision in relation to treatment naïve cohort. As a clinical utility, we hope that this article will facilitate discussions around the prevention of adverse chronic outcomes associated with diabetes through the improvement of diabetes care provisions in primary care.</jats:p>

<jats:p>ObjectivesDeveloping a rigorous application process for COVID-19 linked data within the UK Longitudinal Linkage Collaboration’s Trusted Research Environment (TRE) included balancing the requirements of 20+ longitudinal studies and several data providers with the need to provision data to researchers in a safe and efficient manner.&#x0D; ApproachThe UK LLC is an innovative project, developing a new approach to linking &gt;20 well-established longitudinal studies to routine health and administrative records. Setting up the process included consultation with stakeholders and other multifaceted data resources in order to implement the best practices available. We also had to acknowledge constraints imposed by a variety of data controllers and ensure meaningful public contribution. The initial process had to be flexible while we assessed the first applications, and because the TRE was regularly being updated with new data. This required an adaptive approach where the process was created, adopted, and re-evaluated in real-time.&#x0D; ResultsThe UK LLC has a three-stage application and review process, starting with an enquiry through the HDRUK Innovation Gateway. Enquiries are triaged to ensure the applications meet basic requirements (UK-based, COVID-19 related, scientifically sound, and feasible) before the applicants are invited to submit a full application. The full application is then reviewed internally against the Five Safes model, in communication with applicants. Approved applications move on to the third stage which includes a review by Data Access Committees (DACs) for each study whose data is requested, and the UK LLC DAC - comprised of a public review panel advising on lay summaries and public involvement in the proposed project, and, if applicable, an additional linked data review panel on behalf of NHS Digital.&#x0D; ConclusionOur process is still evolving as new datasets are added and in response to feedback. However, the team has risen to the challenge of balancing this with the needs of 20+ studies, stakeholders, applicants, and participant/public contributors, to create an agile process that can handle complex needs and requirements.</jats:p>

<jats:p>Educational attainment disparities across social groups remain at the forefront of contemporary UK society. Despite this, Northern Ireland reflects a somewhat different context to the rest of the UK due to its transition to a post-conflict society and its dually selective education system (academically and religiously). In Northern Ireland, post-primary (GCSE) attainment differences are often reported according to gender, religious affiliation and socio-economic background. However, due to the lack of available education data that encompasses a wide range of pupil- and school-level factors, discourse informed by the statistical testing of such factors has been limited. This study aims to overcome this current gap by examining the effects of socio-demographics, namely gender, religious affiliation and socio-economic background (through eight measures), and school-level factors on GCSE attainment, using the first linked administrative dataset for education in Northern Ireland. The data combined the household Census (2011) with the School Census (2010-2014) and School Leavers Survey (2010-2014) for the first time in Northern Ireland. To this end, this paper discusses data analytics of the study including data linkage, cohort size, constructed GCSE attainment measures, socio-demographic measures and school-level factors. The multilevel modelling (including interaction models) construction, execution and results will also be discussed. The paper concludes with a reflection upon whether the results of this analysis support existing literature in the Northern Ireland context and wider UK GCSE attainment trends.</jats:p>