Administrative Data Systems Research Articles

Protecting Privacy and Expanding Access in a Modern Administrative Tax Data System

Protecting Privacy and Expanding Access in a Modern Administrative Tax Data System

Return to Soccer After Anterior Cruciate Ligament Reconstruction: An Outcome or a Decision?

The return-to-sports rate is often used as an outcome measure after anterior cruciate ligament (ACL) reconstruction (ACLR). Although most soccer players want to return to sports after their ACL injury, up to 40% do not believe they will return to soccer after their ACL injury. To investigate whether self-reported presurgical beliefs regarding return to soccer correspond to registered official match participation after ACLR. Cohort study; Level of evidence, 2. Included were soccer players with a primary unilateral ACL injury who were ≥15 years at the time of ACLR and who had completed a presurgical question about their future sports participation beliefs. ACLR surgical data were extracted from the Swedish National Knee Ligament Registry, and game participation data were extracted from the Swedish Football Association's administrative data system. A total of 959 soccer players (617 [64%] male and 342 [36%] female) were included. The follow-up time after ACLR ranged from 18 months to 5.5 years. Of 720 soccer players (75%) who believed that they would return to play (RTP), 462 (64%) players actually did. Of the players who believed that they would not RTP, 181 (76%) did not. Presurgical beliefs predicted RTP (odds ratio [OR], 5.59; P < .001). Younger age at the time of ACLR favored RTP, where 61% of the players aged 15-20 years had RTP (OR, 3.85; P < .001). At the top competitive level, 84% of the players RTP compared to 14% at the recreational level. Players active at higher levels were more likely to believe that they would RTP, and they also actually did so (OR, 33.06; P < .001). Overall, 67% of players followed their presurgical intention to return to soccer. The findings indicated that presurgical beliefs, age, and level of play helped to predict the actual RTP of soccer players after ACLR.

Machine Learning For Risk Prediction After Heart Failure Emergency Department Visit or Hospital Admission Using Administrative Health Data.

Patients visiting the emergency department (ED) or hospitalized for heart failure (HF) are at increased risk for subsequent adverse outcomes, however effective risk stratification remains challenging. We utilized a machine-learning (ML)-based approach to identify HF patients at risk of adverse outcomes after an ED visit or hospitalization using a large regional administrative healthcare data system. Patients visiting the ED or hospitalized with HF between 2002-2016 in Alberta, Canada were included. Outcomes of interest were 30-day and 1-year HF-related ED visits, HF hospital readmission or all-cause mortality. We applied a feature extraction method using deep feature synthesis from multiple sources of health data and compared performance of a gradient boosting algorithm (CatBoost) with logistic regression modelling. The area under receiver operating characteristic curve (AUC-ROC) was used to assess model performance. We included 50,630 patients with 93,552 HF ED visits/hospitalizations. At 30-day follow-up in the holdout validation cohort, the AUC-ROC for the combined endpoint of HF ED visit, HF hospital readmission or death for the Catboost and logistic regression models was 74.16 (73.18-75.11) versus 62.25 (61.25-63.18), respectively. At 1-year follow-up corresponding values were 76.80 (76.1-77.47) versus 69.52 (68.77-70.26), respectively. AUC-ROC values for the endpoint of all-cause death alone at 30-days and 1-year follow-up were 83.21 (81.83-84.41) versus 69.53 (67.98-71.18), and 85.73 (85.14-86.29) versus 69.40 (68.57-70.26), for the CatBoost and logistic regression models, respectively. ML-based modelling with deep feature synthesis provided superior risk stratification for HF patients at 30-days and 1-year follow-up after an ED visit or hospitalization using data from a large administrative regional healthcare system.

Adapting to evidence users: accessibility and incentives

Adapting infrastructure for evidence creation and dissemination involves taking new technological, social, and environmental realities into account. Adaptations should also respond to what hasn’t changed: structural disincentives to evidence use among intended users. This presentation addresses the understudied problem of how administrative data systems go beyond enabling access to information – how data linkages or other infrastructure impacts policymakers or other intended information users. It addresses data sharing and data use, as well as public engagement and involvement. A recent practitioner convening examined how new administrative data infrastructure can catalyze evidence-building, and, specifically explored the incentives and disincentives for stakeholders' use of evidence from evaluations. Participants discussed the multiple functions of evaluations for nonprofits, funders, policymakers, and communities, and shared lessons in capacity building from the recent history of nonprofit &amp; human services evaluation. Preliminary results from this convening will be discussed in terms of their broader implications for the use of information resulting from future large-scale data linkages. Broadly, the convening validated the hypothesis that efforts to increase the availability of evidence often proceed without sufficient consideration of the history of non-uptake of evidence, and that there has been limited acknowledgement of this problem by practitioners of evidence generation. Much work is needed within the evaluation &amp; evidence, and nonprofit sector to address this challenge. Key themes for this work that emerged during the convening will illustrate what might work and what won’t work to address the non-uptake of evidence more broadly.

Race in Integrated Data Systems: Why, Wherefrom, and How

Integrated administrative data systems (IADs) are a powerful resource to implement research and analysis for social policy. While IADs may capture racial identity data from multiple administrative sources, there is no agreed upon criteria for whether and how to synthesize this information in a way that (1) produces knowledge to advance racial equity, while (2) underscoring race as a social construct. This study leverages a county-level IAD to test and analyze an event-table design to represent racial identity that, when informed with historical and community knowledge, may meet both of the aforementioned goals. We illustrate this approach applied to the development of a registry of youth experiencing homelessness, as captured by linked administrative data from vital records, homeless service agencies, schools, food support programs, and other systems. The event table design includes race identity for each person in the registry across all systems included. We develop criteria to inform the hierarchy of one source over another when there are discrepancies in race identity across systems, and when categories for this variable differ across administrative systems. We provide historical and social context behind potential discrepancies and discuss approaches to missing data. Furthermore, we highlight the value of including qualitative knowledge from agency data managers, users, and those represented in the data to inform the synthesis of information around race. Finally, we illustrate how this approach can guide research analysis and contextualize results, thus enhancing the research process and advancing racial equity with IADs.

Prevalence and incidence measures for schizophrenia among commercial health insurance and medicaid enrollees

Given the chronic nature of schizophrenia, it is important to examine age-specific prevalence and incidence to understand the scope of the burden of schizophrenia across the lifespan. Estimates of lifetime prevalence of schizophrenia have varied widely and have often relied upon community-based data estimates from over two decades ago, while more recent studies have shown considerable promise by leveraging pooled datasets. However, the validity of measures of schizophrenia, particularly new onset schizophrenia, has not been well studied in these large health databases. The current study examines prevalence and validity of incidence measures of new diagnoses of schizophrenia in 2019 using two U.S. administrative health databases: MarketScan, a national database of individuals receiving employer-sponsored commercial insurance (N = 16,365,997), and NYS Medicaid, a large state public insurance program (N = 4,414,153). Our results indicate that the prevalence of schizophrenia is over 10-fold higher, and the incidence two-fold higher, in the NYS Medicaid population compared to the MarketScan database. In addition, prevalence increased over the lifespan in the Medicaid population, but decreased in the employment based MarketScan database beginning in early adulthood. Incident measures of new diagnoses of schizophrenia had excellent validity, with positive predictive values and specificity exceeding 95%, but required a longer lookback period for Medicaid compared to MarketScan. Further work is needed to leverage these findings to develop robust clinical outcome predictors for new onset of schizophrenia within large administrative health data systems.

Fibrosis-4 score for screening of metabolic dysfunction-associated steatotic liver disease (MASLD): Data from a population-based sample in Tuscany

Background and aimTo verify the prevalence of positive Fibrosis-4 (FIB-4) score, a screening test for metabolic-associated liver disease, in a large population-based sample in the Tuscany Italian Region, and to identify sub-populations at higher risk which could be targeted by specific screening programs. Methods and resultsPopulation-based survey performed in the Italian region of Tuscany, with Tuscany health informative system's administrative data. We included 594,923 subjects, of which 32% had available data for the FIB-4 calculation. The overall proportion of subjects with an FIB-4 value > 1.3, was 41.6% of those with available exams, and 12,8% of the whole population, whereas 5.4% and 1.7% had FIB-4 >2.67. In those younger than 80 years, FIB >1.3 had a 33.1% and 9.4%. People with diabetes mellitus had higher figures (52.8.% and 28.9% for FIB>1.3). Among subjects aged 70 years or over, 74.9% of those with available data and 38.4% of the general population had a FIB-4>1.3, whereas 32% and 16% had a FIB-4 > 2. ConclusionsThe relevant proportion of FIB-4 positivity in the general population poses a significant burden for further screening with liver elastography. Targeting people with diabetes, excluding people older than 80 years and/or adopting a FIB-4 threshold of 2 in those aged more than 70 years could increase the cost-effectiveness of the screening procedures.

The Nature and Scope of Reported Child Maltreatment in Euro-CAN Countries: Current Evidence and Future Opportunities

Most European Cooperation on Science and Technology (COST) affiliated countries aim to advance the goal of protecting children from maltreatment. However, despite the increasing numbers of population-based surveys, the development of administrative data systems has lagged. In this study, we aimed to examine the current state of development of administrative data systems in a sample of countries represented in the COST Action 19106 network, Multi-Sectoral Responses to Child Abuse and Neglect in Europe: Incidence and Trends (Euro-CAN). A structured questionnaire was distributed to researchers and professionals within Euro-CAN-affiliated countries, which captured economic, legislative, systemic, and data infrastructure characteristics. Thematic trends for 13 sampled countries were presented descriptively. The implementation of legislative measures such as banning corporal punishment varied substantially, with some countries decades apart. Almost all sampled countries mandate reports of suspected child maltreatment for all or some professionals in contact with children. In most countries, public child protection, health, or law enforcement systems are decentralized, and unsubstantiated/inconclusive incidents of suspected child maltreatment are not systematically collected at the national level. Child maltreatment data is not routinely collected in health sectors in all sampled countries. Where data is collected in different sectors, such as police and child protection agencies, different descriptions are often used. Systematic data linkage remains a seldom occurrence with only a few countries offering this capability. The call for Euro-CAN countries to develop multi-sectoral data systems to capture recorded instances of child maltreatment remains relevant.

Trends in harmful drug exposure during pregnancy in France between 2013 and 2019: A nationwide cohort study.

Describe the trends of exposure to harmful drugs during pregnancy over recent years in France. Nationwide cohort study. The French National administrative health Data System (SNDS). Pregnancies starting between 2013 and 2019 and outcomes corresponding to live births, medical terminations of pregnancy, and stillbirths. Each pregnancy was divided into a preconceptional period of 90 days before conception and three trimesters from conception to birth. Harmful drugs were defined according to their risks to the fetus: teratogenicity or fetotoxicity. Exposure was defined using the critical period during pregnancy for each type of harmful drug: preconceptional period or first trimester for teratogenic drugs and second or third trimesters for fetotoxic drugs. Prevalence of pregnancies exposed to at least one harmful drug. Among 5,253,284 pregnancies, 204,402 (389 per 10,000) pregnancies were exposed to at least one harmful drug during the critical periods: 48,326 (92 per 10,000) pregnancies were exposed to teratogenic drugs during the preconceptional period or the first trimester, and 155,514 (299 per 10,000) pregnancies were exposed to fetotoxic drugs during the second or third trimesters. Teratogenic drugs were mainly retinoids for topical use (44 per 10,000 pregnancies), antiepileptics (13 per 10,000 pregnancies) and statins (13 per 10,000 pregnancies). Fetotoxic drugs were mainly non-steroidal anti-inflammatory drugs (NSAIDs), for systemic (128 per 10,000 pregnancies) and topical use (122 per 10,000 pregnancies). Exposure to teratogenic drugs decreased from the preconceptional period to the first trimester. Exposure to fetotoxic drugs decreased from the second to the third trimester. Between 2013 and 2019, we found a decrease in harmful drug exposure overall, mainly for topical and systemic NSAIDs and for topical retinoids. In this nationwide study, about one in 25 pregnancies was exposed to at least one harmful drug, mainly NSAIDs and topical retinoids. Although the prevalence of harmful drug exposure decreased over the study period, NSAID exposure in the second and third trimester remains of concern.

Assessing the role of social determinants of health in health disparities: The need for data on work.

Work is a key social determinant of health. Without the collection of work-related information in public health data systems, the role of social determinants in creating and reinforcing health disparities cannot be fully assessed. The Centers for Disease Control and Prevention (CDC) maintains or supports a number of public health surveillance and health monitoring systems, including surveys, case-based disease and exposure systems, vital status records, and administrative data systems. We evaluated a convenience sample of these systems for inclusion of information in three work-related domains: employment status, industry and occupation, and working conditions. While 12 of 39 data systems were identified as collecting work-related data, this information was often minimal (e.g., only employment status), restricted to a subset of respondents, or only gathered periodically. Information on working conditions was particularly sparse. Historically, the limited and inconsistent collection of work-related information in public health data systems has hindered understanding of the role work plays in health disparities. Current CDC data modernization efforts present opportunities to enhance the identification and mitigation of health disparities by prioritizing inclusion of an expanded set of work-related data elements.

Noncognitive Factors and Student Long-Run Success: Comparing the Predictive Validity of Observable Academic Behaviors and Social-Emotional Skills

Noncognitive constructs such as self-efficacy, social awareness, and academic engagement are widely acknowledged as critical components of human capital, but systematic data collection on such skills in school systems is complicated by conceptual ambiguities, measurement challenges and resource constraints. This study addresses this issue by comparing the predictive validity of two most widely used metrics on noncogntive outcomes—observable academic behaviors (e.g., absenteeism, suspensions) and student self-reported social and emotional learning (SEL) skills—for the likelihood of high school graduation and postsecondary attainment. Our findings suggest that conditional on student demographics and achievement, academic behaviors are several-fold more predictive than SEL skills for all long-run outcomes, and adding SEL skills to a model with academic behaviors improves the model’s predictive power minimally. In addition, academic behaviors are particularly strong predictors for low-achieving students’ long-run outcomes. Part-day absenteeism (as a result of class skipping) is the largest driver behind the strong predictive power of academic behaviors. Developing more nuanced behavioral measures in existing administrative data systems might be a fruitful strategy for schools whose intended goal centers on predicting students’ educational attainment.

Recording and analysing ethnicity in public health research: a bibliographical review and focus group discussions with young migrants and refugees in the UK

BackgroundThe ethnicity data gap hinders public health research from addressing ethnic health inequity in the UK, especially for under-served young, migrant populations. We aimed to review how ethnicity was captured, reported, analysed, and theorised within policy-relevant research. MethodsFor this bibliographical review, we reviewed a selection of the 1% most highly cited population health papers reporting UK ethnicity data in MEDLINE and Web of Science databases between Jan 1, 1946, and July 31, 2022, and extracted how ethnicity was recorded and analysed. We included cross-sectional, longitudinal cohort studies, and randomised trials using only UK populations, which were peer-reviewed, were written in English, and reported ethnicity and any health-related outcomes. We held three focus groups with ten participants aged 18–25 years, from Nigeria, Turkistan, Syria, Yemen, and Iran to help us shape and interpret our findings, and asked “How should ethnicity be asked inclusively, and better recorded?” and “Does ethnicity change over time or context? If so, why?”. We consolidated feedback from our focus groups into a co-created poster with recommendations for researchers studying ethnicity and health. Written informed consent was obtained for focus group participation. FindingsOf 44 papers included in the review, 19 (43%) used self-reported ethnicity, but the number of ethnic categories provided varied. Of 27 papers that aggregated ethnicity, 13 (48%) provided justification. Only eight (18%) explicitly theorised how ethnicity related to health. The focus groups agreed that (1) ethnicity should not be prescribed by others (individuals could be asked to describe their ethnicity in free-text, which researchers could synthesise to extract relevant dimensions of ethnicity for their research) and (2) Ethnicity changes over time and context according to personal experience, social pressure, and nationality change. The lived experience of ethnicity of migrants and non-migrants is not fully interchangeable, even if they share the same ethnic category. InterpretationResearchers should communicate clearly how ethnicity is operationalised in their studies, with appropriate justification for clustering and analysis that is meaningfully theorised. Our study was limited by its non-systematic nature. Implementing the recommendation to capture ethnicity via free text remains challenging in administrative data systems. FundingUCL Engagement Beacon Bursary.

Mortality benefits of reduction fine particulate matter in Vietnam, 2019.

Studies assessing the health benefits of air pollution reduction in Vietnam are scarce. This study quantified the annual mortality burden due to PM2.5 pollution in Vietnam above the World Health Organization recommendation for community health (AQG: 5 μg/m3) and the proposed National Technical Regulation on Ambient Air Quality (proposed QCVN: 15 μg/m3). This study applied a health impact assessment methodology with the hazard risk function for non-communicable diseases (NCDs) and lower respiratory infections (LRIs) in the Global Exposure Mortality Model (GEMM) to calculate attributable deaths, Years of Life lost, and Loss of Life expectancy at birth due to air pollution in the Vietnamese population above 25 years of age in 11 provinces. We obtained annual average PM2.5 concentrations for Vietnam in 2019 at a 3x3 km grid modeled using Mixed Linear regression and multi-data sources. Population and baseline mortality data were obtained from administrative data system in Vietnam. We reported the findings at both the provincial and smaller district levels. Annual PM2.5 concentrations in all studied provinces exceeded both the AQG and the proposed QCVN. The maximum annual number of attributable deaths in the studied provinces if they had complied with WHO air quality guidelines was in Ha Noi City, with 5,090 (95%CI: 4,253-5,888) attributable deaths. At the district level, the highest annual rate of attributable deaths if the WHO recommendation for community health had been met was 104.6 (95%CI: 87.0-121.5) attributable deaths per 100,000 population in Ly Nhan (Ha Nam province). A much larger number of premature deaths in Vietnam could potentially be avoided by lowering the recommended air quality standard. These results highlight the need for effective clean air action plans by local authorities to reduce air pollution and improve community health.

Harmonizing Disability Data To Improve Disability Research And Policy.

Disability is complex and multifaceted, complicating governments' efforts to collect the high-quality, comprehensive data necessary for developing, implementing, and monitoring policies. Yet data are needed to obtain information on functioning in the population, to identify the population with disabilities, and to disaggregate indicators of well-being by disability to determine whether people with disabilities are participating in society to the same extent as those without disabilities. In this article we discuss the need for data harmonization to improve disability research and policy. We describe standard question sets on disability developed for inclusion in surveys and administrative systems, as well as the need for coordination of both statistical and administrative data systems. Until disability data become more harmonized, it will not be possible to support the development of comprehensive, evidence-based policies and programs to address the needs of the population with disabilities.

Linkage of people experiencing homeless using two consent models.

ObjectivesAdministrative data linkage is relatively under-utilised as a way of generating evidence to guide homelessness policy and service delivery in the UK. Our objective is to contribute insight into the ethical, legal, and practical challenges of using data linkage with data from people experiencing homelessness (PEH). ApproachWe outline the data collection and linkage methodologies for two UK-based studies related to PEH. The first design aimed to explore the acceptability and feasibility of consented linkage of trial data (‘Moving On’ trial) to NHS Digital records in a cohort of recruited PEH in two English local authorities (n=50). The second design used administrative data originating from a local authority homelessness service in Wales (n=17,000 cases) to explore educational outcomes of children in homeless households. The resultant data linkage rates are contrasted and discussed in relation to the mechanisms for obtaining and linking personal data. ResultsThe Moving On trial demonstrated high rates of consent for data linkage and the ability to collect sufficient personal identifiable data to increase the chance of successful matching. Aggregate match rates will be discussed. Of the roughly 17,000 cases included in the local authority administrative data, 75% could be linked to unique individuals using probabilistic matching and were therefor ‘useable’ in linkage research. The proportion of useable cases rapidly decreased as the cut-off for matching quality was increased, to roughly 50% of cases being useable when a 99% match probability cut-off was used. Matching rates were higher amongst priority need homeless cases, possibly reflecting business need to identify and work closely with these people. ConclusionWhere homelessness administrative data systems are not designed to enable data linkage, low matching rates can result, reducing study sample sizes and potentially leading to bias towards more extreme cases of homelessness if missed-matches are not random. Consented linkage within large-scale trials offers one possibility for generating long-term evidence.

How can we measure the developmental impacts of the COVID-19 pandemic on young children?

Objectives • Present a conceptual data model for understanding the impacts of COVID on children 0-5 • Outline administrative datasets with relevant indicators that are commonly collected, integrated at the individual level, and include relevant 2-generation linkages • Provide specific considerations for research using the conceptual model ApproachThis session presents findings from a paper that was collaboratively generated by a workgroup of five integrated administrative data systems (IDS) in communities across the US with expansive data holdings in early education. Contributors have decades of experience in utilizing administrative data for social policy planning and analysis. We started with a thorough literature review, used a national survey data of IDS data holdings collected in 2020, and co-created a conceptual data model for administrative data reuse based upon current practices. Specific care was taken to include community-level changes and adult factors that serve as mediators for children’s outcomes. ResultsThe COVID-19 pandemic made its mark on the entire world, yet little is known about children, aged 0-5, whose lives have been exponentially affected. This conceptual data model presents five developmental resilience pathways (i.e. early learning, safe and nurturing families, health, housing, and financial/employment) that include direct and indirect influencers related to COVID-19 impacts, and the contexts and community supports that can affect child-level outcomes. We then overview commonly available administrative datasets with relevant indicators to be linked at the individual and household level, with discussion of data specific considerations (e.g. access, insufficiency, availability, quality, and linkage). The US response creates a natural experiment of policy and implementation decisions across boundaries, and while US-centric, this approach could be applied internationally. ConclusionThe development and use of IDS for research has great potential to inform solutions to our most pressing social problems. Cross-site research studies and maintaining sustainable capacities for long-term research is paramount to understanding the implications of response interventions and to study impacts on child development over time.

Expanding the Understanding of School Concentrated Disadvantage Using Free and Reduced-Price Meals Data: Links to College and Labor Market Outcomes in Maryland

School concentrated disadvantage has been linked to poorer academic achievement and psychosocial functioning in prior research. The current study expands upon prior examinations of school concentrated disadvantage by applying a measurement approach first described by Michelmore and Dynarski in 2017, where eligibility for free and reduced-price meals (FRPM) is examined over time and the duration of eligibility serves as the key indicator of student disadvantage. We used data from a linked longitudinal administrative data system in Maryland, and we measured disadvantage using the proportion of years a student was eligible for FRPM between 6th and 12th grades (see ref. Michelmore & Dynarski). This measure was aggregated to the school level to measure school concentrated disadvantage. We found that school-level concentrated disadvantage was uniquely, and more strongly related to college enrollment than individual student-level disadvantage. However, early labor market outcomes tended to be more strongly linked to race/ethnicity than experiences with disadvantage. We highlight the need for additional targeted resources for students in schools with high concentrations of disadvantaged students.

Epidemiology of cerebral venous sinus thrombosis and cerebral venous sinus thrombosis with thrombocytopenia in the United States, 2018 and 2019

BackgroundPopulation‐based data about cerebral venous sinus thrombosis (CVST) are limited. ObjectivesTo investigate the epidemiology of CVST in the United States. Patients/MethodsThree administrative data systems were analyzed: the 2018 Healthcare Cost and Utilization Project National Inpatient Sample (NIS) the 2019 IBM MarketScan Commercial and Medicare Supplemental Claims Database, and the 2019 IBM MarketScan Multi‐state Medicaid Database. CVST, thrombocytopenia, and numerous comorbidities were identified using the International Classification of Diseases, Tenth Revision, Clinical Modification codes. Incidence rates of CVST and CVST with thrombocytopenia were estimated (per 100,000 total US population [NIS] and per 100,000 population aged 0 to 64 years covered by relevant contributing health plans [MarketScan samples]). Comorbidity prevalence was estimated among CVST cases versus total inpatients in the NIS sample. Recent pregnancy prevalence was estimated for the Commercial sample. ResultsIncidence rates of CVST in NIS, Commercial, and Medicaid samples were 2.85, 2.45, and 3.16, respectively. Incidence rates of CVST with thrombocytopenia were 0.21, 0.22, and 0.16, respectively. In all samples, CVST incidence increased with age; however, peak incidence was reached at younger ages in females than males. Compared with the general inpatient population, persons with CVST had higher prevalences of hemorrhagic stroke, ischemic stroke, other venous thromboembolism (VTE), central nervous system infection, head or neck infection, prior VTE, thrombophilia, malignancy, head injury, hemorrhagic disorder, and connective tissue disorders. Women aged 18 to 49 years with CVST had a higher pregnancy prevalence than the same‐aged general population. ConclusionsOur findings provide recent and comprehensive data on the epidemiology of CVST and CVST with thrombocytopenia.

Developmental impacts of the COVID-19 pandemic on young children: a conceptual model for research with integrated administrative data systems.

The COVID-19 pandemic made its mark on the entire world, upending economies, shifting work and education, and exposing deeply rooted inequities. A particularly vulnerable, yet less studied population includes our youngest children, ages zero to five, whose proximal and distal contexts have been exponentially affected with unknown impacts on health, education, and social-emotional well-being. Integrated administrative data systems could be important tools for understanding these impacts. This article has three aims to guide research on the impacts of COVID-19 for this critical population using integrated data systems (IDS). First, it presents a conceptual data model informed by developmental-ecological theory and epidemiological frameworks to study young children. This data model presents five developmental resilience pathways (i.e. early learning, safe and nurturing families, health, housing, and financial/employment) that include direct and indirect influencers related to COVID-19 impacts and the contexts and community supports that can affect outcomes. Second, the article outlines administrative datasets with relevant indicators that are commonly collected, could be integrated at the individual level, and include relevant linkages between children and families to facilitate research using the conceptual data model. Third, this paper provides specific considerations for research using the conceptual data model that acknowledge the highly-localised political response to COVID-19 in the US. It concludes with a call to action for the population data science community to use and expand IDS capacities to better understand the intermediate and long-term impacts of this pandemic on young children.

Inequalities and short-term outcome among patients assessed as non-urgent in a Swedish ambulance service setting

BackgroundWithin the ambulance service, assessment and referral of patients, especially those with non-urgent conditions, is a difficult and complicated task. Studies indicate that 12 to 20 percent of all patients are subjected to non-conveyance and discharged at the scene. There is lack of knowledge of what characterizes conveyed and non-conveyed patients. The aim of this study was to explore non-urgent patients who are conveyed or not conveyed to hospital and the short-term outcome of non-conveyance in a Swedish Ambulance Service setting. MethodsThis study has a descriptive, cross-sectional design. All patients who were prioritized as non-urgent were eligible for the study and 1,048 patients were followed-up in an administrative data system that stores information about the patients’ trajectory in both primary and hospital care. ResultsMore women than men were subjected to non-conveyance and most of the non-conveyed patients were left at home out-of-hours. 53% sought care again within 72 h. A large proportion of the non-conveyed patients were assessed as having unspecific symptoms. ConclusionsThere are prominent gender differences in the context of non-conveyance where unspecific symptoms seem to be the main reason for being left at home. As many of the non-conveyed patients who did not receive any advice about further investigation or intervention sought care again within 72 h, the assessments may be insufficient or inaccurate.