Combination Of Data Sources Research Articles

Past-year mental health issues: discrepant survey & administrative data

Abstract Introduction Accurately measuring mental health disorder prevalence is crucial for public health planning. Administrative and survey data are common methods, but often yield differing results. This study compares case identification of mental health issues across these data sources in Slovenia. Methods We linked the 2019 European Health Interview Survey (EHIS) with three Slovenian health administrative databases: the National Hospital Health Care Statistics Database, Outpatient Prescription Drugs Database, and Absence From Work Database. Case identification utilized self-report of any mental health issue in the past 12 months and healthcare utilization records in 12 months preceding participation in EHIS. Multinomial logistic regression was used to examine the association between socio-demographic factors and case identification across data sources. Results Significant differences were found in 12 month prevalence of any mental health issue estimates between data sources. Only 45.9% of self-reported cases were identified in administrative data, and 36.6% of administrative cases self-reported mental health issues (Kappa = 0.302). Socio-demographics, including age, gender, education, and employment status, were significantly associated with the likelihood of identification in specific data sources. Discussion Our findings underscore the importance of data source and assessment tool selection on mental health prevalence estimates. Variations in identification across population subgroups could reflect differences in healthcare access and self-disclosure biases. Linking data sources and considering their inherent limitations is crucial for accurate burden estimation. Key messages • Data source and assessment methods significantly impact how we identify individuals with mental health issues in the past year. • Combining data sources and understanding their limitations is essential for improving mental health burden estimation.

Assessing the legal, illegal, and gray ornamental trade of the critically endangered helmeted hornbill.

Monitoring wildlife trade dynamics is an important initial step for conservation action and demand reduction campaigns to reduce illegal wildlife trade. Studies often rely on one data source to assess a species' trade, such as seizures or the Convention on International Trade in Endangered Species of Wild Fauna and Flora (CITES) trade data. Each database provides useful information but is often incomplete. Combining information from multiple sources helps provide a more complete understanding of trade. A recent rapid increase in demand for helmeted hornbill (Rhinoplax vigil) casques (a brightly colored, solid keratinous rostrum) led to its uplisting to critically endangered on the International Union for Conservation of Nature Red List in 2015. However, there is little current information on what factors influence trade trends and what current levels of demand are. We combined data from CITES, seizure records, and previously underused, yet abundant, art and antique auction data to examine the global trade in helmeted hornbill casque products (HHPs). Three decades of auction data revealed that 1027 individual HHPs had been auctioned since 1992; total auction sales were over US$3 million from 1992 to 2021. The number of HHPs auctioned was greatest from 2011 to 2014, just after the global art boom (2009-2011), followed by a decline in volume and price. The auction data also revealed 2 possible markets for HHPs: true antique and speculative, defined by era, price, and trade patterns. Trends in illegal trade matched those of the auction market, but legal trade remained consistently low. Combining data sources from legal, illegal, and gray markets provided an overview of the dynamics of illegal trade in an endangered species. This approach can be applied to other wildlife markets to provide a more complete understanding of trade and demand at the market level to inform future demand reduction campaigns.

A unified approach to long-term population monitoring of grizzly bears in the Greater Yellowstone Ecosystem

Long-term wildlife research and monitoring programs strive to maintain consistent data collections and analytical methods. Incorporating new techniques is important but can render data sets incongruent and limit their potential to discern trends in demographic parameters. Integrated population models (IPMs) can address these limitations by combining data sources that may span different periods into a unified statistical framework while providing a holistic view of population dynamics. We developed an IPM in a Bayesian framework for grizzly bears (Ursus arctos) in the Greater Yellowstone Ecosystem. We coupled demographic data with multiple, independent population count data to link annual changes in abundance with vital rates over 4 decades (1983–2023). Abundance increased threefold from an estimated 270 individuals in 1984 to 1030 individuals in 2023. Parameter estimates indicated survival of bears ≥2 years of age was high, contributing to robust population growth during the 1980s (λ = 1.023 [50 % interquartile range = 0.993–1.082]) and 1990s (λ = 1.064 [1.023–1.103]). A slowing of population growth started around 2000 (2000s: λ = 1.030 [0.989–1.068]) and continued into the 2010s (λ = 1.021 [0.985–1.057]), due primarily to reductions in survival of bears <2 years of age. These findings corroborate previous research that identified density-dependent effects as a likely cause. The IPM framework provided greater certainty and understanding regarding the dynamic demographic characteristics of the population and serves as a powerful monitoring tool for this long-lived species. Implementation of the IPM allows timely dissemination of demographic data to help inform adaptive management strategies and policy decisions necessary for the continued management and conservation of this population. This robust and flexible monitoring system allows scientists to investigate the effects of a changing ecosystem on population dynamics, incorporate new data sources and statistical models, and respond to changes in monitoring needs for the population. We highlight the efficacy of the IPM in estimating and tracking demographic parameters for a long-lived species, while accommodating shifts in monitoring techniques and data collections typical of long-term wildlife conservation programs worldwide.

Response to ‘Improving abundance estimation by combining capture–recapture and presence–absence data: Example with a large carnivore’

Abstract A key area of research in ecological statistics involves combining data sources from multiple streams to improve population estimates. One such model attempts to integrate capture–recapture and presence–absence data to estimate the population size of Eurasian lynx in the Jura Mountains, eastern France. This model has been observed to underestimate population sizes. We conducted an extensive simulation study to evaluate the model's performance. We describe our methods for generation of simultaneous capture–recapture and presence–absence data and demonstrate that the model is flawed. Finally, we give an outline of our hypothesis on why the model underestimates population sizes.

Harnessing the potential of shared data in a secure, inclusive, and resilient manner via multi-key homomorphic encryption

In this manuscript, we develop a multi-party framework tailored for multiple data contributors seeking machine learning insights from combined data sources. Grounded in statistical learning principles, we introduce the Multi-Key Homomorphic Encryption Logistic Regression (MK-HELR) algorithm, designed to execute logistic regression on encrypted multi-party data. Given that models built on aggregated datasets often demonstrate superior generalization capabilities, our approach offers data contributors the collective strength of shared data while ensuring their original data remains private due to encryption. Apart from facilitating logistic regression on combined encrypted data from diverse sources, this algorithm creates a collaborative learning environment with dynamic membership. Notably, it can seamlessly incorporate new participants during the learning process, addressing the key limitation of prior methods that demanded a predetermined number of contributors to be set before the learning process begins. This flexibility is crucial in real-world scenarios, accommodating varying data contribution timelines and unanticipated fluctuations in participant numbers, due to additions and departures. Using the AI4I public predictive maintenance dataset, we demonstrate the MK-HELR algorithm, setting the stage for further research in secure, dynamic, and collaborative multi-party learning scenarios.

A multi-source change detection algorithm supporting user customization and near real-time deforestation detections

The abundance of free and accessible satellite data has revolutionized our ability to study deforestation with remote sensing. Recent advances have enabled us to monitor deforestation in near real-time (NRT), and a number of operational NRT alert systems using both optical and synthetic aperture radar (SAR) data have been developed. Yet despite their success, there are three primary issues with current systems. First, alerts often require multiple observations to confirm deforestation, which does not help enforcement if deforestation events are short-lived. Second, the methods can be computationally intensive and often focus on presenting binary detection alerts. Third, results are often reported as static accuracies or temporal lags, with relatively little discussion about how either could be improved for different use-cases. We therefore propose a novel, multisource NRT monitoring algorithm that addresses these issues by quickly achieving high-accuracy detections, using simple data harmonization to combine data sources while yielding probabilities of deforestation, and can be customized to meet management goals where users can explicitly change the algorithm to prioritize latency or accuracy in their detections based on their chosen data inputs. To demonstrate the algorithm's potential, we used combined Landsat-8, Sentinel-2, and Sentinel-1 SAR data in northern Myanmar to monitor areas with contrasting forest types. When testing all possible parameterizations, we achieved comparable results to previous studies, with median false positive rates and false negative rates under 10% and 25%, respectively, and median F1 scores consistently above 75% across the sensor combinations. Median post-data acquisition detection times were faster than other studies and ranged from 2 to 8 days, though the majority were between 2 and 5 days. In attempting to optimize the algorithm for our demonstrated region, we discovered there was no optimal parameterization that simultaneously provided the best accuracy and detection. In addition, we found that including Sentinel-1 data did not improve latency or accuracy, most likely due to relatively high levels of noise in our monsoonal study system that experienced high seasonal precipitation variation. We finally chose an example parameterization to create daily landscape maps of deforestation at 10 m resolution in two regions within the study area (∼60,000 ha). We assert this new, customizable method with the quantification of accuracy and latency trade-offs represents a clear advance for NRT change detection and can improve current operational systems.

Enhancing detection of SARS-CoV-2 re-infections using longitudinal sero-monitoring: demonstration of a methodology in a cohort of people experiencing homelessness in Toronto, Canada

BackgroundAccurate estimation of SARS-CoV-2 re-infection is crucial to understanding the connection between infection burden and adverse outcomes. However, relying solely on PCR testing results in underreporting. We present a novel approach that includes longitudinal serologic data, and compared it against testing alone among people experiencing homelessness.MethodsWe recruited 736 individuals experiencing homelessness in Toronto, Canada, between June and September 2021. Participants completed surveys and provided saliva and blood serology samples every three months over 12 months of follow-up. Re-infections were defined as: positive PCR or rapid antigen test (RAT) results > 90 days after initial infection; new serologic evidence of infection among individuals with previous infection who sero-reverted; or increases in anti-nucleocapsid in seropositive individuals whose levels had begun to decrease.ResultsAmong 381 participants at risk, we detected 37 re-infections through PCR/RAT and 98 re-infections through longitudinal serology. The comprehensive method identified 37.4 re-infection events per 100 person-years, more than four-fold more than the rate detected through PCR/RAT alone (9.0 events/100 person-years). Almost all test-confirmed re-infections (85%) were also detectable by longitudinal serology.ConclusionsLongitudinal serology significantly enhances the detection of SARS-CoV-2 re-infections. Our findings underscore the importance and value of combining data sources for effective research and public health surveillance.

The CLARION study: first report on safety findings in patients newly initiating treatment with cladribine tablets or fingolimod for multiple sclerosis

Objectives: As part of the CLARION study: (1) characterize the incidence of severe infections, herpes zoster, and malignancies in patients newly initiating cladribine or fingolimod for relapsing multiple sclerosis (MS); (2) estimate the incidence of severe lymphopenia among cladribine users; and (3) describe prior/subsequent disease-modifying therapy (DMT) in both cohorts. Methods: Patients were identified from seven participating MS registries/data sources. The incidence rate (IR) of each outcome per 1000 patient-years and its 95% confidence interval (95%CI) were estimated for cohorts using Poisson regression. Results: By cut-off date (01-April-2020), 742 cladribine and 867 fingolimod users were included. Mean follow-up was ∼1 year. The IR for severe infections from all contributing sources (except Denmark) was: cladribine, 7.37 (2.76,19.6); fingolimod, 6.55 (2.46,17.4). The corresponding IR for herpes zoster was 5.51 (1.78,17.1) and 3.27 (0.82,13.1), respectively, while values for opportunistic infections were 0 (0,6.76) and 1.63 (0.23,11.6), respectively. There were no events of progressive multifocal leukoencephalopathy in either cohort. The IR of severe lymphopenia was 63.9 (40.7,100.1) in 349 cladribine users from contributing sources. The IR of malignancies (cut-off date 01-April-2022) was 3.55 (1.59,7.90) for the cladribine cohort (N = 1035) and 3.55 (1.48,8.52) for the fingolimod cohort (N = 843) from three MS registries/data sources. In the combined data sources, 36.8% of cladribine and 27.4% of fingolimod users were DMT-naïve; after initiation of study treatment, 2.5% and 20.2% switched to another DMT, respectively. Conclusion: No new safety signal was observed in patients treated with cladribine tablets, although results are limited by a relatively short duration of follow-up.

Využití administrativních zdrojů dat při vymezení obyvatelstva ve sčítání lidu 2021 v Česku

The article presents the administrative data sources used in the 2021 population census in Czechia and describes the main steps in data processing leading from raw input data to the final census population. Special attention is paid to signs-of-life analysis, which deals with the issue of overcoverage in the population register. The article ties in with a previous article (published in 2022) that discussed the use of registers in European population censuses and explained the process of combining data sources in the previous 2011 Czech population census.

The Virtual Community of Practice Facilitation Model

This study examines the instructional design, learning experiences, and outcomes of a virtual community of practice (VCoP). In 2019, the Northern Ontario School of Medicine launched a continuing professional development program consisting of an asynchronous online module followed by an optional series of facilitated case-based videoconference workshops, designed as a VCoP. This program evaluation study employed a convergent parallel mixed methods design and combined data sources from participant pre- and post-program surveys and reflections with a content analysis of semi-structured interviews. The paper reports key enablers that contributed to the following outcomes: the value of an online module as a baseline of knowledge; the impact of the shared case studies, experiences, and peer support on reflection and modifications to medical practice; and skill development and patient-centered care as a result of module and VCoP participation. A model for the effective design and delivery of VcoPs is proposed that results in acquisition of new knowledge and skills and promotes patient-centred practice.

Hypospadias prevalence by severity and associated risk factors: A population-based active surveillance analysis

Hypospadias is one of the most common congenital anomalies. Trends of hypospadias prevalence by severity are not well reported. Most prevalence studies consist of pooled data from different countries or states suffer from low data quality due to inconsistent methodologies, limited variables, and lack of categorization of hypospadias severity. The objective of this study is to examine the prevalence of hypospadias by degrees of severity and associated risk factors using combined data sources from a stable and well-defined population. The study population includes infants born with hypospadias to mothers residing in Arkansas from 1997 to 2016. Cases were identified from an active population-based surveillance program of birth defects. Identified hypospadias cases from surveillance data were linked to birth certificate and to a clinical database. These two data sources provide more details on the location of the defect and maternal and infant characteristics. The prevalence and 95% confidence intervals were calculated using total male live births as denominator. Chi-square test was used to assess the association of nominal variables. Logistic regression was used to calculate adjusted odds ratio. A total 3230 hypospadias cases were identified from 1997 to 2016. The overall prevalence is 83.0 per 10,000 male births. A majority of cases (56.7%) were classified as first degree with the others having second degree (22.8%), third degree (4.7%) or not otherwise specified (15.6%). The highest prevalence of hypospadias was observed among Non-Hispanic whites. Higher prevalence also was observed among mothers in the older age group with no prenatal care in the first trimester and with gestational hypertension or diabetes. Premature or small for gestational age infants tend to have higher prevalence across all levels of severity. The number of hypospadias cases increased over time. After maternal age, race and education were adjusted, higher risk persisted for infants of restricted fetal growth, mothers with gestational hypertension or diabetes and the cohort of 2013-2016. There is an increase of hypospadias cases in Arkansas. Several maternal and infant characteristics associated with higher prevalence for all levels of severity are worth further investigation.

Completeness of endocrine therapy information in the Primary Care Prescription Database (PCPD) and secondary care treatment datasets: A national population-based cohort study using routine healthcare data

BackgroundEndocrine therapy (ET) is a widely used treatment for breast cancer. In the UK, use is typically initiated in secondary care, with subsequent treatment in primary care. Evaluating use of ET depends on data sources containing accurate and complete information. This study aimed to evaluate the completeness and consistency of ET recorded in primary and secondary care data (SCD) and determine the value of combining data sources in describing use of ET. MethodsThis cohort study included women (50 + years) diagnosed with hormone receptor-positive invasive breast cancer in England, April-2015 to December-2019. Concordance of ET recorded in SCD and the Primary Care Prescription Database (PCPD) was evaluated. Factors associated with recording of ET in each setting were assessed using statistical models. ResultsOverall 110,529 women were included. 94% had ET recorded in either SCD or PCPD. ET captured in SCD varied from 3% (in Systemic Anti-Cancer Therapy data) to 52% (in the Cancer Outcomes and Services Dataset; COSD). By contrast, 93% of patients had an ET prescription in PCPD. Among patients with ET recorded, this was not captured in COSD for 45%. Capture in COSD was lowest for younger women, those with no comorbidity/frailty, with lower stage or HER2-positive disease, or with other treatments recorded. Overall concordance between COSD and PCPD was 57%, but varied substantially across NHS trusts (lowest decile≤28%; highest decile≥86%). Among women with ET recorded in both settings, the earliest record was in COSD for 97%; 59% of initial ET prescriptions recorded in COSD were not captured in PCPD. Combining PCPD and COSD data enabled estimation of ET duration. ConclusionsPCPD is vital for understanding the use of ET within this population. Completeness of SCD could be improved by ensuring information on first ET prescription is recorded. PCPD (linked to SCD) is a valuable resource for examining patterns of care for patients with cancer, including treatment duration and adherence.

Epidemiology of Robin sequence in the UK and Ireland: an active surveillance study

BackgroundBirth prevalence of Robin sequence (RS) is commonly reported as 1 case per 8000–14 000 live births. These estimates are based on single-source case ascertainment and may miss infants who...

Bringing micro to the macro: how citizen science data enrich geospatial visualizations to advance health equity

ABSTRACT Social and spatial contexts affect health, and understanding nuances of context is key to informing successful interventions for health equity. Layering mixed methods and mixed scale data sources to visualize patterns of health outcomes facilitates analysis of both broad trends and person-level experiences across time and space. We used micro-scale citizen scientist-collected data from four Bay Area communities along with aggregate epidemiologic and population-level data sets to illustrate barriers to, and facilitators of, physical activity in low-income aging adults. These data integrations highlight the synergistic value added by combining data sources, and what might be missed by relying on either a micro- or macro-level data source alone. Mixed methods and granularity data integration can generate a deeper understanding of environmental context, which in turn can inform more relevant and attainable community, advocacy, and policy improvements.

Quality framework for combining survey, administrative and big data for official statistics

Creating statistics by combining data sources allows for the production of new, more timely and/or more detailed statistics. With an intended statistical output in mind, and various potentially useful data sources, there is a need to assess the potential of each source to contribute to the intended statistic. Quality frameworks provide tools for such tasks. This paper proposes a quality framework that includes dimensions applicable to survey, administrative and big data to support the assessment of the potential of each source to contribute to the intended statistic. The framework is applied to a case study of mobility data and a case study of virus particle detection in sewage data.

Identifying individuals with complex and long-term health-care needs using the Johns Hopkins Adjusted Clinical Groups System: A comparison of data from primary and specialist health care.

This study aimed to present the Johns Hopkins Adjusted Clinical Groups (ACG) System risk stratification profile of a total adult population of somatic health-care users when using data from either general practitioners (GPs) or hospital services and to compare the number and characteristics of individuals identified as having complex and long-term health-care needs in each data source. This was a registry-based study that included all adult residents (N=168,285) in four municipalities in Central Norway who received somatic health care during 2013. Risk profiles were generated using the ACG System based on age, sex and diagnoses registered by GPs or the local hospital. ACG output variables on number of chronic conditions, frailty and concurrent resource utilisation were chosen as indicators of complexity. Nearly nine out of 10 (83.9%) of the population had been in contact with a GP, and 35.4% with the hospital. The mean number of diagnoses (3.0) was equal in both sources. A larger proportion of the population had higher risk scores in all variables except frailty when comparing hospital data to GP data. This was also found when comparing individuals identified as having complex and long-term health-care needs. A similar proportion of the population was found to have complex and long-term health-care needs (hospital 6.7%, GP 6.3%), but only one in five (21.5%) were identified in both data sets. As data from GPs and hospitals identified mostly different individuals with complex and long-term health-care needs, combining data sources is likely to be the best option for identifying those most in need of special attention.

Development of a Machine Learning Model to Estimate US Firearm Homicides in Near Real Time

Firearm homicides are a major public health concern; lack of timely mortality data presents considerable challenges to effective response. Near real-time data sources offer potential for more timely estimation of firearm homicides. To estimate near real-time burden of weekly and annual firearm homicides in the US. In this prognostic study, anonymous, longitudinal time series data were obtained from multiple data sources, including Google and YouTube search trends related to firearms (2014-2019), emergency department visits for firearm injuries (National Syndromic Surveillance Program, 2014-2019), emergency medical service activations for firearm-related injuries (biospatial, 2014-2019), and National Domestic Violence Hotline contacts flagged with the keyword firearm (2016-2019). Data analysis was performed from September 2021 to September 2022. Weekly estimates of US firearm homicides were calculated using a 2-phase pipeline, first fitting optimal machine learning models for each data stream and then combining the best individual models into a stacked ensemble model. Model accuracy was assessed by comparing predictions of firearm homicides in 2019 to actual firearm homicides identified by National Vital Statistics System death certificates. Results were also compared with a SARIMA (seasonal autoregressive integrated moving average) model, a common method to forecast injury mortality. Both individual and ensemble models yielded highly accurate estimates of firearm homicides. Individual models' mean error for weekly estimates of firearm homicides (root mean square error) varied from 24.95 for emergency department visits to 31.29 for SARIMA forecasting. Ensemble models combining data sources had lower weekly mean error and higher annual accuracy than individual data sources: the all-source ensemble model had a weekly root mean square error of 24.46 deaths and full-year accuracy of 99.74%, predicting the total number of firearm homicides in 2019 within 38 deaths for the entire year (compared with 95.48% accuracy and 652 deaths for the SARIMA model). The model decreased the time lag of reporting weekly firearm homicides from 7 to 8 months to approximately 6 weeks. In this prognostic study of diverse secondary data on machine learning, ensemble modeling produced accurate near real-time estimates of weekly and annual firearm homicides and substantially decreased data source time lags. Ensemble model forecasts can accelerate public health practitioners' and policy makers' ability to respond to unanticipated shifts in firearm homicides.

Standardizing fishery-dependent catch-rate information across gears and data collection programs for Alaska sablefish (Anoplopoma fimbria)

Abstract Indices of abundance used to inform stock assessment models are commonly derived from fishery-dependent data sources. However, fishery catch-per-unit-effort (CPUE) are often confounded by a myriad of factors for which corrections must be made using model-based standardization methods. The Alaska sablefish (Anoplopoma fimbria) fishery provides a fitting case study of such issues, wherein a regulatory change in 2017 disrupted historic fishery dynamics, promoting a rapid transition in use of pot gear over demersal hook-and-line gear in the Gulf of Alaska. To address this, we combined across both observer and logbook programs (data sources) and gear types to develop an intercalibrated abundance index. We first regressed observer records against vessel logbooks to understand potential biases that may arise from combining data sources during the CPUE standardization process. Here, we found that both data sources exhibited strong agreement in reported CPUEs when compared on a set-by-set basis. Therefore, we intercalibrated both CPUE data sources and developed an index of abundance that incorporated catch records from both demersal hook-and-line and pot gear fisheries for sablefish in Alaska, to account for the recent rapid change in gear use. This standardized index of abundance compared favourably with an index generated from a fishery-independent hook-and-line survey currently used in management, suggesting it is representative of sablefish population trends. Our findings not only represent a valuable contribution to the management of sablefish in Alaska, but also provide a widely applicable framework for standardizing fishery-dependent CPUE data to support the management of multi-gear fisheries.

Genetic Validation of Psoriasis Phenotyping in UK Biobank Supports the Utility of Self-Reported Data and Composite Definitions for Large Genetic and Epidemiological Studies

Genetic Validation of Psoriasis Phenotyping in UK Biobank Supports the Utility of Self-Reported Data and Composite Definitions for Large Genetic and Epidemiological Studies

Analysis of types of and language used in online information available to patients with periodontitis.

Introduction This research investigates framing in online patient information for those newly diagnosed with periodontitis.Methods This study is a cross-sectional analysis of websites using corpus linguistic techniques. A Google search was conducted with the term 'gum disease.' Ten pages of search results were reviewed and information available was separated into types of resource: retail, healthcare, and dental practice websites. The dataset was analysed in terms of word frequency, collocation and keyness as compared to the British National Corpus Written Sampler. Differences between sources were assessed.Results Across combined data sources, there was a tendency for the most advanced symptoms of periodontitis to be given prominence. There was also a negative skew towards avoidance of negative outcomes of treatment rather than achieving positive ones. When comparing types of resource, retail websites tended to be more positive, with a focus on improving 'milder' stages of disease.Conclusions Negative framing could potentially induce engagement with treatment and self-care by the process of 'fear-appeal'; however, there is a risk that negativity demotivates an already anxious patient. Further research is required to evaluate patient perceptions of the information and to investigate effects this could have on behaviour change.