A Study to Evaluate the Effectiveness and Safety of Prephase Steroid Treatment before Remission Induction Chemotherapy in Patients with Pediatric Acute Lymphoblastic Leukemia Using Common Data Model-Based Real-World Data: A Retrospective Observational Study.

Background: Rapid reduction of leukemic cells in the bone marrow during remission induction chemotherapy (RIC) can lead to significant complications such as tumor lysis syndrome (TLS). We investigated whether prephase steroid treatment before RIC could decrease TLS incidence and improve overall survival in pediatric patients with acute lymphoblastic leukemia (ALL). Methods: Data were extracted from the Common Data Model databases in two tertiary-care hospitals in Seoul, South Korea. Patients were classified into the treated or untreated group if they had received RIC with prephase steroid treatment ≥7 days before RIC in 2012-2021 or not, respectively. Stabilized Inverse Probability of Treatment Weighting (sIPTW) was applied to ensure compatibility between the treated and untreated groups. The incidence of TLS within 14 days of starting RIC, overall survival (OS), and the incidence of adverse events of special interest were the primary endpoints. Multiple sensitivity analyses were performed. Results: Baseline characteristics were effectively balanced between the treated (n=308.4) and untreated (n=246.6) groups after sIPTW. Prephase steroid treatment was associated with a significant 88% reduction in the risk of TLS (OR 0.12, 95% CI: 0.03-0.41). OS was numerically greater in the treated group than in the untreated group although the difference was not statistically significant (HR 0.64, 95% CI 0.25-1.64). The treated group experienced significantly elevated risks for hyperbilirubinemia and hyperglycemia. The reduction in TLS risk by prephase steroid treatment was maintained in all of the sensitivity analyses. Conclusion: Prephase steroid treatment for ≥7 days before RIC in pediatric patients with ALL reduces the risk of TLS, while careful monitoring for toxicities is necessary. If adequately analyzed, real-world data can provide crucial effectiveness and safety information for proper management of pediatric patients with ALL, for whom prospective randomized studies may be difficult to perform for ethical and practical reasons.

Text Extraction and Standardization System Development for Pathological Records in the Korea Biobank Network.

In Korea, the Korea Centers for Disease Control and Prevention operates the Korea BioBank Network (KBN). KBN has pathological records that collected in Korea and it is useful dataset for research. In this study, we established system that time efficient and reduced error by step-by-step data extraction process from KBN pathological records. We tested the extraction process by 769 lung cancer cohorts and 1292 breast cancer cohorts and accuracy is 91%. We expect this system can be used to efficiently process data from multiple institutions, including Korea BioBank Network.

Development of Integrated Data Quality Management System for Observational Medical Outcomes Partnership Common Data Model.

The amount of research on the gathering and handling of healthcare data keeps growing. To support multi-center research, numerous institutions have sought to create a common data model (CDM). However, data quality issues continue to be a major obstacle in the development of CDM. To address these limitations, a data quality assessment system was created based on the representative data model OMOP CDM v5.3.1. Additionally, 2,433 advanced evaluation rules were created and incorporated into the system by mapping the rules of existing OMOP CDM quality assessment systems. The data quality of six hospitals was verified using the developed system and an overall error rate of 0.197% was confirmed. Finally, we proposed a plan for high-quality data generation and the evaluation of multi-center CDM quality.

Data Quality Assessment for Observational Medical Outcomes Partnership Common Data Model of Multi-Center.

The amount of research on the gathering and handling of healthcare data keeps growing. To support multi-center research, numerous institutions have sought to create a common data model (CDM). However, data quality issues continue to be a major obstacle in the development of CDM. To address these limitations, a data quality assessment system was created based on the representative data model OMOP CDM v5.3.1. Additionally, 2,433 advanced evaluation rules were created and incorporated into the system by mapping the rules of existing OMOP CDM quality assessment systems. The data quality of six hospitals was verified using the developed system and an overall error rate of 0.197% was confirmed. Finally, we proposed a plan for high-quality data generation and the evaluation of multi-center CDM quality.

Text Extraction and Standardization System Development for Pathological Records in the Korea Biobank Network.

In Korea, the Korea Centers for Disease Control and Prevention operates the Korea BioBank Network (KBN). KBN has pathological records that collected in Korea and it is useful dataset for research. In this study, we established system that time efficient and reduced error by step-by-step data extraction process from KBN pathological records. We tested the extraction process by 769 lung cancer cohorts and 1292 breast cancer cohorts and accuracy is 91%. We expect this system can be used to efficiently process data from multiple institutions, including Korea BioBank Network.

Study of the factors influencing the use of MyData platform based on personal health record data sharing system.

BACKGROUND: The application of telemedicine and electronic health (eHealth) technology has grown in importance during the COVID-19 pandemic, and a new approach in personal data management and processing MyData, has emerged. Data portability and informational self-determination are fundamental concepts of MyData. This study analysed the factors that influence acceptance of the MyData platform, which, reflects the right to self-determine personal data. METHODS: The study involved participants having experience using the MyData platform, and the key factors of the unified theory of acceptance and use of technology were used in the research model (performance expectancy, effort expectancy, social influence, facilitation condition and behavioural intention to use). The questionnaire comprided 27 items, and system usage log data were used to confirm that behavioural intention to use affected actual use behaviour through structural equation modeling. RESULTS: In total, 1153 participants completed the survey. The goodness of fit in the structural equation model indices indicates that the data fit the research model well. Performance expectancy, social influence, and facilitating conditions had direct effects on behavioural intention to use. We used system usage log data to confirm that behavioural intention to use positively affected actual use behaviour. The impact of the main factors in the unified theory of acceptance and use of technology was not moderated by age or gender, except for performance expectancy. CONCLUSIONS: This study is the first to examine the factors influencing the use of the MyData platform based on the personal health record data sharing system in Korea. In addition, the study confirmed the use behaviour of the MyData platform utilising the system's actual usage log for each function and analysing the effect of the intention of use on actual use. Our study serves as a significant foundation for the acceptance of data portability and sharing concepts. It also lays the foundation for expanding the data economy and ecosystem in the pandemic era.

missForest with feature selection using binary particle swarm optimization improves the imputation accuracy of continuous data.

BACKGROUND: Missing data are a common problem in large-scale datasets and its appropriate handling is crucial for data analyses. Missingness can be categorized as (1) missing completely at random (MCAR), (2) missing at random (MAR), and (3) missing not at random (MNAR). Different missingness mechanisms require different imputation strategies. Multiple imputation, an approach for averaging outcomes across multiple imputed data, is more suitable than single imputation for dealing with various missing mechanisms. missForest, a nonparametric missing value imputation strategy using random forest, is one of the most prevalent multiple imputation methods for missing-data because it can be applied to mixed-type data and does not require distributional assumptions. However, a recent study found that missForest can produce biased results for non-normal data. In addition, missForest is computationally expensive. OBJECTIVE: Therefore, we aimed to further develop the missForest algorithm by combining a binary particle swarm optimization (BPSO)-based feature-selection strategy. METHODS: The BPSO is an evolutionary algorithm that is well known for global optimization and computational efficiency. By using the BPSO-based feature selection step prior to imputing missing values with missForest, the imputation accuracy for continuous variables could be increased by pruning redundant variables. RESULTS: In this study, missForest with BPSO (BPSOmf) showed better imputation accuracy than missForest alone with respect to continuous variables by feature selection prior to the imputation step. CONCLUSIONS: BPSOmf is an appropriate and robust method when the imputation target data consist mainly of continuous variables.