Common Data Model Research Articles

Mapping UK Pain Datasets to the OMOP Common Data Model: Lessons Learned

ObjectiveChronic Pain is defined as pain lasting for three months or longer and a report by the British Pain Society in 2016 found that 28 million adults live with this condition. Chronic pain research is problematic as it is poorly represented in health data and often in silos. Alleviate is the Advanced Pain Discovery Platform (APDP) Data Hub which is removing barriers to access to data. ApproachThe aim for Alleviate is to transform pain data sets within the UK to be Findable, Accessible, Interoperable and Reusable (FAIR) and be more discoverable for researchers via Health Data Research UK’s Cohort Discovery Tool. The tool enables identification of potential cohorts in real-time from datasets Alleviate provides. The onboarding requires datasets to be transformed to the open and widely used Observational Medical Outcomes Partnership (OMOP) Common Data Model (CDM). ResultsWe collaboratively developed CaRROT tools and used our mapping expertise to transform datasets to OMOP-CDM. The tools have improved efficient mapping of clinical and research data, covering ~10 million individuals’ records. Subsequently we learned valuable lessons on how to accurately and consistently map across data sets to support federated discovery. We also identified a lack of standard vocabulary representations for pain specific data. ConclusionsMapping data to OMOP is not straightforward. Attention to detail is required to ensure that consistent mapping decisions are made to maximise data reusability and interoperability. ImplicationsIn sharing the lessons learned of mapping datasets to OMOP we aim to help improve interoperability of disparate data sources.

Evaluation of a Multi-Component Phenotype Algorithm for Systemic Lupus Erythematosus across the PCORnet, OMOP, and i2b2 Common Data Models

The aim of this study was to evaluate the performance of a computable phenotype for systemic lupus erythematosus (SLE) patients when it is ported from a local data warehouse to the i2b2, OMOP, and PCORnet CDMs. We adapted the SLE phenotype to the Northwestern Medicine (NM) Enterprise Data Warehouse (EDW) and NU i2b2, OMOP, and PCORnet instances. Each of the phenotype’s 17 components were determined by a rules-based algorithm built on diagnosis, medication, lab, and procedure codes. We assessed the phenotype over 168 clinician-confirmed SLE patients and 100 healthy controls and calculated agreement between the EDW and CDMs for overall SLE classification, individual SLICC criteria membership, and individual code occurrence level aggregates - count, first and last date, lab values using Cohen’s kappa (Cκ) and intraclass correlation coefficients (ICC). For overall SLE classification, agreement of the OMOP and PCORnet datamarts with the EDW was high (Cκ 0.928, 0.802, respectively) while that of i2b2 was low (Cκ 0.328). For the panel of SLICC criteria, OMOP had high agreement for 14/17 SLICC criteria (Cκ 0.792-1.000). PCORnet had high agreement for 13/17 criteria (Cκ 0.708-1.000). i2b2 had high agreement for 10/17 criteria (Cκ 0.813-1.000) with the exception of lab-based criteria. While agreement between the EDW and CDMs at the level of individual codes was relatively poor, agreement at the levels of overall SLE classification and individual SLICC criteria were reasonably high for OMOP and PCORnet, suggesting that inconsistencies at a micro, code level become less apparent at the macro, phenotype level.

Providing data analytic services for national knowledge users in a federated system: learnings from two Canadian use cases

Objective and ApproachCanada’s federated health data system along with access pathways geared toward academic research, pose challenges for knowledge Users (KUs) requiring timely, pan-Canadian evidence to inform decisions. To understand challenges and explore solutions, we conducted two use cases for a pan-Canadian health technology assessment organization and a federal agency, using health administrative data at one federal and six provincial data centres. The population-based cohort studies described socio-demographics, comorbidities, treatment patterns, service utilization and costs. One focused on spinal muscular atrophy, a rare disease and the other on dementia, a complex chronic disease. ResultsAdministrative challenges included aligning varied ethical review and data access policies/procedures including requiring local and/or academic principal investigators and differing definitions of “research” vs. planning, evaluation, and monitoring. Data-related challenges included differences in structure, timeliness, and completeness across regions resulting in difficulty aligning constructs such as incident cases and episodes of care. Privacy requirements prohibited pooling jurisdictional estimates resulting in “small cells” that couldn’t be shared with KUs. ConclusionsWe provided analytic outputs from multiple regions, albeit with some differences and gaps, increasing knowledge around both diseases while developing capacity for combined analyses and gaining insight into national possibilities for data access. ImplicationsAs decision makers must rely on best available information, some data is better than none. However, to improve data-analytic services for Pan-Canadian KUs, next steps will include improving data harmonization, expanding data assets and filling data gaps, implementing common data models, and exploring options for federated and/or pooled analyses.

Limitations of NHIC claim code-based surveillance and the necessity of UDI implementation in Korea

The E-Health Big Data Evidence Innovation Network (FeederNet) in Korea, based on the observational medical outcomes partnership (OMOP) common data model (CDM), had 72.3% participation from tertiary hospitals handling severe diseases as of October 2022. While this contributes to the activation of multi-institutional research, concerns about the comprehensiveness of device data persist due to the adoption of national health insurance corporation (NHIC) claim codes as device identifiers in the medical device field. This study critically evaluated the effectiveness and compatibility of NHIC claim codes and unique device identifier (UDI) within FeederNet to identify the optimal identifier for efficient Post-market surveillance (PMS). Specifically, this study addressed three main questions: (1) the number of UDIs classified as NHIC-covered items, (2) the number of UDIs included in each NHIC claim code, and (3) the number of NHIC claim codes each UDI covers. Among the 1,979,655 UDIs registered domestically, only 36.02% (712,983) were classified as covered by National Health Insurance. NHIC-covered medical devices were limited to categories (A) medical devices, (B) medical supplies, and (C) dental materials, excluding most software and in vitro diagnostics (IVD). Multiple UDIs could be registered under a single NHIC claim code, and a single UDI could be registered under multiple NHIC claim codes. Only 32.62% (13,756/42,171) of NHIC claim codes had registered UDIs, with an average of 53 UDIs per claim code. Of the UDIs listed as NHIC covered, 92.39% (659,046/713,341) had one claim code, while 7.25% (51,652) had multiple claim codes. Additionally, 2643 UDIs were listed as NHIC covered but had no registered claim codes. Due to this complex relationship, NHIC claim code-based PMS may pool safe and unsafe models or disperse problematic models across multiple claim codes, leading to a lower problem rate or insignificant differences between claim codes, thus reducing signal detection sensitivity compared to UDI-based PMS. In conclusion, NHIC claim code-based PMS has limitations in granularity and signal detection sensitivity, necessitating the adoption of UDI-based PMS to address these issues. The UDI system can enhance the accuracy of medical device identification and tracking, playing a crucial role in generating real-world evidence (RWE) by integrating data from various sources. Future research should explore specific strategies for integrating and utilizing UDI with NHIC claim codes, contributing to the implementation of a more reliable and comprehensive PMS in Korea’s healthcare system.

THE INTERFACE BETWEEN THE BIOCEANIC ROAD AND THE BELT AND ROAD INITIATIVE (BRI) FROM A BRAZILIAN PERSPECTIVE ON CRIMINAL ORGANIZATIONS

The current changing in world dynamics leads to an increased complexity on countries relations. From this perspective, the article's general objective analyses the impact of Latin America and East Asia integration through the Bioceanic Corridor and the Belt and Road Initiative on illicit markets. The specific objectives focus on identifying complicating elements for regional security, stemming from the collaborative and competitive engagements of actors within illegal networks. The adopted methodology is predominantly qualitative, involving bibliographic review, document analysis and case studies based on data from academic and government documents, and reports from international organizations. The addressed topics include drug and weapons trafficking, infrastructure initiatives, regional and global interconnection, and their relations to the criminal phenomena in South America. From the results obtained, the article emphasizes the anticipatory dimension, providing preventive recommendations in international collaboration, standardization of informational capabilities through common data models in road logistics hubs, and the application of security protocols in maritime hubs, with emphasis on the ARESP methodology (Risk Analysis with Emphasis on Port Security), contributing to regional integration on global security, and proposing practical and collaborative measures to anticipate and mitigate potential challenges arising from the interconnection of illicit networks on a global scale.

Abstract P365: Prevalence of Hypertension in Minnesota Adults in 2023 Using Data from a Statewide Electronic Health Record Consortium

Background: Data on prevalence of hypertension in the United States come primarily from in-person data collected by the National Health and Nutrition Examination Survey and self-reports in the Behavioral Risk Factor Surveillance Survey. Electronic phenotypic algorithms for hypertension that include diagnosis codes, blood pressure values, and antihypertensive treatment have been proposed and tested. Objective: To describe prevalence of hypertension in Minnesota adults using data from a statewide electronic health record (EHR) consortium, the Minnesota EHR Consortium (MNEHRC). Methods: Data from 11 health systems participating in the MNEHRC were standardized to a common data model allowing for distributed querying of EHR data. Data were de-duplicated using a one-way hash algorithm. Preliminary prevalence of hypertension in 2023 was estimated in Minnesota residents with ≥ 1 encounter in the past 3 years who had ≥1 diagnosis of hypertension in the past 5 years. Medicaid insurance status was determined via linkage with state enrollment files. Residential addresses of patients were geocoded to census tracts and social vulnerability index (SVI) of census tracts was determined. Results: Among 4,369,470 Minnesotan age 18 and older in 2023, 23.6% had hypertension based on ≥1 diagnosis code. Prevalence increased with age, was higher in men (26% vs. 22% in women), and adjusted for age and sex was highest in Black (30%) and Native American (27%) compared with White (23%) and Hispanic (21%) Minnesotans. Within each race/ethnic group, those insured by Medicaid had a higher adjusted hypertension prevalence than those with other insurance types. The prevalence of hypertension in census tracts with moderate to high SVI ranged from 3% to 43%, but on average was similar to the statewide prevalence of 24% (see image 1). Prevalence of hypertension in race/ethnicity, sex, and age subgroups in higher SVI census tracts was similar to the statewide prevalence in those subgroups. Conclusions: The prevalence of hypertension in Minnesota was described in detail by demographic characteristics, geography and social vulnerability. Future work will expand the electronic hypertension phenotype and examine quality of care.

Extracting lung cancer staging descriptors from pathology reports: A generative language model approach

BackgroundIn oncology, electronic health records contain textual key information for the diagnosis, staging, and treatment planning of patients with cancer. However, text data processing requires a lot of time and effort, which limits the utilization of these data. Recent advances in natural language processing (NLP) technology, including large language models, can be applied to cancer research. Particularly, extracting the information required for the pathological stage from surgical pathology reports can be utilized to update cancer staging according to the latest cancer staging guidelines. ObjectivesThis study has two main objectives. The first objective is to evaluate the performance of extracting information from text-based surgical pathology reports and determining pathological stages based on the extracted information using fine-tuned generative language models (GLMs) for patients with lung cancer. The second objective is to determine the feasibility of utilizing relatively small GLMs for information extraction in a resource-constrained computing environment. MethodsLung cancer surgical pathology reports were collected from the Common Data Model database of Seoul National University Bundang Hospital (SNUBH), a tertiary hospital in Korea. We selected 42 descriptors necessary for tumor-node (TN) classification based on these reports and created a gold standard with validation by two clinical experts. The pathology reports and gold standard were used to generate prompt-response pairs for training and evaluating GLMs which then were used to extract information required for staging from pathology reports. ResultsWe evaluated the information extraction performance of six trained models as well as their performance in TN classification using the extracted information. The Deductive Mistral-7B model, which was pre-trained with the deductive dataset, showed the best performance overall, with an exact match ratio of 92.24% in the information extraction problem and an accuracy of 0.9876 (predicting T and N classification concurrently) in classification. ConclusionThis study demonstrated that training GLMs with deductive datasets can improve information extraction performance, and GLMs with a relatively small number of parameters at approximately seven billion can achieve high performance in this problem. The proposed GLM-based information extraction method is expected to be useful in clinical decision-making support, lung cancer staging and research.

Open Access Data Repository and Common Data Model for Pulse Oximeter Performance Data.

The OpenOximetry Repository is a structured database storing clinical and lab pulse oximetry data, serving as a centralized repository and data model for pulse oximetry initiatives. It supports measurements of arterial oxygen saturation (SaO2) by arterial blood gas co-oximetry and pulse oximetry (SpO2), alongside processed and unprocessed photoplethysmography (PPG) data and other metadata. This includes skin color measurements, finger diameter, vital signs (e.g., arterial blood pressure, end-tidal carbon dioxide), and arterial blood gas parameters (e.g., acid-base balance, hemoglobin concentration). Data contributions are encouraged. All data, from desaturation studies to clinical trials, are collected prospectively to ensure accuracy. A common data model and standardized protocols for consistent archival and interpretation ensure consistent data archival and interpretation. The dataset aims to facilitate research on pulse oximeter performance across diverse human characteristics, addressing performance issues and promoting accurate pulse oximeters. The initial release includes controlled lab desaturation studies (CLDS), with ongoing updates planned as further data from clinical trials and CLDS become available.

Financial Benefits of Renal Dose-Adjusted Dipeptidyl Peptidase-4 Inhibitors for Patients with Type 2 Diabetes and Chronic Kidney Disease.

Dipeptidyl peptidase-4 (DPP4) inhibitors are frequently prescribed for patients with type 2 diabetes; however, their cost can pose a significant barrier for those with impaired kidney function. This study aimed to estimate the economic benefits of substituting non-renal dose-adjusted (NRDA) DPP4 inhibitors with renal dose-adjusted (RDA) DPP4 inhibitors in patients with both impaired kidney function and type 2 diabetes. This retrospective cohort study was conducted from January 1, 2012 to December 31, 2018, using data obtained from common data models of five medical centers in Korea. Model 1 applied the prescription pattern of participants with preserved kidney function to those with impaired kidney function. In contrast, model 2 replaced all NRDA DPP4 inhibitors with RDA DPP4 inhibitors, adjusting the doses of RDA DPP4 inhibitors based on individual kidney function. The primary outcome was the cost difference between the two models. In total, 67,964,996 prescription records were analyzed. NRDA DPP4 inhibitors were more frequently prescribed to patients with impaired kidney function than in those with preserved kidney function (25.7%, 51.3%, 64.3%, and 71.6% in patients with estimated glomerular filtration rates [eGFRs] of ≥60, <60, <45, and <30 mL/min/1.73 m2, respectively). When model 1 was applied, the cost savings per year were 7.6% for eGFR <60 mL/min/1.73 m2 and 30.4% for eGFR <30 mL/min/1.73 m2. According to model 2, 15.4% to 51.2% per year could be saved depending on kidney impairment severity. Adjusting the doses of RDA DPP4 inhibitors based on individual kidney function could alleviate the economic burden associated with medical expenses.

Early Clinical Experience of Finerenone in People with Chronic Kidney Disease and Type 2 Diabetes in Japan-A Multi-Cohort Study from the FOUNTAIN (FinerenOne mUltidatabase NeTwork for Evidence generAtIoN) Platform.

Background: In the phase 3 clinical trials FIGARO-DKD and FIDELIO-DKD, finerenone reduced the risk of cardiovascular and kidney events among people with chronic kidney disease (CKD) and type 2 diabetes (T2D). Evidence regarding finerenone use in real-world settings is limited. Methods: A retrospective cohort study (NCT06278207) using two Japanese nationwide hospital-based databases provided by Medical Data Vision (MDV) and Real World Data Co., Ltd. (RWD Co., Kyoto Japan), converted to the OMOP common data model, was conducted. Persons with CKD and T2D initiating finerenone from 1 July 2021, to 30 August 2023, were included. Baseline characteristics were described. The occurrence of hyperkalemia after finerenone initiation was assessed. Results: 1029 new users of finerenone were included (967 from MDV and 62 from RWD Co.). Mean age was 69.5 and 72.4 years with 27.3% and 27.4% being female in the MDV and RWD Co. databases, respectively. Hypertension (92 and 95%), hyperlipidemia (59 and 71%), and congestive heart failure (60 and 66%) were commonly observed comorbidities. At baseline, 80% of persons were prescribed angiotensin-converting-enzyme inhibitors or angiotensin-receptor blockers. Sodium-glucose cotransporter 2 inhibitors and glucagon-like peptide 1 receptor agonists were prescribed in 72% and 30% of the study population, respectively. The incidence proportions of hyperkalemia were 2.16 and 2.70 per 100 persons in the MDV and RWD Co. databases, respectively. There were no hospitalizations associated with hyperkalemia observed in either of the two datasets. Conclusions: For the first time, we report the largest current evidence on the clinical use of finerenone in real-world settings early after the drug authorization in Japan. This early evidence from clinical practice suggests that finerenone is used across comorbidities and comedications.

Association of the protective effect of telmisartan on hearing loss among patients with hypertension.

Hearing loss, affecting a significant portion of the global population, is prevented with peroxisome proliferator-activated receptor γ agonism. Understanding potential protective treatments is crucial for public health. We examine the effect of telmisartan, an antihypertensive drug and partial peroxisome proliferator-activated receptor γ agonist, on hearing loss in patients with hypertension. This retrospective cohort analysis used data from the OMOP Common Data Model database, encompassing information from three tertiary institutions in South Korea. The study included a substantial sample size of 860,103 people diagnosed with hypertension. The study included individuals who had been medically diagnosed with hypertension and had been prescribed antihypertensive drugs, including telmisartan. The study design was established to evaluate the comparative effects of telmisartan and other hypertension medications on hearing loss. We used propensity score matching (PSM) to create a balanced cohort, reducing potential biases between the telmisartan and non-telmisartan groups. From the initial 860,103 patients with hypertension, a propensity score matched cohort was derived from 20,010 patients, with 2,193 in the telmisartan group. After PSM, lower incidence of total hearing loss was observed in the telmisartan group compared to the non-telmisartan group during the 3-year follow-up (0.5% vs. 1.5%, log-rank p = 0.005). In subgroup analysis, this study showed consistent results that lower incidence of total hearing loss was higher in the telmisartan group than in the non-telmisartan group. Telmisartan was associated with reducing certain types of hearing loss in patients with hypertension. Further research is needed to confirm these findings and understand the mechanisms.

Adoption of the OMOP CDM for Cancer Research using Real-world Data: Current Status and Opportunities.

The Observational Medical Outcomes Partnership (OMOP) common data model (CDM) that is developed and maintained by the Observational Health Data Sciences and Informatics (OHDSI) community supports large scale cancer research by enabling distributed network analysis. As the number of studies using the OMOP CDM for cancer research increases, there is a growing need for an overview of the scope of cancer research that relies on the OMOP CDM ecosystem. In this study, we present a comprehensive review of the adoption of the OMOP CDM for cancer research and offer some insights on opportunities in leveraging the OMOP CDM ecosystem for advancing cancer research. Published literature databases were searched to retrieve OMOP CDM and cancer-related English language articles published between January 2010 and December 2023. A charting form was developed for two main themes, i.e., clinically focused data analysis studies and infrastructure development studies in the cancer domain. In total, 50 unique articles were included, with 30 for the data analysis theme and 23 for the infrastructure theme, with 3 articles belonging to both themes. The topics covered by the existing body of research was depicted. Through depicting the status quo of research efforts to improve or leverage the potential of the OMOP CDM ecosystem for advancing cancer research, we identify challenges and opportunities surrounding data analysis and infrastructure including data quality, advanced analytics methodology adoption, in-depth phenotypic data inclusion through NLP, and multisite evaluation.

OntoBridge Versus Traditional ETL: Enhancing Data Standardization into CDM Formats Using Ontologies Within the DATOS-CAT Project.

Common Data Models (CDMs) enhance data exchange and integration across diverse sources, preserving semantics and context. Transforming local data into CDMs is typically cumbersome and resource-intensive, with limited reusability. This article compares OntoBridge, an ontology-based tool designed to streamline the conversion of local datasets into CDMs, with traditional ETL methods in adopting the OMOP CDM. We examine flexibility and scalability in the management of new data sources, CDM updates, and the adoption of new CDMs. OntoBridge showed greater flexibility in integrating new data sources and adapting to CDM updates. It was also more scalable, facilitating the adoption of various CDMs like i2b2, unlike traditional methods reliant on OMOP-specific tools developed by OHDSI. In summary, while traditional ETL provides a structured approach to data integration, OntoBridge offers a more flexible, scalable, and maintenance-efficient alternative.

Enhancing Healthcare Informatics: Integrating Category Theory Reasoning into OMOP-CDM Ontology Model.

The task of managing diverse electronic health records requires the consolidation of data from different sources to facilitate clinical research and decision-making support, with the emergence of the Observational Medical Outcomes Partnership - Common Data Model (OMOP-CDM) as a standard relational database schema for structuring health records from different sources. Working with ontologies is strongly associated with reasoners. Implementing them over expansive and intricate Ontologies can pose computational challenges, potentially resulting in slow performance. In this paper, we propose the implementation of a new reasoner based on categorical logic over a translation of OMOP-CDM into an ontology model. This enables enhancements to the efficiency and scalability of implementing such models.

Converting Entity-Attribute-Value Data Sources to OMOP's CDM: Lessons Learned.

Clinical data repositories often use entity-attribute-value (EAV) data models. To be valuable for secondary use, these health data can be transformed to the Observational Medical Outcomes Partnership (OMOP) Common Data Model (CDM). The present paper describes the lessons learned from such an endeavour based on the concept of registering transformation functions on source data elements. We further provide future work directions for follow-up projects.

Coverage of physical therapy assessments in the Observational Medical Outcomes Partnership Model common data model.

High-value care aims to enhance meaningful patient outcomes while reducing costs and is accelerated by curating data across healthcare systems through common data models (CDMs), such as Observational Medical Outcomes Partnership Model (OMOP). Meaningful patient outcomes, such as physical function, must be included in these CDMs. However, the extent that physical therapy assessments are covered in the OMOP CDM is unclear. Examine the extent that physical therapy assessments used in neurologic and orthopaedic conditions are in the OMOP CDM. After identifying assessments, two reviewer teams independently mapped the neurologic and orthopaedic assessments into the OMOP CDM. We quantified agreement within the reviewer team by the number of assessments mapped by both reviewers, one reviewer but not the other, or neither reviewer. The reviewer teams then reconciled disagreements, after which we examined agreement and the average number of concept ID numbers per assessment. Of the 81 neurologic assessments, 48.1% (39/81) were initially mapped by both reviewers, 9.9% (8/81) were mapped by one reviewer but not the other, and 42% (34/81) were unmapped. After reconciliation, 46.9% (38/81) were mapped by both reviewers and 53.1% (43/81) were unmapped. Of the 79 orthopaedic assessments, 46.8% (37/79) were initially mapped by both reviewers, 12.7% (10/79) were mapped by one reviewer but not the other, and 48.1% (38/79) were unmapped. After reconciliation, 48.1% (38/79) were mapped by both reviewers and 51.9% (41/79) were unmapped. Most assessments that were mapped had more than one concept ID number (2.2±1.3 and 4.3±4.4 concept IDs per neurologic and orthopaedic assessment, respectively). The OMOP CDM includes some assessments recommended for use in neurologic and orthopaedic conditions, but many have multiple concept IDs. Including more functional assessments in the OMOP CDM and creating guidelines for mapping would improve our ability to include functional data in large datasets.

Real-World Data in Greece: Mapping the Papageorgiou General Hospital Data to the OMOP Common Data Model.

This paper showcases the results of the Extract-Transform-Load process mapping the Electronic Health Record of Papageorgiou General Hospital in Thessaloniki, Greece, to the Observational Medical Outcomes Partnership Common Data Model. We describe the staged process utilized to account for the intricate structure of the database, along with some general findings from the mapping. Finally, we investigate potential directions for future research.

DATOS-CAT: OMOP-Common Data Model for the Standardization, Integration and Analysis of Population-Based Biomedical Data in Catalonia.

Transforming the population based biomedical cohort into the Common Data Model (OMOP-CDM) empowers researchers to access direct sources of information, enabling a deeper understanding of how genetic profiles relate to clinical outcomes and providing new knowledge that can significantly influence health care practices around the world.

Intensive Care Quality Indicators Dashboard Using Observational Medical Outcomes Partnership Common Data Model.

This paper presents the development of a visualization dashboard for quality indicators in intensive care units (ICUs), using the Observational Medical Outcomes Partnership (OMOP) Common Data Model (CDM). The dashboard enables the user to visualize quality indicator data using histograms, pie charts and tables. Our project uses the OMOP CDM, ensuring a seamless implementation of our dashboard across various hospitals. Future directions for our research include expanding the dashboard to incorporate additional quality indicators and evaluating clinicians' feedback on its effectiveness.

Key Research Areas for Building and Deploying a Common Data Model for an Intensive Medicine Data Space in Europe and Beyond.

Key Research Areas (KRAs) were identified to establish a semantic interoperability framework for intensive medicine data in Europe. These include assessing common data model value, ensuring smooth data interoperability, supporting data standardization for efficient dataset use, and defining anonymization requirements to balance data protection and innovation.