The FAIR principles, first articulated in a landmark 2016 paper in Scientific Data, were conceived as a response to a paradox of modern science: the sheer abundance of data coexisting with an inability to put it to work. The principles do not mandate that data be open, but rather that it be described in ways that machines and humans can act upon. In biomedicine, however, turning FAIR into reality has proved notoriously difficult. Hospital datasets are governed by privacy law, stored in legacy systems, and annotated with local coding conventions that resist harmonization. Previous studies, including work by the European Commission expert group on FAIR data and analyses of hospital data during the COVID-19 pandemic, have documented how even well-resourced institutions struggle to make medical datasets genuinely interoperable.

The EHDEN Catalogue emerged from the European Health Data and Evidence Network, a project funded through the Innovative Medicines Initiative 2 Joint Undertaking under grant agreement No 806968. EHDEN’s ambition is to create a federated network of European data partners whose databases share a common analytical language. That language is the OMOP Common Data Model, or CDM, developed by the Observational Health Data Sciences and Informatics community. The OMOP CDM defines a standardized set of tables and relationships for observational health data, so that a prescription record in Lisbon is structured the same way as one in Leiden or Barcelona. Once data are mapped into this model, the same analysis code can run across many institutions, enabling large-scale studies of drug safety, disease outcomes and treatment patterns without moving patient-level records across borders.

Technically, the EHDEN Catalogue is a web application that guides data custodians through the process of profiling and documenting their databases. Institutions register their data sources, describe their provenance and content, and run standardized characterization scripts that extract summary statistics about the underlying OMOP CDM implementation. These summaries, covering everything from vocabulary coverage to the distribution of patient counts and observation periods, are aggregated into machine-readable profiles. The platform builds on established web standards, including the W3C’s Data Catalog vocabulary, which provides a semantic framework for describing datasets, their publishers, licenses and access conditions. The result is a catalogue that is not merely a list of databases but a structured, queryable map of the European observational data landscape.

Each letter of the FAIR acronym maps onto a concrete design decision in the platform. Findability is served by rich metadata and persistent identifiers that make data sources discoverable through search. Accessibility is addressed through clearly documented access procedures rather than unrestricted downloads, respecting the governance constraints that necessarily surround patient data. Interoperability flows from the OMOP CDM itself, which gives heterogeneous hospital systems a shared schema, and from the catalogue’s use of standard vocabularies. Reusability is supported by detailed data quality assessments and characterization reports that tell prospective researchers exactly what a database contains, how complete it is, and whether it suits a given study. The authors emphasize that this combination transforms the platform into a comprehensive solution, offering data custodians and researchers a suite of tools and services rather than a single static registry.

The platform also incorporates analytical capabilities that extend beyond documentation. By connecting to visualization and exploration tools, including the Apache Superset business intelligence framework, the catalogue allows users to interrogate aggregated database profiles and compare institutions before committing to a collaboration. This matters because choosing the right data source is one of the most consequential decisions in observational research. A study of a rare adverse drug reaction, for example, requires databases with sufficient patient numbers, adequate follow-up time and reliable recording of the relevant exposures. The catalogue’s profiling reports make such feasibility assessments possible at a distance, reducing the trial-and-error that has historically characterized multi-database studies.

The work builds on a lineage of earlier systems developed by the same group. The EMIF Catalogue, described in 2019 in the International Journal of Medical Informatics, was a collaborative platform for sharing and reusing biomedical data across European medical informatics projects. Subsequent contributions refined the approach: a methodology for cohort harmonization in multicentre clinical research, a secure architecture for exploring patient-level databases distributed across institutions, and more recently a web platform called MONTRA2 for profiling distributed health databases, along with chatbot-like tools to enhance the discovery of OMOP CDM databases. The EHDEN Catalogue consolidates these threads into a production-grade application deployed across the EHDEN network, which has grown to encompass dozens of European data partners mapping billions of patient records to the common data model.

The timing of this work is significant. The European Health Data Space, a legislative initiative currently reshaping health data governance across the European Union, aims to establish a single market framework for the secondary use of health data. Workshops at the European Journal of Public Health have debated whether Europe is ready to maximize the reuse of health data under this framework, and pilot infrastructures for cross-border data exchange are already being tested. Platforms like the EHDEN Catalogue supply a crucial piece of the puzzle: the technical scaffolding that makes data discoverable and interoperable at a continental scale. Without such infrastructure, legal frameworks alone cannot deliver the promised research benefits, because researchers would still face the practical impossibility of locating and assessing suitable data sources.

The study’s authors report that the EHDEN Catalogue significantly improves the visibility, searchability and reusability of biomedical data across Europe, paving the way for more effective and efficient healthcare research with potential impact on the broader European healthcare landscape. The work received funding from the EU/EFPIA Innovative Medicines Initiative 2 Joint Undertaking, and Silva was supported by the European Commission under grant agreement 101081813 for the Genomic Data Infrastructure project. The authors declare no conflicts of interest, and the research involved no direct handling of patient data, which is consistent with the platform’s federated philosophy: aggregated metadata and characterization summaries travel, while patient-level records stay behind institutional firewalls.

For the wider scientific community, the message is that FAIR compliance is achievable in health data when it is engineered into workflows rather than bolted on afterward. The EHDEN Catalogue demonstrates that a carefully designed application, grounded in a common data model and open metadata standards, can convert fragmented national databases into a coherent research resource. As observational studies increasingly inform regulatory decisions, drug safety surveillance and health technology assessment, the ability to find, evaluate and reuse real-world data will only grow in importance. The Portuguese team’s platform offers a template for how that transformation can happen, one documented, standardized and searchable database at a time.

Subject of Research: FAIR-compliant sharing and management of observational health data through the EHDEN Catalogue platform

Article Title: Sharing observational health data following FAIR principles

Article References: Sharing observational health data following FAIR principles. (n.d.). https://doi.org/10.1007/s11042-026-21974-4

Image Credits: AI Generated

DOI: 10.1007/s11042-026-21974-4

Keywords: FAIR principles, observational health data, EHDEN Catalogue, OMOP Common Data Model, data sharing, data interoperability, health informatics, biomedical data discovery, European Health Data Space, data standardization, electronic health records, University of Aveiro

The concept of a learning health system is deceptively simple. In traditional medicine, research and practice exist in separate worlds: clinical trials produce evidence, guidelines distill that evidence, and clinicians are expected to apply it, but the loop often takes years or decades to close. A learning health system collapses that loop. Electronic health records, registries, wearable devices, and patient-reported outcomes become a living laboratory, and artificial intelligence and machine learning tools sift through the resulting data to surface insights that are rapidly embedded in clinical decision support systems at the point of care. The result is a feedback cycle in which evidence generation, decision-making, and quality improvement happen continuously rather than episodically.

The review highlights why cardiovascular medicine is particularly well suited to this approach. Heart attacks, strokes, and heart failure emergencies are intensely time-sensitive, so systems that can shorten the distance between new evidence and bedside action could save lives directly. At the same time, cardiovascular conditions require decades of longitudinal management, generating exactly the kind of rich, repeated data that machine learning models thrive on. Guideline-directed medical therapies for heart failure and post-myocardial infarction care are well established, yet underuse of these therapies remains widespread and measurable, providing a concrete target for automated quality monitoring. Cardiovascular care also already possesses mature registry infrastructure, including the National Cardiovascular Data Registry, which offers a foundation on which learning loops can be built.

Technically, the review identifies several building blocks that a cardiovascular learning health system requires. Interoperable data infrastructures and common data models, such as the Observational Medical Outcomes Partnership common data model and standards like Fast Healthcare Interoperability Resources, allow information from different hospitals and systems to be combined and analyzed. Natural language processing can extract structured information from clinical notes, while patient-generated health data from wearables and remote monitoring devices extend the data stream beyond the clinic walls. Patient-reported outcomes, measured with validated instruments such as the Kansas City Cardiomyopathy Questionnaire, bring the patient’s own experience of symptoms and function into the analytic loop. Clinical decision support then translates model outputs into actionable prompts for clinicians.

The authors ground their analysis in real-world case studies that demonstrate both the feasibility and the limits of these systems. The Veterans Health Administration has built integrated data platforms and programs such as the Clinical Assessment Reporting and Tracking system for cardiac procedures, showing how a single-payer integrated network can close learning loops at national scale. PCORnet, the National Patient-Centered Clinical Research Network, and the Patient-Centered Outcomes Research Institute have demonstrated how distributed data networks can support comparative effectiveness research across many health systems. Kaiser Permanente’s integrated model has long used its comprehensive electronic health record for population health management and surveillance. The American Heart Association’s Get With The Guidelines program illustrates a registry-driven learning cycle that has measurably improved adherence to evidence-based care in participating hospitals.

Yet the review is candid about why these successes have not simply been replicated everywhere. Fragmented multipayer systems make it difficult to follow patients across care settings, and misaligned financial incentives mean that organizations investing in data infrastructure may not be the ones that reap the savings. The upfront costs of interoperable data platforms, analytics teams, and decision support tools are substantial. Data quality remains heterogeneous across institutions, and machine learning models trained in one population can suffer from algorithmic bias, performing worse in groups that were underrepresented in training data, and from model drift, in which performance degrades over time as populations, practices, and coding habits change. These are not minor technical wrinkles; they are central obstacles that determine whether an algorithm helps or harms.

The human and organizational dimensions receive equally careful attention. Clinician workload is a recurring concern: poorly designed decision support can add alerts and documentation burdens that fuel burnout rather than improve care. Regulatory and liability questions surrounding artificial intelligence, particularly software as a medical device, remain unsettled, leaving hospitals uncertain about who is responsible when an algorithm errs. The review also emphasizes implementation science frameworks, including the Consolidated Framework for Implementation Research, the Practical, Robust Implementation and Sustainability Model, and the RE-AIM framework, as practical tools for designing and evaluating deployment strategies. Community engagement through community-based participatory research is highlighted as essential for ensuring that learning systems earn trust and address the priorities of the populations they serve.

Health equity emerges as one of the most consequential themes of the analysis. Cardiovascular disease already imposes a disproportionate burden on communities facing barriers of access, income, geography, and discrimination, and the review warns that digital technologies could widen these gaps if wearable devices, broadband connectivity, and algorithmically well-served populations are concentrated among the advantaged. The authors call for systematic measurement of health equity as a core function of any cardiovascular learning health system, so that disparities in access, quality, and outcomes are tracked with the same rigor as clinical metrics, and for governance structures that continuously monitor algorithmic performance across demographic subgroups rather than treating a model as finished once it is deployed.

The governance question extends to the lifecycle of the artificial intelligence tools themselves. Because learning systems are designed to change constantly, a model that is safe and accurate at deployment may become unreliable as inputs shift, and regulators accustomed to static, locked medical devices are still developing frameworks for adaptive algorithms. The review points toward the need for continuous post-market surveillance of algorithmic performance, clear liability arrangements, and pragmatic regulatory pathways that preserve innovation without sacrificing patient safety. Sustainable financing models are equally critical: without mechanisms that reward evidence generation and quality improvement rather than only volume of services, learning systems risk remaining grant-funded demonstrations rather than durable infrastructure.

The overall message of the review is one of conditional optimism. Cardiovascular learning health systems have genuine potential to accelerate the translation of evidence into practice, tighten clinical decision-making, and drive continuous quality improvement in the world’s leading cause of death. But the authors are clear that technology alone will not deliver this transformation. Success depends on integrating robust technical infrastructure with effective governance, sustainable financing, systematic equity measurement, ongoing monitoring of algorithmic performance, pragmatic implementation strategies, and rigorous evaluation of patient-centered outcomes across diverse and fragmented care environments. In other words, the hardest problems are not computational but institutional, and solving them will determine whether the promise of a system that learns from every patient becomes routine reality or remains an aspiration confined to a handful of integrated health networks.

Subject of Research: Application of learning health system principles, including artificial intelligence and data infrastructure, to cardiovascular disease prevention and care

Article Title: Learning health systems for cardiovascular health and health care

Article References: Ye, J., Bronstein, S., & Hashish, M. A. (2026). Learning health systems for cardiovascular health and health care. BMC Medicine. https://doi.org/10.1186/s12916-026-05255-3

Image Credits: AI Generated

DOI: 10.1186/s12916-026-05255-3

Keywords: learning health system, cardiovascular disease, artificial intelligence, machine learning, clinical decision support, electronic health records, data interoperability, implementation science, health equity, patient-reported outcomes, wearable devices, health informatics

The theoretical backbone of the study is Organizational Information Processing Theory, a framework first articulated in the 1970s by Jay Galbraith and later refined by Richard Daft and colleagues. The theory holds that organizations are, at their core, information-processing machines. When the uncertainty and ambiguity, or equivocality, of the information they face exceeds their processing capacity, performance suffers. Hospitals are perhaps the most extreme example of this mismatch. A single patient encounter can generate data from laboratory systems, imaging archives, electronic health records, insurance databases, and referral letters, each in different formats and each carrying its own uncertainties. Clinicians must reduce this flood of raw data into a diagnosis, a treatment plan, and a discharge decision, often under time pressure. The authors describe these mismatches as information contingencies: situations where the amount, clarity, or reliability of information does not match what decision-makers need.

To translate this old theory into the AI era, the researchers conceptualize HAIC as a third-order construct, a layered structure in which a single overarching capability emerges from the interaction of two second-order pillars: information processing and information decisiveness. This architectural choice matters methodologically. Rather than treating AI capability as a vague, monolithic asset, the model forces researchers and executives to measure distinct, testable components. Information processing captures how well an organization absorbs, integrates, and makes sense of the data flowing through its systems. Information decisiveness captures how confidently and coherently the organization converts that processed information into timely decisions and coordinated action. In the authors’ framing, an AI system that generates brilliant predictions but leaves clinicians paralyzed by ambiguity has failed on the second pillar, no matter how sophisticated its models are.

Within the information processing pillar, the study identifies two critical drivers: predictive accuracy and data interoperability. Predictive accuracy is the familiar face of AI, the statistical fidelity with which a model forecasts outcomes such as patient length of stay, readmission risk, or disease progression. Interoperability is the less glamorous but arguably more decisive factor. It refers to the ability of AI systems to exchange data seamlessly across platforms, departments, and even institutions. The authors argue that both must advance together. A highly accurate model fed by fragmented, incompatible data sources produces confident nonsense, while perfectly interoperable data flowing into weak models produces noise at scale. Only the combination augments an organization’s genuine capacity to process information, reducing both the uncertainty of not knowing enough and the equivocality of having too many ambiguous interpretations at once.

The information decisiveness pillar rests on two different supports: decision support and organizational collaboration. Decision support refers to AI systems designed to augment human judgment, presenting clinicians with ranked options, risk scores, and explanations rather than opaque outputs. Research on physician interaction with AI has shown that the way recommendations are framed shapes whether doctors adopt, adapt, or ignore them. Organizational collaboration, meanwhile, extends decisiveness beyond the individual clinician. Healthcare delivery is a team sport involving physicians, nurses, administrators, insurers, and external partners. The study finds that when AI capability is paired with structures that encourage cross-functional coordination, the organization becomes more decisive as a whole, converting shared information into unified action rather than siloed hesitation.

The payoff of this capability, according to the study, is concrete and operational. The authors establish a positive association between HAIC and timely, cost-efficient service delivery, which in turn reinforces overall operational effectiveness in healthcare. In practical terms, a hospital with strong HAIC is better positioned to shorten waiting times, allocate beds and staff more intelligently, and reduce the ancillary costs that accumulate when decisions stall. This finding echoes a growing body of operations management research showing that analytics capabilities in hospitals influence cost, quality, and patient satisfaction, but the new study’s contribution is to specify the mechanism: it is the reduction of information contingencies that unlocks the operational gains, not the mere presence of AI tools.

Methodologically, the study is notable for how it handles the messy reality of measuring an organizational capability. The authors ground their construct development in a content-analytic examination of the healthcare AI landscape, drawing on documented evidence from the sector, and then validate the higher-order structure using partial least squares structural equation modeling, a statistical technique well suited to complex, multidimensional constructs. The appendices of the published paper reveal the depth of this work: correlation matrices, cross-loadings to establish discriminant validity, and robustness checks including endogeneity assessments using Gaussian copula approaches, a technique increasingly favored in management research when randomized experiments are impossible. The authors also employed structural topic modeling to analyze textual data from the healthcare AI ecosystem, ensuring that their constructs reflected real-world configurations rather than armchair theorizing.

The practical implications reach hospital executives directly. The study offers AI-enabled configurations, essentially evidence-based recipes, for how to assemble the components of HAIC. An organization strong on predictive accuracy but weak on interoperability should prioritize data integration infrastructure before buying more sophisticated models. An organization with excellent analytics teams but poor cross-departmental collaboration should invest in governance and coordination mechanisms, because decisiveness will remain the bottleneck. The framework also suggests a diagnostic use: by measuring the four sub-dimensions, administrators can identify precisely where their information pipeline leaks value, whether at the point of data capture, in the analytical engine, or in the human decision that follows.

For scholars, the study extends Organizational Information Processing Theory in a direction that has waited decades for the right technology. Classic OIPT research treated information processing capacity as a property of organizational structure, such as formalization, decentralization, and the richness of communication media. The new work shows that AI systems can serve as a fundamentally new information processing mechanism, one that operates at machine speed and scale, but only when embedded in complementary organizational arrangements. This reframing connects the study to the resource-based view of the firm, in which capabilities, not technologies, are the sources of sustained advantage. An AI model can be copied; the organizational capability to exploit it cannot.

The timing of this research is significant. Studies of AI implementation across United States hospitals published in 2026 reveal a landscape of uneven adoption, with many institutions deploying tools without the surrounding capabilities to extract value. Meanwhile, healthcare costs continue to climb globally, and universal health coverage remains out of reach for billions of people, according to World Bank monitoring. Against this backdrop, the message of Bizalwan, Kumar, and Bose is quietly radical: the future of AI in medicine will not be decided by the cleverness of algorithms alone, but by whether healthcare organizations can build the twin disciplines of processing information decisively and deciding on processed information. Hospitals that master both may finally convert the long-promised AI revolution into something patients can feel in waiting rooms and wallets alike.

Subject of Research: Healthcare artificial intelligence capability and its role in reducing information contingencies to improve decision-making and operational effectiveness in hospitals

Article Title: Healthcare Artificial Intelligence Capability: Translating Information Contingencies into Decision Effectiveness and Operational Impact

Article References: Bizalwan, D. S., Kumar, R., & Bose, I. (2026). Healthcare Artificial Intelligence Capability: Translating Information Contingencies into Decision Effectiveness and Operational Impact. Information Systems Frontiers. https://doi.org/10.1007/s10796-026-10820-5

Image Credits: AI Generated

DOI: 10.1007/s10796-026-10820-5

Keywords: healthcare AI, artificial intelligence capability, Organizational Information Processing Theory, information processing, information decisiveness, predictive accuracy, data interoperability, decision support, operational effectiveness, hospital management, health informatics, Information Systems Frontiers