Why healthcare data still doesn’t talk

When policymakers, hospital CIOs, and tech vendors talk about the promise of digital health, they often highlight seamless data flow: a clinician in one system pulls up lab results generated across the country, or a patient’s smartwatch data automatically adjusts a care plan. Yet a new systematic review of 161 studies finds that reality remains elusive. Despite decades of work, the global health system is still struggling with three stubborn barriers: interoperability, patient-centered care, and genomic data integration.

The study, published in Frontiers in Health Services in June 2025, explores how even widely promoted standards such as HL7 FHIR and SNOMED CT are adopted inconsistently, leaving health systems vulnerable to delays, redundant testing, and patient frustration. “Semantic misalignment across commonly used healthcare standards continues to block seamless data exchange,” wrote lead author Radha Ambalavanan and colleagues from the Self Research Institute.

The EHR revolution — and its limits

Electronic health record (EHR) adoption has been one of the most dramatic shifts in modern medicine. In the United States, adoption among office-based physicians jumped from 18% in 2001 to nearly 80% by 2021, fueled by billions in federal incentives through the Meaningful Use program. But as the Office of the National Coordinator for Health IT has acknowledged, moving from paper to digital didn’t solve the interoperability problem. Instead, many systems created new silos.

That fragmentation has real-world costs. A 2009 study in the New England Journal of Medicine warned that lack of interoperability would undermine the efficiency gains expected from EHRs. Fifteen years later, the new review finds those concerns still valid, with inconsistent adoption of APIs and data standards hampering cross-system communication.

Patient-centered care that isn’t

The second major theme of the review is patient-centered care (PCC). PCC is supposed to empower patients with shared decision-making and personalized care. But most EHR systems remain provider-centric.

The review found poor integration of patient-generated health data (PGHD), such as information from fitness trackers, home monitoring devices, and mobile apps. Even when patients have access to portals, real-time feedback loops are rare. 

For chronic disease patients — who account for the bulk of U.S. healthcare spending — that gap means fewer opportunities to intervene early. “Without real-time exchange, personalized care adjustments are often delayed,” the review concluded.

Genomics: promise meets privacy

Perhaps the thorniest challenge the researchers identified is genomic data integration. Precision medicine depends on linking phenotypic data in EHRs with genetic profiles, but that linkage raises enormous privacy and security concerns.

The U.S. HIPAA framework and Europe’s GDPR provide partial guardrails, but cross-border data exchange remains a patchwork. A 2021 Cell Genomics policy paper from the Global Alliance for Genomics and Health warned that “inconsistent international standards” limit global research and clinical adoption. The new review echoes that warning, pointing to consent complexity, computational cost, and the risk of re-identification when AI models process large datasets.

Blockchain-based consent models and federated learning — where algorithms learn across decentralized data without moving the raw information — are among the proposed fixes. But as the review notes, few of these have been validated at scale.

The COVID-19 inflection point

The pandemic made clear just how damaging fragmented health data can be. In 2020, public health officials struggled to merge hospital records, lab data, and contact tracing systems into a coherent picture of the crisis. The Centers for Disease Control and Prevention has since acknowledged that lack of standardized reporting slowed response times.

COVID-19 also accelerated some progress: hospitals expanded telehealth and remote monitoring, and federal agencies promoted FHIR-based APIs through the 21st Century Cures Act. But progress has been uneven. A 2023 BMC Medical Informatics review found that even within the same country, interoperability maturity varied sharply between health systems.

Global gaps

The review doesn’t just look at the U.S. and Europe. It highlights work in low- and middle-income countries (LMICs), where modular standards like FHIR have been adopted for their flexibility. In Africa, initiatives like H3Africa and ACEGID are trying to build genomic infrastructure tailored to local populations. Yet barriers remain: underrepresentation in global datasets, lack of population-specific genomic references, and limited capacity for secure data sharing.

Those disparities matter not just for equity but for science: without genomic diversity, precision medicine risks being optimized for only a subset of the world’s population.

What’s next?

The authors recommend three main paths forward:

• Standardization: push for universal adoption of HL7 FHIR, SNOMED CT, and LOINC, reinforced by regulatory mandates.
• AI-driven patient engagement: build mobile-first, user-friendly apps that integrate PGHD into EHRs and provide real-time decision support.
• Secure genomic governance: deploy federated learning, blockchain, and zero-knowledge proofs to balance privacy with utility.
  

As the authors caution, many of the most promising solutions — from ontology-driven data models to NLP-based patient feedback systems — exist mainly in research settings, not clinical practice. Without sustained funding, regulatory clarity, and buy-in from providers, the risk is that the next decade will look much like the last: pockets of innovation, but persistent fragmentation.