Data Without a Diagnosis: How Fragmented Health Records Are Failing American Patients

She had been tracking her symptoms for eight months. Her fitness tracker logged her resting heart rate, her sleep fragmentation, and her daily step count. A separate app catalogued her food intake and flagged potential dietary triggers. A third platform, provided by her insurance company, stored the results from her last two blood draws. Her telehealth provider had notes from three virtual visits. Her in-person internist had notes from two more.

None of these systems spoke to one another. When she finally saw a specialist — after a four-month wait — she arrived with screenshots on her phone and a handwritten summary she had prepared herself. The specialist, working from an incomplete picture, ordered tests that had already been run.

This is not an unusual story. It is, for a significant proportion of Americans engaged with digital health tools, a recognizable one.

The Illusion of Connected Care

The term "connected care" implies a coherent, integrated system in which health information flows seamlessly between providers, platforms, and patients. The reality of American digital health infrastructure in 2024 is considerably more fragmented. Wearables generate proprietary data stored in manufacturer clouds. Telehealth platforms maintain their own electronic records. Hospital systems run on legacy EHR platforms — Epic, Cerner, Meditech — that interoperate imperfectly even with one another, let alone with consumer wellness applications.

The 21st Century Cures Act, enacted in 2016 and substantially expanded through subsequent federal rulemaking, established interoperability requirements intended to address this fragmentation. The legislation mandated that certified health IT systems support standardized application programming interfaces — specifically the HL7 FHIR standard — to enable data sharing between platforms. Progress has been made. The problem has not been solved.

Large health systems have generally complied with the letter of federal requirements. Consumer wellness platforms, which are not classified as health IT under current regulatory frameworks, are largely exempt from interoperability mandates. The result is a two-tier data ecosystem: a regulated tier in which clinical records can, in theory, be exchanged, and an unregulated tier in which the vast majority of consumer health data sits in siloed proprietary systems.

The Patient as Data Integrator

In the absence of system-level interoperability, the burden of health data synthesis has increasingly fallen on patients themselves. This is a significant and underappreciated form of health labor — one that is distributed unevenly across the population.

Patients who are health-literate, technologically fluent, and have the time and cognitive bandwidth to manage multiple digital health accounts are able, to some degree, to compensate for system failures. They screenshot their wearable data. They export CSV files from nutrition apps. They print portal records and carry them to appointments. They become, in effect, the integration layer that the technology has not provided.

Patients who lack these resources — those who are older, less educated, managing cognitive impairment, or simply overwhelmed by illness — cannot perform this function. For them, fragmentation is not an inconvenience. It is a clinical hazard.

Research published in the Journal of the American Medical Informatics Association has documented that patients from lower socioeconomic backgrounds are significantly less likely to successfully transfer health records between providers, even when electronic transfer mechanisms exist. The digital divide in health data access mirrors, and in some cases amplifies, existing disparities in health outcomes.

How Fragmentation Causes Clinical Harm

The downstream consequences of data fragmentation are not merely administrative. They are medical.

Duplicate diagnostic testing — the ordering of labs, imaging, or procedures that have already been performed elsewhere — is one of the most quantifiable consequences. The Office of the Inspector General has estimated that duplicate testing contributes billions of dollars annually to unnecessary healthcare expenditure. Beyond cost, repeated testing carries direct patient burden: additional radiation exposure from imaging, venipuncture, time, and the anxiety that accompanies repeated diagnostic uncertainty.

More serious is the risk of diagnostic delay or error. When a clinician lacks access to a patient's complete symptom history, prior test results, or medication record, the clinical picture they are working from is incomplete. Conditions that would be apparent from a longitudinal data review may be missed in a cross-sectional snapshot. Medication interactions that would be flagged by a complete medication list may go unnoticed when that list is distributed across three platforms and two providers.

The diagnostic odyssey — the protracted, often years-long process through which patients with complex or rare conditions eventually receive accurate diagnoses — is in part a fragmentation problem. Data that could accelerate pattern recognition exists. It simply does not exist in one place, in one format, accessible to one physician at the moment of clinical decision-making.

Wearables and the Unintegrated Evidence Base

Consumer wearables represent perhaps the most striking example of clinically relevant data that exists outside the clinical record. Devices capable of detecting atrial fibrillation, tracking blood oxygen saturation, measuring continuous glucose levels, and logging detailed sleep architecture are worn by tens of millions of Americans. The data they generate is, in aggregate, an extraordinary epidemiological resource. At the individual level, it is largely invisible to the healthcare system.

Some health systems have begun piloting integrations that pull wearable data into EHR platforms — Apple Health Records, for instance, supports connections with a growing number of hospital systems through the FHIR standard. But these integrations remain the exception rather than the rule, and even where they exist, the clinical workflow for reviewing and acting on wearable data is underdeveloped.

Physicians, already managing documentation burdens that contribute to widespread burnout, are rarely equipped with the time or tools to review months of step-count data or sleep graphs within a standard appointment. The data arrives without clinical contextualization, without AI-assisted pattern recognition, and without a structured workflow for integration into diagnostic reasoning.

What Genuine Connectivity Requires

Addressing health data fragmentation requires coordinated action across regulatory, technical, and clinical domains.

On the regulatory side, extending interoperability requirements to consumer health platforms — currently outside the scope of HIPAA and federal health IT certification — would be a meaningful step, though politically complex given industry opposition. Strengthening enforcement of existing information-blocking prohibitions, which have been applied inconsistently since their introduction under the Cures Act, would improve data portability within the regulated sector.

On the technical side, the development of patient-controlled health data aggregation tools — platforms that allow individuals to consolidate records from multiple sources into a single portable health record — represents a promising architectural approach. Apple's Health app and Google Health have made partial progress in this direction, but neither has achieved the comprehensive integration that genuine clinical utility requires.

On the clinical side, healthcare organizations need structured protocols for incorporating patient-generated data into diagnostic and treatment workflows. This means training, workflow redesign, and the development of AI-assisted tools capable of surfacing clinically relevant patterns from large, heterogeneous datasets.

The Standard Worth Holding

Connected care, as a concept, demands more than digital access to healthcare services. It requires that the information generated across a patient's entire health journey — from wearable devices, from telehealth visits, from in-person encounters, from laboratory results — be available, coherent, and actionable at the moment a clinical decision is being made.

By that standard, the American healthcare system's digital infrastructure remains substantially incomplete. The data exists. The connections, too often, do not. And it is patients — particularly those with the least capacity to compensate — who bear the cost of that gap.