Karthikeya Rekulapalli, a data architect at Midland Memorial Hospital
Photo: Karthikeya Rekulapalli
Although there have been significant advancements in healthcare interoperability over the past 10 years, the actual situation for information sharing is still far more fragmented than policy milestones may indicate.
The adoption of FHIR as a contemporary data exchange standard has accelerated, and the 21st Century Cures Act and the creation of the Trusted Exchange Framework and Common Agreement have produced significant regulatory scaffolding.
The Office of the National Coordinator for Health IT reports participation in health information exchange has increased steadily and the vast majority of non-federal acute care hospitals currently use certified EHR systems. These are noteworthy accomplishments.
Beneath these top-line figures, however, legacy protocols continue to play a major role in the daily exchange of clinical data.
The majority of real-time data transfer between hospital systems is still powered by HL7 v2 messaging, a standard that was established in the late 1980s, says Karthikeya Rekulapalli, a data architect at Midland Memorial Hospital and consultant at consulting firm Cardinal Solutions Tech.
Midland Memorial Hospital, part of Texas' Midland Health, is a not-for-profit institution serving the Permian Basin region of West Texas.
Millions of messages are processed by integration engines every day, and when set up properly, they are incredibly efficient, said Rekulapalli.
"However, this infrastructure is delicate: A single case-sensitivity error in a file name, a mismatched patient identifier, or a version mismatch during a system upgrade can cause entire workflows to stop processing messages," he explained.
"To catch what the interfaces miss, many hospitals still rely on manual tracking and reconciliation. In reality, the integration layer is more brittle than most non-health IT professionals are aware, often leading to significant disruptions in patient care and data management when errors occur.
"Newer capabilities are also developing at the same time," he continued. "Barcode scanning in pharmacies and supply chains has enhanced point-of-care data collection, while patient-facing portals and apps have increased record access. However, there is still a large gap between what is operationally deployed and what is technically feasible."
The shift to modern APIs and real-time data access is uneven, moving more quickly at large academic centers while community hospitals and rural systems frequently lag behind, he added.
"Many health systems are operating on legacy infrastructure, which necessitates considerable downtime for version upgrades," he said. "All things considered, the state of interoperability is one of real advancements coexisting with enduring, deeply ingrained operational fragility."
Healthcare IT News spoke with Rekulapalli to discuss interoperability in healthcare, and what can be done to improve it.
Q. Why do you consider interoperability still often to be just a buzzword in healthcare?
A. In the healthcare industry, the term "interoperability" is widely used in policy documents, vendor marketing and conference keynotes – but the patient experience frequently paints a very different picture. Take a straightforward example: Whether you have a receipt or not, the system instantly recognizes your transaction when you return an item at a large retailer.
However, when a patient moves from one healthcare facility to another, they are still frequently required to carry physical printouts of their records, repeat their medication history from memory or wait while staff members manually ask a previous provider for information. Despite years of investment and industry promises, Graham Grieve, the founder of FHIR, has publicly discussed this disconnect – the suffering a patient experiences when navigating across systems that do not genuinely communicate.
The economics of data exchange is one of the main reasons interoperability is still more of an ideal than a reality. Accessing and transferring data between EHR platforms frequently entails significant licensing fees, custom development costs and continuous maintenance overhead – despite the existence of common frameworks and standards.
Health systems often have to bargain with vendors for access to data they already own in order to pull their data, create their own clinical applications or establish connections with outside organizations. True, frictionless data sharing is frequently impractical due to contractual and financial barriers, even though the infrastructure for exchange may be technically available. This is especially true for smaller and mid-sized health systems with limited IT budgets.
Q. Why do physicians still face fragmented data?
A. Physicians today often make clinical decisions without a complete picture of their patients' histories, and this issue arises from structural problems rather than individual failures. When patients transition between care settings – such as moving from a primary care office to a specialist or from an emergency department to an outpatient imaging center – there is no guarantee their full medical records follow them.
Important documents like radiology reports, medication lists, surgical notes and allergy histories are frequently locked within the systems where they were originally recorded. If a patient switches health systems or visits a provider outside their network, the continuity of their records heavily relies on whether the patient remembers to bring the necessary documentation or if someone from the clinical team will manually request it.
The user interface layer further complicates this challenge. Physicians face significant time constraints, and their workflow may not always inform the design of the tools they use to access patient information. While electronic health record systems have become more sophisticated over the years, each update or redesign introduces new navigation patterns that clinicians must learn.
This task can cause even experienced physicians to struggle with quickly finding the information they need. The cognitive burden of navigating complex, often unintuitive software – while simultaneously managing patient interactions – adds to documentation fatigue and can lead to crucial information being overlooked.
Physicians ultimately experience fragmentation due to the storage, sharing and governance of data within the healthcare ecosystem. Without a universal mechanism that enables a patient's complete longitudinal record to follow them wherever they seek care, clinicians will continue to operate with incomplete information.
This is not merely an inconvenience; it has serious implications for clinical accuracy, patient safety and the overall quality of care. To address this issue, the industry needs to rethink data ownership and patient-centered access, as well as enhance technology.
Q. What can CIOs and other health IT leaders at hospitals and health systems do to overcome these various challenges to advance closer to what you would call true interoperability?
A. The most impactful step health IT leaders can take is to conduct an honest, rigorous assessment of their current technology landscape and eliminate redundant or underutilized systems. Many hospitals have accumulated layers of legacy applications over the years – point systems added to address specific needs that were never fully integrated or retired when better alternatives became available.
This complexity creates maintenance overhead, increases security risk and makes true interoperability harder to achieve. CIOs who focus on having a clear and up-to-date technology system – based on open standards and allowing easy data sharing – will be much better able to work with outside partners and provide a smooth experience for both doctors and patients.
Equally important is investing in robust data governance, AI-assisted documentation and workflow optimization. Artificial intelligence is rapidly maturing in its ability to support clinical documentation, automate coding, and surface relevant patient information at the point of care. Health IT leaders who embrace these tools can reduce the manual burden that heavily contributes to fragmentation, as long as they ensure responsible implementation with clinician input.
At the same time, making the scheduling and registration process easier for routine appointments frees up time and space for what matters most: handling complex and emergency cases quickly and completely, as patient safety requires.
Furthermore, to achieve true interoperability without exorbitant expenses, health IT leaders should embrace modern data architectures that utilize supported open-source technologies. Deploying platforms like Apache Druid for real-time analytics and Apache Superset for advanced data visualization – whether hosted natively or distributed across a cross-cloud environment – enables health systems to process and analyze massive volumes of interoperable data instantly.
This agile strategy supports highly specific clinical and operational use cases with low latency while maintaining a lean, cost-effective infrastructure. By using open-source tools that are supported by businesses in a cross-cloud setup, CIOs can prevent being tied to a single vendor, keep their systems reliable, and provide real-time, useful information right where it's needed.
Follow Bill's health IT coverage on LinkedIn: Bill Siwicki
Email him: bsiwicki@himss.org
Healthcare IT News is a HIMSS Media publication.
WATCH NOW: Price transparency is vital for lowering healthcare costs
