Centralization of Medical Data: New Opportunities and Implementation Challenges

Currently, in 2017, digital medicine is undergoing a breakthrough. We are moving from an era of local automation (where each hospital purchased a server and installed its own isolated program) to building a national healthcare ecosystem. Local databases lead to patients undergoing the same tests twice because hospitals cannot see each other’s data, and managers lack real statistics on morbidity.

Healthcare urgently needs integrated, managed, and secure centralized registries. The team at IQusion IT LLC is already designing architectural solutions today that not only solve the problem of data storage but also lay the technological foundation for healthcare reform in the coming years.

Information Consolidation: Architecture Instead of a “File Dump”

A centralized database is not just a large server where text documents flow. It is a complex microservice or service-oriented architecture (SOA) that ensures traceability of changes and consistency of medical history regardless of where the patient received the service.

What we lay at the foundation of our solutions:

• Hybrid Data Storage: relational databases (e.g., PostgreSQL) for financial transactions and strict accounting, and NoSQL solutions for flexible storage of unstructured medical records.
• Integration Bus (ESB): configuring data exchange between hundreds of different medical information systems (MIS) and the central component, routing millions of requests per day without performance loss.
• Data-driven Management: analytical modules that transform raw data into BI dashboards for management (real workload on doctors, patient routes, medication accounting).

Interoperability: The End of the “Format Zoo” Era

The biggest technical barrier today is data heterogeneity. In one system, a diagnosis is recorded as text; in another, it’s a code from an outdated local directory. For data to become functional, we implement strict standardization.

IQusion IT Technology Stack: we design data exchange based on the modern standard HL7 FHIR. This allows medical events to be transmitted as structured resources. For exchanging medical images (MRI, CT, X-ray), we implement DICOM protocols.

Special attention is given to master data directories (MDM — Master Data Management). We build systems that enforce data validation against international classifiers: МКХ-10 (ICD-10) for secondary care and ICPC-2 for primary healthcare. This ensures that a clinical case in Kyiv and Lviv will be described in a machine-readable, identical language.

Reliability for a Critical Industry: Banking Sector Level SLA

A medical system has no right to downtime. When an emergency doctor requests a patient’s allergy history, the system must respond in milliseconds.

• Horizontal Scaling and Load Balancing: our architecture is ready for peak loads (e.g., during mass signing of declarations with doctors).
• Fault Tolerance (RPO/RTO): configuring synchronous and asynchronous data replication between geographically dispersed data centers so that the loss of one node does not stop the system.

Data Security: Presumption of Distrust (Zero Trust)

Medical data is the most sensitive information. We move away from the concept of simple “perimeter defense” and implement deep control at the application level.

• Role-Based Access Control (RBAC): a family doctor sees the full picture, a specialized professional sees only relevant episodes, and a statistician receives exclusively depersonalized data.
• Cryptographic Protection: encryption of data both in-transit (TLS) and at-rest. Mandatory use of a doctor’s electronic digital signature (EDS) for validating each medical record.
• Total Audit: immutable event logging. The system records who, when, and from which IP address viewed a patient’s card, preventing unnoticed data leakage.

Forecast for 2018–2020: Architecture That Will Withstand the Future

Analyzing the current market state and laying down the architecture today, we at IQusion IT clearly understand the challenges eHealth will face in the next three years:

• 2018 — Scaling of Primary Care: a mass launch of electronic declarations signing between patients and doctors will occur. The load on central databases will grow exponentially; only systems built on microservices and ready for cloud scaling will survive.
• 2019 — Electronic Prescriptions and Referrals: the focus will shift to transactional accuracy. The introduction of e-prescriptions will require ideal integration between hospital MIS, pharmacy networks, and the central registry. Data will become the basis for the real movement of funds (“money follows the patient”).
• 2020 — Crisis Testing and Telemedicine: we predict a sharp jump in the need for remote interaction and real-time epidemiological monitoring. The system will have to withstand colossal unpredictable loads, and it is then that data standardization will allow for the rapid deployment of new modules (e.g., for tracking infectious diseases or mass vaccination) in a matter of weeks.

Looking Ahead

In 2017, we no longer have the right to build local, incompatible IT solutions. Centralization and standardization of medical data (according to FHIR, DICOM, ICPC-2 standards) is the only path to transparent and effective medicine.

IQusion IT’s solution is an architecture being developed today but designed for the realities of the 2020s. We are building a robust foundation capable of integrating thousands of institutions, protecting the data of millions of patients, and providing the state with a real management tool.

Preparing for integration with the national eHealth system? Contact our engineers, and we will prepare a detailed audit of your infrastructure and a plan for transitioning to a modern medical data exchange bus.