Timely analysis of primary medical data for making objective decisions is one of the key objectives when it comes to consistent healthcare digitalization the industry has seen lately. This objective requires implementing structured electronic medical records and structured electronic medical documents, which constitute the digital method for transferring and storing information in a specific format. Enormous amounts of structured electronic medical records have been accumulated to date at the national and regional levels, as well as by medical organizations in general, in a format determined by the Russian Healthcare Ministry’s Central Research Institute for Healthcare Data. Data for analysis can be automatically retrieved from structured electronic medical records using special tools.
Medical data: From collection to practical application
The goals of healthcare digitalization must prioritize maximizing benefits and convenience for both healthcare providers and patients. This process should be supported by health authorities and the Medical Information and Analytical Center. Every implementation of a digital medical solution should enhance the accuracy and timeliness of diagnoses, ensure higher quality and better-monitored treatments, and ultimately lead to saved lives and improved patient outcomes.
For instance, patients in our country can now conveniently schedule doctor appointments remotely through regional patient portals and the public services portal. In 2023, over 9.6 million appointments were made in St. Petersburg alone via the Petersburger’s Health service. Nationwide, the success rate of appointments through the State Services portal has increased from 30% to 59% in the past year and a half. These numbers represent millions of people who were able to book appointments with just a few clicks, receiving timely medical care. During the consultation, the patient’s electronic medical record is available to the doctor, providing a comprehensive overview of their health, including past illnesses, contraindications, previous treatments, and tests.
All of this has been made possible through the extensive collaborative efforts of medical organizations, regional health authorities, and IT solution developers. Analytical platforms have played a key role in elevating the quality of healthcare operations. These platforms enable real-time tracking and monitoring of medical institution performance metrics, such as identifying and correcting errors in doctor appointments. Additionally, analytics are now being used to address other critical healthcare tasks, including monitoring doctor workloads, ensuring the availability of medications, and more.
The next phase in healthcare digitalization involves extracting necessary information for analytics directly from medical documents as primary data. Structured electronic medical records form the foundation for creating a digital patient profile and unlock new possibilities for analytical systems. For instance, by comparing a patient’s blood test results over time, trends can be identified, allowing doctors to monitor health changes and intervene with treatment if needed. Additionally, AI-powered predictive tools can evaluate the risk of developing specific diseases, enhancing early detection and prevention.
The volume of structured electronic medical documents stored in healthcare systems is steadily increasing. For example, during the operation of the N3.Healthcare integration platform in Kuzbass, a distributed repository containing over 100 million documents has been created, with each document containing between 50 to 1,000 data points. On some days, up to 400,000 SEMDs are added.
At the federal level, the total number of accumulated electronic medical documents surpassed 2 billion this year and continues to grow exponentially. The variety of SEMDs is also expanding rapidly: about 100 unique types were recorded in 2022, around 200 in 2024, and approximately 250 are expected by 2025. This immense growth underscores the need for tools that can quickly and accurately extract valuable insights from these vast datasets.
Healthcare analytics for everyone
Today, structured electronic medical documents are exchanged in Russia based on the HL7 CDA international standard. A modern structured electronic medical record is an XML-based document, with each value associated with a specific reference such as a disease classifier. This applies to all types of electronic documents, including SEMD 194 (a medical report following a preliminary/regular medical examination/checkup, used for implementing the Incident 7 project related to the introduction of electronic health records); since the summer of 2024, it has been replaced by SEMD 230 (a medical report following an employee health checkup to be submitted to the electronic health records subsystem).
Along with XML data fields, those now include a CSV file that stores data for XPath queries. Previously, we had to compare each field with the reference, while now, starting with SEMD 194, this task has already been implemented by the Central Research Institute for Healthcare Data of the Russian Healthcare Ministry.
As regards data extraction and analytical processing capabilities, structured electronic medical documents compare favorably with the previously utilized unstructured PDF format. For instance, all the dynamics of the aforementioned indicators of a standard laboratory research can be monitored solely with a SEMD.
The procedure of creating rules to extract values and build necessary reports is a clear yet time-consuming task that has to be repeated with a new type or edition of an electronic medical document. A SEMD data extractor is a tool that can help solve this issue: you need to upload a template of an electronic medical document, configure value extraction in the user interface, and then apply these rules to the accumulated bulk of documents.
Specialists involved in working with SEMDs and XML documents at medical centers for data and analysis can independently obtain data marts to create new reports, a proper example being the aforementioned integration platform implemented in Kuzbass. At Zelkovich Kuzbass Regional Medical Center for Data and Analytics, analytics segments are created mostly by two specialists who independently build analytical data marts and promptly compile reports using the existing bulk of data.
Prospects for electronic document analytics
Digital healthcare solutions allow medical institutions to implement the patient-centric concept. The Russian Healthcare Ministry is engaged in a number of the Incident federal and specialized projects, some of them involving SEMDs. The Incident 7 health records project includes SEMD 103 (a medical report following a preliminary/regular medical examination or checkup), as well as SEMD 194, 230, and related documents. The Incident 9 project, which aims to reduce excess deaths, is based on monitoring of six indicators described by a number of SEMDs.
The efforts to improve performance indicators of the digital healthcare system and ultimately the quality of medical care require a profound knowledge of processes occurring in the acquired SEMD databases. Obviously, in order to rely on primary data provided by SEMDs, we as an industry will have to undergo a stage of development that involves extensive identification and elimination of errors within SEMDs for local institutions to obtain more quality and reliable information.
By Andrei Dyukov, Head of Sales and Development, Netrika Healthcare