The Luxembourg Institute of Health (LIH) is revolutionising health research with PIANISSIMO, its first openEHR-native application. Designed to support osteoarthritis patients in staying active, this smartphone-based intervention leverages structured health data to personalise physical activity plans.
Abstract
The Luxembourg Institute of Health (LIH) is pioneering the adoption of the open Electronic Health Record (openEHR) standard across its applications, with Pianissimo being the first instance. This initiative aims to enhance data interoperability and quality in both clinical settings and research projects by utilizing structured models like archetypes from the international Clinical Knowledge Manager (CKM). By embracing openEHR, the LIH seeks to facilitate seamless data exchange, promote the reuse of predefined models, integrate with other healthcare standards, and ultimately improve research outcomes through better data handling practices.
Luxembourg is a country in Western Europe, bordered by Belgium, France, and Germany. Despite its small size, it holds significant economic and political influence, serving as a major financial centre and a founding member of the European Union. The country has three official languages – Luxembourgish, French, and German – reflecting its diverse cultural heritage. Luxembourg City, the capital, is a key financial hub in Europe. The nation consistently ranks among the highest in the world for quality of life and economic prosperity. As of 2024, Luxembourg has a population of approximately 660,000 people. It is one of the most multicultural countries in Europe, with nearly 50% of its residents being foreign nationals. The largest expatriate communities come from Portugal, France, Italy, Belgium, and Germany. Additionally, the country’s strong economy and job market attract a significant number of cross-border workers from France, Belgium, and Germany, who contribute to its daily workforce.
The Luxembourg Institute of Health (LIH) is a public biomedical research organization dedicated to advancing precision health. Its mission focuses on patient-centric translational research, particularly in the areas of cancer and immune-related disorders. The institute also invests in key topics such as digital health, preventive medicine, clinical research, and precision health, aligning with its commitment to transforming scientific excellence into meaningful patient outcomes. The LIH actively collaborates with healthcare providers, hospitals, and both public and private biomedical organizations, acting as a catalyst for research partnerships. It also serves as a primary source of public health information in Luxembourg, aiding public authorities in making informed decisions based on scientific data.
As a biomedical research institute, the LIH relies on data from different sources – hospitals (each with their own different applications), third-party apps and electronic data capture (EDCs) applications (e.g. REDCap™, Ennov Clinical™), wearables (e.g. Apple and Android based devices, Withings™) and local developments, among many others. The collected data includes medical history, current health status (such as body measurements, vital signs, imaging reports, lab results, substance use, allergies), and informed consent for secondary data use. This heterogeneous data collection setting leads to the same data being collected multiple times from the same study participants and sources. Moreover, this data is modelled by individuals with different backgrounds and areas of expertise, each with their own approach to modelling information. Unfortunately, these models are often misaligned, as each group tailors their models to their very specific needs – often ignoring the greater picture. As a result, data from different projects is stored in separated silos, formatted differently (with variations in variable names, data types, and contextual meaning), and created using different applications – many of which lock the data into their systems and do not follow any standardized data specifications or ISO formats. Even within the same application, different projects may collect identical clinical data in different ways, further contributing to fragmentation and inconsistency. As a consequence, extensive efforts are required to extract, transform, and re-map the data while attempting to preserve its original context and meaning. This process is very complex and time-consuming, and in many cases, valuable data is lost.
Vision for the LIH
The LIH is currently undergoing a paradigm shift to improve data definitions, annotations, completeness and governance, with the overarching goal of increasing data quality and reusability across projects and studies, while avoiding redundancy. The LIH is aiming to build a data ecosystem with a strong foundation based on structured data and standardization – from data models to medical ontologies, guidelines and data queries, in an open and collaborative manner. In Luxembourg, there is still a long way to go in achieving health data standardization. Currently, much of the data is recorded in an unstructured manner, not tied to any standard, stored in proprietary formats, and frequently kept on paper or digitized as PDFs, with rare communication and exchange of data across medical institutions. Many stakeholders seem to be waiting for an “application solution” to emerge before acting, while the data is still the biggest pain point. Although there are initiatives to adopt HL7 FHIR, this is unlikely to fully address the core issue. What is truly needed is a persistent, detailed, structured, and high-quality clinical data repository with consistent data models. For example, instead of having ten different FHIR profiles collecting blood pressure data in various ways in different clinical settings, a single reusable openEHR “blood pressure” archetype should be used across different use cases. This would enable consistent and standardized queries of the same concept across studies and collaborating institutions. To support this initiative, collaborations with other partners and international consortia will be leveraged, such as with the federated European openEHR network.
PIANISSIMO – Objectives and Protocol
The name PIANISSIMO stands for “Personalised physIcal Activity promotioN in osteoarthrItis patientS uSIng a sMartphone-based sOlution.” PIANISSIMO is a research project designed to help people with hip and knee osteoarthritis stay active through a smartphone-based intervention. In the study, participants are provided with a mobile companion app, serving as a tool for patient education, data collection, and personalized physical activity promotion. Upon joining the study, participants provide demographic information and complete a validated instrument, the Self-Administered Comorbidity Questionnaire (SCQ), followed by weekly responses to a custom pain questionnaire and monthly completion of the standardized Western Ontario and McMaster Universities Arthritis Index (WOMAC). Additionally, the app continuously tracks daily step counts throughout the intervention.
The application actively supports patients in staying motivated – every day, participants receive a notification with useful information, motivation, or advice on physical activity.
During the six months intervention program, each week, the participant is invited to set personalized daily step goals for the week and receives feedback on their progress.
The overall philosophy behind the intervention is reflected as helping patients to gradually increase their activity levels, step by step, at their own pace. To ensure that the intervention is effective, it was developed according to physical activity recommendations from rheumatology scientific societies as well as with input from healthcare providers in Luxembourg. The program is structured using the COM-B behaviour change framework, which emphasizes three key components:
- Capability – Providing the right knowledge and tools for physical activity
- Opportunity – Creating an environment that encourages movement
- Motivation – Helping patients stay engaged through goal setting and feedback
All notifications sent to participants were carefully designed to reflect these principles.
For too long, many osteoarthritis patients have been told that movement is harmful, when in fact, staying active improves joint mobility and can reduce pain. Unfortunately, general practitioners often lack the time to emphasize the benefits of physical activity, and many patients struggle to maintain rehabilitation exercises once they return home. By delivering an intervention through a mobile app, PIANISSIMO hopes to bridge this gap, helping patients build an active lifestyle in a way that fits their daily routine.
How openEHR played a role the study (modelling, protocol definition)
During the first review of the study protocol and case report forms (CRFs) to implement, a common pattern across screening questionnaires emerged – covering patient history, physical examination and standard Patient Reported Outcomes (PRO) questionnaires. The objective for this study was to move beyond the “traditional” approach of simply replicating paper forms. Instead, a reusable solution that can be applied across different use-cases was targeted. This led to a standardized approach of choosing and reusing certain clinical concepts. An analysis of the archetypes in the international Clinical Knowledge Manager (CKM) was performed in order to asses which existing archetypes are fit to use out-of-the-box. This showed that only a few archetypes were missing and would have to be modelled after acquiring the necessary licences and copyrights to use them. For the remaining existing archetypes, the work focused on terminology bindings with the terminologies standards that are starting to be used at the LIH across different projects (e.g. WHO-ATC, SNOMED, ISCED, LOINC). To follow the collaborative spirit of openEHR, the modelled archetypes were submitted to the CKM for evaluation, and to share them with the community. Being able to reuse existing archetypes and constrain them to the use case of this project at the template level saved a lot of time in the modelling phase of the project. It also helped to discover additional relevant context and data points for the study that were not initially considered but were present in the archetypes.
Project contribution for openEHR
Two new archetypes were developed and shared for evaluation in the international openEHR CKM, the Western Ontario and McMaster Universities Arthritis Index (WOMAC) and Self-Administered Comorbidity Questionnaire (SCQ). To comply with the language requirements of the study, all of the archetypes used in the project have been translated to French and German.
Perspectives
PIANISSIMO is the first openEHR-native application in production at the LIH and in Luxembourg and it has demonstrated that development with openEHR is not as complicated as it may seem. The keys are understanding the basics, properly educating involved stakeholders, including clinical teams in modelling discussions, and ensuring that developers know how to interact with the openEHR APIs. Most of the heavy lifting is already done using openEHR’s predefined archetypes. The remaining effort is focused on aligning data requirements, business logic and applications to the openEHR philosophy and specifications. For the implementation, mature openEHR server implementations are available (some even free of charge). This project should prove that it is possible to build a functional research project on top of openEHR. Hopefully, this also helps to debunk the myth that openEHR development is overly complex.
The added value of openEHR for research and for Luxembourg
openEHR offers a lot of value and potential to research. The reference model is strong, stable, and covers most “core” data concepts, with published components that are tightly governed. However, this does not mean that everything will be ready for immediate use in all use cases. To ensure sustainability, it is important to contribute to the community efforts, just as we benefit from its contributions. Over the years, numerous challenges with vendor lock-in applications and poorly designed data models that lack a solid data structure were encountered. This is where openEHR shines and provides solutions. It starts by having the cumulative power of the data modelling community expertise combined with a strong reference model. Free access to a vast set of archetypes developed by experts from both medical and IT fields worldwide is a major asset, reducing cost of high quality models and eliminating the “reinvent the wheel” cycle for the same clinical concept – something that often happens across different research teams and departments, depending on the people and tools involved. In the long run, the LIH is aiming to avoid redundant data representation, where the same information is expressed in countless ways. By implementing the PIANISSIMO application using openEHR, the first step in this direction has been undertaken, enabling our researchers to prioritize the data itself, recognizing it as a core asset for research. This approach will allow researchers to quickly understand the data they will be working with – all the data definitions are always tied in the data model with consistent constraints and template validation – on the server side and application level. The recycling of 90% of the archetypes from the international CKM is essential for interoperability. Additionally, the ability to easily integrate different ontologies that are relevant for research studies makes it a quite powerful combination. With the recent community work from openEHR mapping with HL7 FHIR and OMOP CDM, this integration is even more effective, working as a win-win combination for research, facilitating integration with hospitals or federated research consortia.
At this stage, openEHR CDRs and top-plugin applications are more prominently used in clinical settings, not directly in biomedical research institutions, since research institutes usually extract parts of EHRs to feed the CRF of the studies using EDC or custom developed applications. This tends to lead to different formats and models, very basic proprietary applications for data collection, custom made apps for a specific use and other so called feral systems in between. To facilitate interactions between care and research, it is essential to establish a standardized format or “lingua franca” across all different settings. The fact that openEHR provides highly defined and structured clinical models can increase the data quality across all settings and seems to be the best fit at this point in time. It is then possible to dedicate more time building and designing the application that will be done on top of the openEHR server and data models.
More cases are possible and will be explored
PIANISSIMO is only the first application at the LIH using the openEHR as its data information format and more are expected. Another project using openEHR is currently ongoing – Clinnova is a European cross-border initiative with a vision to establish a data enabling environment in which translational research projects can be launched based on data standardisation, interoperability and federation. Partner countries include Luxembourg, France, Switzerland, and Germany, and the initial proof-of-concept use cases focus on patient cohorts in three immune-mediated diseases: Inflammatory Bowel Disease (IBD), Rheumatoid Diseases (RD) and Multiple Sclerosis (MS). Although the data is not registered immediately into openEHR, it will be mapped at a later stage. For the IBD and RD cohorts, 16 archetypes have been created, such as EULAR/ACR SLE, SLEDAI-2K, JAFS, PRQL, Brief IPQ, Lemann index, Wexner score, Harvey Bradshaw Index, CDAI, MARS-5, among others. For the MS cohort, in collaboration of Basel University Hospital (USB), eight additional archetypes have been created: Symbol Digit Modalities Test (SDMT), Digit span backwards (DSB), Trail Making Test (TMT A & B), Stroop Victoria Test, verbal learning and memory test (VMLT), Contrast vision test, EDSS/Neurostatus, among others. All of these archetypes have been submitted to the international CKM for editorial review and will hopefully soon be available to the community. Additionally more projects are in development and plans of further openEHR applications as part of our institutional roadmap – so stay tuned!
Join the PIANISSIMO Journey!
Participants can easily join by downloading the application in the Apple store or Google Play store, or by scanning a QR code in the official webpage to download the app and create an account. Along the way, they will learn more about physical activity and might even discover a new appreciation for movement! Some may even be pleasantly surprised by their own physical abilities.
This study is made possible thanks to the generous support of the Œuvre Nationale de Secours Grande Duchesse Charlotte, which promotes physical activity in Luxembourg.
Thanks to Vanessa Pereira, Michaël Schnell and Mathilde Mura for collaborating on this article.
Leave a Reply