Interoperability and machine learning in primary care - a clinical decision support system for low back pain
Wendy Oude Nijeweme – d’Hollosy is a PhD Student in the research group Biomedical Signals and Systems. Her supervisor is professor Hermie Hermens and co-supervisor is doctor Lex van Velsen from the Faculty of Electrical Engineering, Mathematics and Computer Science (EEMCS).
The term eHealth is commonly defined as "Health services and information delivered or enhanced through the Internet and related technologies". General examples of eHealth applications are websites that make health information accessible to patients, or eConsult environments for secure digital communication between a patient and his/her general practitioner (GP). More specific eHealth applications are web based environments that support patients to work on their own health (self-management), systems that support remote diagnosis and treatment of patients (telemedicine), and clinical decision support systems (CDSSs).
A CDSS can be defined as “Any computer program designed to help healthcare professionals to make clinical decisions”. Over time, CDSSs have been shown to improve both patient outcomes and costs of care by prompting, reminding and cautioning clinicians whether or not to intervene under specific clinical circumstances. Nowadays, some CDSSs are already used in daily primary care, because they are implemented as functionalities of the healthcare information systems of the healthcare professionals. These functionalities are mainly used for prevention and screening, drug dosing, medical management of acute diagnoses and chronic disease management through the usage of alerts and computerized protocols. eHealth technology can benefit the healthcare system in a variety of ways. It can help to achieve efficient management of health data and the possibility to share these data among healthcare professionals, informal caregivers, and patients within patient care processes. The quality and sustainability of healthcare can be improved as well, by supporting the self-management of patients. Next to this, big data solutions, based on connected digital health data from different sources, can support the development of CDSSs.
Despite of its benefits, eHealth is still not widely used in primary care. One reason is that both healthcare providers and patients are not aware of the possibilities of eHealth in primary care. Another reason that hinders the implementation of eHealth in primary care is that barriers still exist around integrated and interoperable technological infrastructures for eHealth.
Interoperability is defined as the ability of two or more systems or components to exchange information and to use the information that has been exchanged. Interoperability between health systems facilitates health information exchange (HIE). Interoperability barriers that hinder smooth HIE are related to technical, organizational, safety, privacy, and security issues. For example, standalone systems are used that store data in different formats and without means for data exchange, despite the existence of available HIE communication standards, like HL7 and terminology standards as SNOMED CT.
The objective of this PhD research was to gain more knowledge on how to achieve interoperable eHealth technology for primary care and how this interoperability can be used for decision support via a data-driven approach with the help of machine learning, focused on the development of a clinical decision support system (CDSS) for optimized patient referral. Patients with low back pain (LBP) were chosen as first application area, as LBP is the most common cause for activity limitation and has a tremendous socioeconomic impact in Western society.
