JMIR Publications

ABSTRACT

Background:

In ageing populations, the demand for care, including care delivery in long-term care facilities, is increasing. This situation highlights the need to optimize care processes through continuous scientific evaluation. The use of artificial intelligence (AI) has potential for use in nursing research, but it suffers from a lack of standardization and structuring of nursing data. Although solutions such as standardized nursing terminologies exist, their use in practice has thus far not been widespread and is often associated with high documentation costs.

Objective:

This article presents the conceptual and technical development of a nursing minimum data set (NMDS) that focuses on a specific "fall" scenario. It aims to improve data standardization and usability for research and AI-based analysis in long-term care settings.

Methods:

The representation of the “fall” use case was developed using literature analyses, co-design workshops and a quantitative survey (n =158). The technical indexing was conducted by translating the results into the technical terminology of the HL7® FHIR® standard.

Results:

The "fall" use case was developed as part of a German NMDS for long-term residential care that included fall risk factors, interventions and outcomes. The relevant content, in the sense of a minimum set of items, was identified and prioritized in collaboration with nursing experts and translated into an FHIR® -based implementation guide. Discussion: The development of the “fall” module demonstrates a method for providing standardized nursing data based on routine care data for AI and research without increasing staff workload. Its co-designed, interoperable approach supports cross-sector data harmonization and offers a transferable model for international nursing contexts.

Conclusions:

This approach addresses the lack of structured nursing data for AI and research and can serve as an example for interoperable, cross-sector solutions in global long-term care.