https://orcid.org/0000-0002-3156-3936
Figures
Abstract
Background
The World Health Organization (WHO) emphasizes digital technologies as a key accelerator for strengthening health emergency preparedness and response. One such application is the early rescue chain (ERC), which coordinates automated alerting, responding, and clinical systems. However, current ERC implementations primarily rely on spoken language and lack standardized terminology. The ERC Terminology (ERC_T) defines the minimum set of concepts required for automated and standardized ERC information exchange. The International Classification of Diseases, 11th Revision (ICD-11) extensively covers adverse events in its sections: “External causes of morbidity or mortality” and “Dimensions of external causes”. However, there is no comprehensive mapping of ERC_T to the ICD-11 Foundation (ICD-11_F), the terminological resource from which the successor to ICD-10, ICD-11 Mortality and Morbidity Statistics (MMS), is derived. Therefore, we aim to: (I) assess semantic relationships between ERC_T and ICD-11_F; (II) evaluate the overall mapping quality; and (III) identify missing concepts and semantic gaps between ERC_T and ICD-11_F.
Methods
We will select ICD-11_F as the target of the mapping because of its greater specificity and its structure as a terminology rather than a classification that is constrained to a single-parent monohierarchy with residual categories. We will systematically map concepts from ERC_T to ICD-11_F using predefined mapping criteria and procedures. The mapping process addresses two complementary aspects of semantic alignment. First, we will classify the semantic relationship between an ERC_T concept and an ICD-11_F entity using HL7 FHIR ConceptMap relationships. Second, we will evaluate the quality of each mapping using ISO 21564:2025 (MapQual) measures, which provide a structured assessment of how well the ICD-11_F entity represents the ERC_T concept. Two coders will independently conduct the mapping following predefined mapping procedures. We will assess intercoder agreement using percentage and Krippendorff’s alpha (α), and will resolve disagreements through structured consensus or, if unresolved, by a senior expert.
Citation: Tegegne MD, Tu SW, Ibrahim I, Cornet R, Niakan Kalhori SR, Deserno TM (2026) Mapping the terminology of the early rescue chain to the Foundation of ICD-11: Registered report protocol. PLoS One 21(6): e0350733. https://doi.org/10.1371/journal.pone.0350733
Editor: Pasyodun Koralage Buddhika Mahesh, Ministry of Health, Sri Lanka, SRI LANKA
Received: March 6, 2026; Accepted: May 16, 2026; Published: June 16, 2026
Copyright: © 2026 Tegegne et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Data Availability: All data underlying this Registered Report Protocol are within the manuscript. Data generated by the study described in this protocol will be made available with the resulting Registered Report article.
Funding: The author(s) received no specific funding for this work.
Competing interests: The authors have declared that no competing interests exist.
1. Introduction
The Seventy-eighth World Health Assembly (WHA78) approved the extension of the global strategy on digital health to 2027 [1]. Under the Thirteenth General Program of Work (GPW13, 2019–2025) [2], the World Health Organization (WHO) strategically strengthened the architecture of health emergency preparedness and response using digital technologies as key accelerators.
A practical implementation of this digital and data-driven approach is Accident & Emergency Informatics (A&EI) [3], an interdisciplinary field that collects, analyses, and communicates data on health emergencies and disasters to support prompt decision-making and to save lives. A core component of A&EI is the early rescue chain (ERC), which coordinates the automated communication among alerting, responding, and clinical systems [4]. However, current ERCs use spoken language for communication, and a significant barrier is the lack of semantic interoperability [5].
Deserno & Jakob [6] emphasized the need for a standardized terminology to automate the ERC. Accordingly, we developed a minimum set of concepts to support automated information exchange in the ERC, referred to as ERC Terminology (ERC_T) [7]. ERC_T follows a mono-hierarchical, multi-axial terminology structure comprising five categories: event, outdoor location, indoor location, type of transport, and type of room (Fig 1). Each category contains a defined concept organized in a multi-level hierarchy, comprising a total of 307 ERC_T concepts, excluding residual concepts (i.e., “other” and “unknown”). The terminology explicitly defines inclusion and exclusion criteria for each concept.
The International Classification of Diseases, 11th Revision (ICD-11) covers adverse events through its dedicated chapter on “External causes of morbidity or mortality” and related Extension code under “Dimensions of external causes” [8]. The informatics architecture of ICD-11 is based on the ICD-11 Foundation (ICD-11_F), a polyhierarchical, multi-level semantic network that comprises all ICD-related content [9]. The ICD-11_F represents each clinical concept as a distinct entity, identified by a unique Uniform Resource Identifier (URI). It is part of the WHO Family of International Classifications (WHO-FIC) Foundation, which comprises over 100,000 entities across ICD, the International Classification of Functioning, Disability and Health (ICF), and the International Classification of Health Interventions (ICHI) [10]. The ICD-11 Mortality and Morbidity Statistics (ICD-11 MMS) is the main linearization of the ICD-11_F and constitutes a selected subset that meets the requirements of a statistical classification. It offers around 17,000 unique codes for diagnoses and causes of death within a fully digital and flexible structure [11]. These codes are supported by a rich set of index terms and synonyms, enabling detailed clinical expression. Accordingly, ICD-11_F contains detailed clinical entities, while residual categories such as “Other specified” and “Unspecified” are features of the ICD-11 MMS and do not exist in ICD-11_F.
Despite the availability of ERC_T and ICD-11_F, the lack of a systematic and objective assessment of their alignment hinders efforts to leverage ICD-11 to support semantic interoperability for automated emergency response. Therefore, we aim to analyze the semantic relationship and mapping quality between ERC_T and the ICD-11_F in describing adverse events.
Assessing mapping between terminologies has been extensively studied. Several studies employed the Fast Healthcare Interoperability Resources (FHIR) ConceptMap resource [12] to assess semantic relationships between terminological systems. Ohlsen et al. [13] developed algorithms that automatically generate machine-readable FHIR ConceptMap relationships from the WHO ICD-10-to-ICD-11 mapping tables to support automated mapping and the transition to ICD-11. Similarly, we conducted an exploratory manual mapping of ERC_T to ICD-11 MMS [14], applying high-level FHIR ConceptMap relationships [15]. Together, these studies demonstrate that FHIR ConceptMap provides a practical framework for implementing computable terminology mappings, thereby enabling semantic interoperability and data exchange.
In addition, several recent studies have applied ISO/TS 21564:2019 Health informatics terminology resource map quality measures (MapQual) to evaluate semantic interoperability in clinical and research datasets. Rinaldi et al. [16] used this standard to assess semantic interoperability between the International Severe Acute Respiratory and Emerging Infection Consortium (ISARIC) freestanding follow-up survey and a large multicohort COVID-19 research dataset. Vorisek et al. [17] applied the same framework to evaluate the suitability of Systematized Nomenclature of Medicine Clinical Terms (SNOMED CT) for encoding the German Corona Consensus Dataset (GECCO). Extending this approach to social and diversity data, Muzoora et al. [18] used the standard to assess the alignment of SNOMED CT and Logical Observation Identifiers Names and Codes (LOINC) with the Diversity Minimal Item Set (DiMIS). Overall, across these studies, researchers applied ISO/TS 21564:2019 equivalence assessment scores and calculated weighted-average scores to evaluate overall semantic equivalence, thereby providing a standardized assessment of semantic interoperability when integrating heterogeneous clinical and research datasets. However, these studies primarily focus on clinical and research datasets.
To our knowledge, no work systematically combines FHIR-based semantic relationship classification with ISO 21564 (MapQual) quality assessment. This registered report protocol specifies the procedure for reproducible mapping of ERC_T to ICD-11_F using FHIR and ISO standards, and a standardized assessment of semantic interoperability. Our objectives are to:
- I. Assess semantic relationships;
- II. Evaluate overall mapping quality; and
- III. Identify missing concepts and semantic gaps.
2. Materials and methods
2.1. Definitions of terms
This section defines the basic terms used throughout this registered report protocol. The definitions reflect their operational use in mapping ERC_T to ICD-11_F (Table 1).
2.2. Study design
This study is a preregistered terminology mapping study. We systematically map concepts from ERC_T to ICD-11_F using predefined mapping criteria and procedure. The units of analysis are all ERC_T concepts eligible for mapping. We execute all our predefined mapping workflows in accordance with the approved registered report protocol.
2.3. Terminology sources and resources
2.3.1. Sources terminology.
Our source terminology is ERC_T [7], which defines a minimum set of concepts for automatic alerts within the ERC across five categories: event, outdoor location, indoor location, type of transport, and type of room (Fig 1). Each category contains multi-level concepts with their own definitions, along with inclusion and exclusion criteria provided where applicable.
2.3.2. Target terminology.
Our target terminology is the ICD-11_F, which includes a dedicated section for the detailed description of “Events” through “External causes of morbidity or mortality” [21] and the related Extension codes under “Dimensions of external causes” [22] (Fig 2). Building on our previous work demonstrating that the ERC_T’s ICD-11 MMS map targets include below-the-shoreline ICD-11_F entities (represented in ICD-11 MMS as index terms) and ICD-11 MMS residual categories that affect semantic interoperability [14,23], we shift the target to ICD-11_F, which is structured as a terminology rather than a classification [9]. This shift avoids these constraints and supporting the direct mapping of ERC_T concepts to ICD-11_F entities. Using the ICD-11_F API (v2.5) [24], we extract approximately 14,754 unique concept URIs that represent the complete set of entities under the “External causes of morbidity or mortality” and the “Dimensions of external causes”.
2.4. Scope and eligibility criteria
The mapping is limited to valid ERC_T entries with explicit definitions. Residual entries such as “other” and “unknown” are excluded. Within ICD-11_F, the scope is restricted to “External causes of morbidity or mortality” and their associated dimensions in the Extension codes section, as other sections are not relevant to the ERC.
We follow a one-to-one mapping cardinality. This is because ERC_T is a hierarchical terminology of predefined concepts (e.g., a specific event, location, transport type, or type of room) rather than a collection of pre-coordinated composite terms. The target, ICD-11_F, is a terminology in which a URI uniquely identifies each entity. Therefore, single-target semantic equivalence is the appropriate approach for assessing semantic alignment between the two terminologies. Compositional cardinality scenarios (one-to-many, many-to-one, or many-to-many) lie outside the scope of this terminology mapping, as they pertain to the combined use of terminologies in real-world coding, such as encoding a complete event record.
2.5. Mapping framework
2.5.1. Semantic relationship.
- I. Definition
The FHIR “ConceptMap” framework [12] defines explicit semantic relationship types between source and target concepts. The relationship element is bound to a value set derived from the FHIR CodeSystem “ConceptMapRelationship” [15] (Table 2). In our work, we depart from the FHIR “ConceptMapRelationship” in two ways. First, the “Related to” relationship is defined as “The concepts are related to each other, but the exact relationship is not known.” However, the subcategories “Equivalent,” “Broader,” and “Narrower” do not cover the case where the source intersects (overlaps with) the target. Instead of adding a subcategory, we include such cases in the “Related to” category. Second, the FHIR “ConceptMapRelationship” defines the semantic relationship type “Not related to”; however, this relationship does not apply to our mapping. We consider only a predefined set of source concepts, while target concepts are identified dynamically and are not predefined.
Therefore, if we identify no suitable target concept, this indicates the absence of a mapping rather than an explicit semantic judgment of non-relatedness. We represent these cases using the “ConceptMap.group.unmapped” [27] element, as defined in the FHIR specification, and label them as “No match”.
- II. Examples
A necessary part of a mapping framework is the criteria for operationalizing the assignment of semantic relationships. This is especially important because of the mismatch between the conceptual models of ERC_T and ICD-11_F: ERC_T focuses on describing adverse events to alert emergency services, whereas ICD-11_F is designed for classifying external causes of injury. In this section, we explore the assignment of semantic relationships with some examples.
The ERC_T concept “Transport accident” is related to the ICD-11_F entity “Unintentional transport injury event.” However, the source concept also includes transport accidents resulting from intentional or assaultive acts. Therefore, it has a broader semantic scope than the target entity “Unintentional transport injury event.”
The ERC_T concept “Pedestrian” is defined as a type of land transport referring to a person traveling on foot, including standing, walking, or running, and encompasses “Pedestrian using a mobility aid” and “Pedestrian without a mobility aid”. The ICD-11_F entity “Pedestrian as mode of transport of person injured in transport event” accurately reflects the ERC_T definition. Although ERC_T defines pedestrians as a separate land transport and models transport accidents independently, ICD-11_F explicitly describes pedestrians as a mode of transport in relation to injury. Even with this difference, both terms refer to the same transport role in traffic-related adverse events and are therefore considered equivalent.
The ERC_T concept “Technical disaster” refers to adverse events caused by direct human action. Its terminal concepts are, for instance, “Building collapse”, “Explosion”, and “Fire”. Although these concepts can be mapped based on their specific definitions, no corresponding parent-level in the ICD-11_F encompasses the full scope of “Technical disaster.” Consequently, the mapping is classified as no match (Table 3).
- III. Pros and cons of FHIR CodeSystem “ConceptMapRelationship.”
Pros
- Defines semantic relationships between concepts across terminologies [15].
- Enables interoperability and data exchange in FHIR systems [28].
- Supports automated terminology mapping through machine-processable semantic relationship types, facilitating integration across systems [13,29].
- Widely adopted and open source [30].
Cons
- Semantic relationships between terminology concepts are not turned into a numeric score, making it challenging to quantify the overall semantic interoperability [15].
- Less suitable for research requiring a detailed evaluation of mapping quality [31].
- Does not explicitly cover the case where the source intersects the target.
- Relies on predefined source and target concepts and represents unmatched cases as mapping outcomes rather than semantic relationships [27].
2.5.2. Mapping quality.
- I. Definition
ISO 21564:2025, Health Informatics — Terminology resource map quality measures (MapQual) provides a framework for quantifying the quality of maps between terminological resources [20]. The framework defines equivalence measures as the degree to which the meanings of a source and a target concept match. Mapping quality is evaluated by assigning an equivalence score to each map, with the lowest score (0) corresponding to the highest quality match (Table 4).
II. Examples
We use the same examples as in Section 2.5.1.2. Mapping the ERC_T concept “Transport accident” to the ICD-11_F entity “Unintentional transport injury event” results in a partial inclusion. The source concept encompasses transport accidents of various intents (i.e., unintentional, intentional, and undetermined). Only unintentional transport accidents fall within the semantic scope of the selected ICD-11_F target entity. Therefore, the source concept can be partitioned into two disjoint subsets: one included in the target and the other excluded from it. Accordingly, this mapping is assigned a quality score of 2.
Mapping the ERC_T concept “Pedestrian” to the ICD-11_F entity “Pedestrian as mode of transport of person injured in transport event” received a quality score of 0, reflecting semantic equivalence despite differences in modeling. Furthermore, mapping the ERC_T concept “Technical disaster” to ICD-11_F receives a quality score of 4, reflecting that no valid ICD-11_F target entity adequately represents the full scope of the source concept (Table 5).
- III. Pros and Cons of ISO 21564:2025
Pros.
- Provides quality assessment of terminology mappings [20].
- Offers numeric or graded measures of mapping quality for evaluating semantic interoperability between two terminology systems [20].
- Designed for research contexts requiring methodological rigor, enabling evaluation and comparison of overall mapping quality across datasets [20].
Cons.
- Evaluates mapping quality but does not specify semantic relationship types (e.g., broader, narrower, related) [20].
- Limited automation requires expert-driven and conservative semantic assessment [16].
- Newly standardized (2025) [20] and integration with FHIR requires separate guidance [32].
- Not open source [20].
2.5.3. Complementarity of mapping frameworks.
Although certain mapping labels may appear conceptually aligned across the two frameworks, for instance, the “Equivalent” semantic relationship and the quality score “0”, these overlaps reflect intentional methodological complementarity rather than redundancy. The frameworks are neither intended to replace one another nor to be used interchangeably. Instead, each provides a distinct perspective on semantic relationship and mapping quality, and will be applied independently to address the separate research questions.
Therefore, our registered report protocol applies both distinct mapping frameworks to address separate objectives. HL7 FHIR ConceptMap framework establishes semantic relationships, while ISO 21564:2025 (MapQual) measures mapping quality.
2.6. Mapping procedure
The mapping procedure follows a sequence of predefined stages (Fig 3). Each stage builds on the output of the previous stage and is applied consistently across all ERC_T concepts.
2.6.1. Searching for target entities.
In the first stage, two ICD-11 coders will independently search for each ERC_T concept in the ICD-11_F using the WHO-FIC Foundation Browser [33], without access to each other’s mappings. They will not consider concepts from the ICF and ICHI included in the browser, as they are outside the scope of this registered report protocol. The ERC_T concept serves as the source. To enhance target retrievals, coders will consider the source concept’s definition, inclusion criteria, and exclusion criteria. In addition, the coders will use the ICD-11 MMS coding tool [34] to determine how the identified ICD-11_F entities are represented within ICD-11 MMS. This process will identify entities represented with a distinct ICD-11 MMS code, those listed as index terms, and those classified in a residual category.
2.6.2. Mapping and semantic gap identification.
After identifying relevant ICD-11_F entities, the coders will classify the semantic relationship for each mapping using the HL7 FHIR ConceptMap framework. They will assign mapping quality scores according to ISO 21564:2025, based on the following predefined assumptions underlying mapping decisions (Section I). When mappings are not equivalent, the coders will identify semantic gaps between the ERC_T concept and the ICD-11_F entities.
- I. Assumptions underlying mapping decisions
These assumptions define methodological decision rules for mapping and do not reflect deficiencies in either ERC_T or ICD-11_F. To ensure reproducibility, we classify mapping decisions based on the following assumptions, informed by our feasibility analysis of the target ICD-11_F entities (Section 2.7).
- A. Absence of a suitable target entity
If no ICD-11_F target entity exists to represent the ERC_T concept, coders will assign the mapping outcome of “No match”, resulting in a quality score of “4” (No map possible). This rule operationalizes the feasibility check and supports the identification of missing ERC_T concepts in the ICD-11_F. Examples are the ERC_T concepts “Technical disaster”; locations such as “Underwater”, “Air”, and “Space”; and “Cableway” as a form of land transport.
- B. Equivalence semantic alignment
The coders will assign the semantic relationship “Equivalent” and the quality score “0” if both concepts are semantically equivalent. However, differences in system modeling perspective, such as the event-based representation in ERC_T and classification of external causes of injury in ICD-11_F, do not, by themselves, preclude equivalence across the two frameworks, provided that they have the same semantic meaning. For instance, “Pedestrian” (event-focused) and “Pedestrian as mode of transport of person injured in transport event” (cause of injury-focused) refer to the same type of transport.
- C. Partial semantic alignment
If a target ICD-11_F entity exists but is not semantically equivalent, the coders will assign the semantic relationship “source is broader than target,” “source is narrower than target,” or “related to” according to the HL7 FHIR ConceptMap framework and a quality score of “1–3” according to ISO 21564:2025.
Given the operational alignment between the two frameworks, the semantic relationship “Source is narrower than target” is interpreted as a quality score of “1” (The source is wholly included in the target). Similarly, the semantic relationship “Source is broader than target” is interpreted as a quality score of “2” (The source is partially included in the target). We illustrate this with the following examples:
- The ERC_T concept “Engine cycle” has the definition “Motorized vehicle without a cabin” and maps to the ICD-11_F entity “Motorized two- or three-wheeled vehicle.” The semantic relationship is classified as “Source is narrower than target,” as every engine cycle is a subset of a motorized two- or three-wheeled vehicle. Since the target ICD-11_F entity fully encompasses the ERC_T concept, the mapping is assigned a quality score of “1” (The source is wholly included in the target). The constraint “without a cabin” semantically narrows the source concept by restricting its set of instances, but the restricted set remains fully contained within the target concept and therefore does not alter the mapping decision.
- The ERC_T concept “Fall,” defined as an adverse event in which a person loses stability and descends to a lower level, is mapped to the ICD-11 target entity “Unintentional fall”. The semantic relationship is classified as “Source is broader than target,” as the ERC_T concept encompasses falls of varying intent. Accordingly, the source concept partitions into two disjoint subsets: unintentional falls within the target, and intentional or undetermined intent falls outside the target. Therefore, this mapping receives a quality score of “2” (The source is partially included in the target).
However, we cannot always interpret the two frameworks directly in relation to each other. The FHIR semantic relationship “related-to” corresponds to an ISO mapping quality score of “2” or “3”, depending on the nature of the relationship. If a “related-to” relationship indicates a well-defined intersection, we will assign a quality score of “2” (The source is partially included in the target). In contrast, if no clear intersection exists, we will assign a quality score of “3” (Source is mapped; however, there were many options of possible concepts and overlaps. The map produced is a best comparison rather than an actual correspondence. Such a map requires significant manual input to create and maintain, and it should be used with care). The following two examples illustrate this distinction:
- The ERC_T concept “Residential area,” defined as “Outdoor land area used for daily living activities,” maps to the ICD-11_F entity “Home.” The semantic relationship between the two is “Related to,” as neither concept fully subsumes the other: ERC_T concept covers all outdoor residential areas, including those that extend beyond homes (e.g., parks, sports and entertainment areas, squares, and marketplaces), whereas ICD-11_F’s “Home” is limited to a single dwelling and its indoor spaces. The two concepts share a well-defined intersection: the outdoor spaces directly attached to a single dwelling (e.g., garden, roof, backyard, garage). The resulting quality score is “2” (the source is partially included in the target).
- The ERC_T concept “Diver-assisted device,” defined as “Underwater transport devices that directly augment a diver’s capabilities,” maps to the ICD-11_F entity “Underwater diving equipment.” The semantic relationship is “Related to” because both refer to equipment used by divers underwater. However, there is no well-defined intersection, as the ERC_T concept refers to devices that actively assist a diver’s movement (e.g., propulsion devices, assist exoskeletons), whereas ICD-11_F refers to the passive gear divers wear underwater (e.g., aqualung, wetsuit, mask, fins). This results in a quality score of “3” (Source is mapped; however, there are many options of possible concepts and overlaps. The map produced is the best comparison rather than an actual correspondence. Such a map requires significant manual input to create and maintain, and should be used carefully).
- II. Documentation
The two coders will independently perform the mappings, without access to each other’s sections, and will document their results in separate, structured spreadsheets (Excel, Microsoft Corporation, Redmond, WA, USA). For each predefined ERC_T concept, they will record the mapped ICD-11_F entity (title and Foundation URL), its representation in the ICD-11 MMS, the assigned semantic relationships, the quality scores, and any identified gaps.
2.6.3 Quality assurance.
Based on this documentation, a WHO-FIC expert will conduct a pilot assessment and, where needed, provide guidance or training to address systematic issues in applying our predefined mapping procedure and ICD-11 coding principles. Then, we will calculate the percentage agreement of target entity selection and evaluate intercoder reliability using Krippendorff’s alpha (α) [35]. We will interpret inter-coder reliability using the predefined thresholds (Table 6), adapted from Krippendorff [35] and Marzi et al. [36]. In cases of disagreement, coders will resolve them through structured consensus discussions. If they cannot reach a consensus, they will refer the case to the WHO-FIC expert for final decision.
2.7. Feasibility of target entities
This section reports a preliminary, single-coder feasibility analysis conducted to determine whether target ICD-11_F entities exist for the ERC_T concepts, without assessing semantic relationships and quality score. We limit the assessment to confirming the availability of potentially relevant target ICD-11_F entities to support the proposed registered report protocol, and it does not constitute a final mapping decision or validated mapping results.
At the ERC_T category level, the ICD-11_F section “External causes of morbidity or mortality” constitutes a potential target for the ERC_T event category. The ICD-11_F extension code section “Dimensions of external causes” is a potential target for the remaining ERC_T categories: indoor and outdoor locations, types of transport, and types of rooms. Within this structure, we group outdoor and indoor locations, as well as room types, under the target ICD-11_F entity “Aspects of place of injury occurrence.” We group transport types under the target ICD-11_F entity “Objects, living things or substances involved in causing injury” (Fig 4). As an illustrative example of the feasibility assessment, we focus on the ERC_T concepts within the “Event” category. The ICD-11_F entities “Unintentional transport injury event” and “Unintentional fall” represent potential target entities for the ERC_T concepts “Transport accident” and “Fall,” because the term “unintentional” closely aligns with “accidental.” However, ERC_T does not encode intent, whereas intent is a central organizing principle in the ICD-11_F section “External causes of morbidity or mortality.” For example, a transport accident in ERC_T does not distinguish between intentional and unintentional events, while ICD-11_F explicitly differentiates between these intent options. These conceptual differences define important constraints for the subsequent mapping process and influence the assignment of the semantic relationship and quality score.
Single-coder example mapping illustrating only potential targets, without assignment of semantic relationships or quality scores.
The ICD-11_F entity “Exposure to extreme forces of nature” represents a potential target for the ERC_T concept “Natural disaster.” In addition, ERC_T concepts such as “Technical disaster” and “External violence” do not correspond to a single consolidated entity or correspond to two target ICD-11_F entities. Although we identify two target ICD-11_F entities during the feasibility assessment (e.g., external violence), we prioritize and select a single target entity in accordance with the predefined one-to-one mapping procedure (Fig 4).
3. Discussion
Anticipated outcomes
The first anticipated outcome of this registered report protocol is the systematic assessment of the semantic relationship between each ERC_T concept and ICD-11_F entities using a predefined HL7 FHIR ConceptMap relationship. We will identify the semantic relationship for each ERC_T-to-ICD-11_F mapping. This establishes a logic that can be easily translated into a machine-processable format, thereby supporting interoperability and quality data exchange in a future FHIR-compliant environment.
The second anticipated outcome is to evaluate mapping quality using the ISO 21564:2025 (MapQual) equivalence assessment score. We will assign a quality score from (0–4) for each ERC_T–ICD-11_F mapping, and calculate a weighted average to summarize the overall mapping quality. This approach enables quantification of the overall semantic alignment between ERC_T and ICD-11_F and facilitates comparison with mapping quality results reported in other ISO-based studies. Lower weighted average scores indicate stronger semantic alignment, while higher scores reflect greater reliance on best-comparison mappings.
Finally, based on the semantic relationships and mapping quality, we will identify the missing ERC_T concepts and characterize semantic gaps between ERC_T and ICD-11_F. These findings will inform future refinement of ERC_T and the proposal of potential enhancements to ICD-11 content, thereby supporting semantic interoperability between ERC and ICD-11. Furthermore, this registered report protocol provides a foundation for future research. Such research may include developing algorithms to automate the mapping process and extending the methodology to other terminologies and classifications in health informatics.
Anticipated limitations
This report protocol has several anticipated limitations. Although “Health emergency” exists as an intermediate concept under the ERC_T event category, we do not operationalize it into a structured set of concepts with inclusion and exclusion criteria suitable for terminology mapping. Consequently, we do not map health emergency events in this registered report protocol. Therefore, interpretations of future findings apply only to the ERC_T concepts explicitly operationalized by this registered report protocol.
We deliberately restricted the scope to the ICD-11_F sections “External causes of morbidity or mortality” and the related extension codes under “Dimensions of external causes,” following a detailed review of the current ERC_T concepts against the ICD-11_F. However, other ICD-11_F sections may capture additional aspects of emergency events, injury mechanisms, or contextual response settings. Extending the ERC_T concepts to cover such entities lies beyond the aim of this registered report protocol and will be considered in future work.
Although the source terminology is a nomenclature, ICD-11_F is semantically rich and may contain several plausible target entities for a single source concept. In these cases, the coder must select the most appropriate single target, a subjective process that may degrade mapping quality. There is also the possibility of coder subjectivity in semantic interpretation. We will mitigate this by applying predefined mapping decision rules and mapping examples. We will further diagnose coder subjectivity through inter-coder reliability analysis (Section 2.6.3) and resolve it through structured consensus discussions. Furthermore, the findings of this study are specific to the terminology mapping between ERC_T and ICD-11_F within the defined scope, and cannot be directly generalized to other emergency informatics contexts.
References
- 1.
World Health Organization. Seventy-eighth World Health Assembly 2025. World Health Organization. https://apps.who.int/gb/ebwha/pdf_files/WHA78/B156_REC1_EXT-en.pdf. 2025. Accessed 2025 October 9.
- 2. World Health Organization. Extending the Thirteenth General Programme of Work, 2019-2023 to 2025. 2023. https://apps.who.int/gb/ebwha/pdf_files/WHA75/A75_8-en.pdf
- 3.
Deserno TM, Haghi M, Al-Shorbaji N. Accident and emergency informatics. Amsterdam: IOS Press. 2022.
- 4. Haghi M, Barakat R, Spicher N, Heinrich C, Jageniak J, Öktem GS, et al. Automatic Information Exchange in the Early Rescue Chain Using the International Standard Accident Number (ISAN). Healthcare (Basel). 2021;9(8):996. pmid:34442133
- 5.
Sotoodeh M. Challenges in semantic interoperability in emergency management. In: Proc 16th Doctoral Consort (CAiSE-DC’09), 2009.
- 6.
Deserno TM, Jakob R. Accident & emergency informatics: terminologies and standards are needed for digital health in the early rescue chain. In: Proc IEEE AICT, 2020. 1–5.
- 7. Tegegne MD, Ibrahim I, Cornet R, Deserno TM. Terminology for Automatic Alerts in the Early Rescue Chain. Stud Health Technol Inform. 2025;329:1580–1. pmid:40776128
- 8. Southern DA, Harrison JE, Romano PS, Le Pogam M-A, Pincus HA, Ghali WA. The three-part model for coding causes and mechanisms of healthcare-related adverse events. BMC Med Inform Decis Mak. 2022;21(Suppl 6):376. pmid:35209889
- 9. Chute CG, Çelik C. Overview of ICD-11 architecture and structure. BMC Med Inform Decis Mak. 2022;21(Suppl 6):378. pmid:35578335
- 10. World Health Organization. WHO Family of International Classifications (FIC). https://www.who.int/standards/classifications. 2023. Accessed 2026 February 16.
- 11. World Health Organization. ICD-11 2022 release. https://www.who.int/news/item/11-02-2022-icd-11-2022-release. 2022. Accessed 2026 February 16.
- 12. HL7 Fast Healthcare Interoperability Resources. ConceptMap. https://build.fhir.org/conceptmap.html. Accessed 2025 November 28.
- 13. Ohlsen T, Ingenerf J, Wiedekopf J. Generating a FHIR ConceptMap from WHO’s ICD-10 to ICD-11 Mapping Tables. Stud Health Technol Inform. 2025;327:1308–12. pmid:40380717
- 14. Tegegne MD, Ibrahim I, Cornet R, Deserno TM. Mapping Automatic Alerting Concepts in the Early Rescue Chain to ICD-11. Stud Health Technol Inform. 2025;329:93–7. pmid:40775826
- 15. HL7 Fast Healthcare Interoperability Resources. CodeSystem. https://build.fhir.org/codesystem-concept-map-relationship.html. 2025. Accessed 2025 November 28.
- 16. Rinaldi E, Thun S, Stellmach C. ISO/TS 21564:2019- based Evaluation of a Semantic Map between Variables in the ISARIC Freestanding Follow Up Survey and ORCHESTRA Studies. J Med Syst. 2023;47(1):115. pmid:37962711
- 17. Vorisek CN, Klopfenstein SAI, Sass J, Lehne M, Schmidt CO, Thun S. Evaluating Suitability of SNOMED CT in Structured Searches for COVID-19 Studies. Stud Health Technol Inform. 2021;281:88–92. pmid:34042711
- 18. Muzoora MR, Haffer N, Thun S, Vorisek CN. How aligned are SNOMED and LOINC with the diversity minimal item set?. Stud Health Technol Inform. 2025;329:1530–4.
- 19. World Health Organization. WHO-FIC classifications and terminology mapping. https://cdn.who.int/media/docs/default-source/classification/who-fic-network/whofic_terminology_mapping_guide.pdf. 2021. Accessed 2025 October 9.
- 20.
International Organization for Standardization. ISO 21564:2025 Health informatics — Terminology resource map quality measures and requirements (MapQual). ISO. 2025. https://www.iso.org/standard/86417.html
- 21. World Health Organization. WHOFIC Foundation: External causes of morbidity or mortality. https://icd.who.int/browse/2025-01/foundation/en#850137482. 2025. Accessed 2025 November 28.
- 22. World Health Organization. WHOFIC Foundation: Dimensions of external causes. https://icd.who.int/browse/2025-01/foundation/en#1004338415. 2025. Accessed 2025 November 28.
- 23. Tegegne MD, Krips M, Ibrahim I, Deserno TM. Challenges of Integrating ICD 11 into Automatic Alerting Systems. Stud Health Technol Inform. 2024;316:267–71. pmid:39176724
- 24. World Health Organization. ICD API. https://icd.who.int/icdapi. 2025. Accessed 2025 December 28.
- 25. World Health Organization. External causes of morbidity or mortality [Internet]. 2026 [cited 2026 Feb 17]. Available from: https://icd.who.int/dev11/f/en#/http%3a%2f%2fid.who.int%2ficd%2fentity%2f850137482?view=V167
- 26. World Health Organization. Dimensions of external causes [Internet]. 2026 [cited 2026 Feb 17]. Available from: https://icd.who.int/dev11/f/en#/http%3a%2f%2fid.who.int%2ficd%2fentity%2f1004338415?view=V167
- 27. Resource ConceptMap – Detailed Descriptions. Fast Healthcare Interoperability Resources. https://build.fhir.org/conceptmap-definitions.html#ConceptMap.group.unmapped. 2026. Accessed 2026 February 20.
- 28. Terminology maps. HL7 Fast Healthcare Interoperability Resources. https://build.fhir.org/ig/HL7/terminology-fundamentals-ig/terminologymaps.html. 2025. Accessed 2026 January 23.
- 29. Tabari P, Costagliola G, De Rosa M, Boeker M. State-of-the-Art Fast Healthcare Interoperability Resources (FHIR)-Based Data Model and Structure Implementations: Systematic Scoping Review. JMIR Med Inform. 2024;12:e58445. pmid:39316433
- 30. The next generation standard has transformed health information technology across the globe. https://blog.hl7.org/hl7-international-celebrates-10-years-of-hl7-fhir. 2021. Accessed 2026 January 23.
- 31. Saripalle R, Sookhak M, Haghparast M. An Interoperable UMLS Terminology Service Using FHIR. Future Internet. 2020;12(11):199.
- 32. International Organization for Standardization. Health informatics — Guidelines for implementation of HL7 FHIR based on ISO 13940:2015, ISO 13606-1:2019 and ISO 13606-3:2019. https://www.iso.org/standard/78390.html. 2025. Accessed 2026 January 23.
- 33. World Health Organization. WHO Family of International Classifications Foundation. https://icd.who.int/browse/2026-01/foundation/en. Accessed 2026 February 18.
- 34. World Health Organization. ICD-11 Coding Tool. https://icd.who.int/ct/icd11_mms/en/2026-01. 2026. Accessed 2026 February 18.
- 35.
Krippendorff K. Content analysis: an introduction to its methodology. 4th ed. Thousand Oaks (CA): SAGE Publications. 2019.
- 36. Marzi G, Balzano M, Marchiori D. K-Alpha Calculator-Krippendorff’s Alpha Calculator: A user-friendly tool for computing Krippendorff’s Alpha inter-rater reliability coefficient. MethodsX. 2024;12:102545. pmid:39669968