A semantic mapping can be technically complete while the decision behind it remains unsettled. That creates a specific control problem in clinical-research interoperability. Information can move through the required transformation even when no explicit authority has established that two representations carry the same meaning. The consequence emerges at acceptance. A completed mapping may appear ready for use while review still lacks a clear basis for deciding whose equivalence judgment governs the interoperable result.
Equivalence is a decision point
Interoperability depends on more than the ability to connect different representations. At some point, the process must determine whether those representations express the same concept or require a more specific transformation. That determination controls the meaning carried forward. The execution step can be small, such as accepting a mapping in a system record, but its effect is larger because it establishes how information will be interpreted after the handoff. The key operational question is whether the person or function making that determination also holds authority to establish the meaning it represents.
Friction begins when those two functions are allowed to collapse into one another. A team can possess the technical ability to implement a semantic relationship without possessing recognized authority to approve the underlying equivalence. If the approval path is unclear, implementation can supply the missing judgment by default. The mapping becomes operational because it can be executed, rather than because the semantic decision has been accepted by the party entitled to make it. At that point, the process has moved information forward before resolving who controls its meaning.
This distinction explains why technical verification cannot settle the semantic question. A relationship may be encoded exactly as specified and still remain disputed if no recognized decision authority supports the specification itself. Execution review determines whether the system performed the stated transformation. It does not establish whether the stated transformation represents the accepted interpretation. [1] That separation matters because a technically correct result can still carry a semantic treatment that is unauthorized or inconsistent with another accepted judgment.
Disagreement exposes the authority problem
The weakness becomes more visible when different participants reach different equivalence determinations for the same representations. Transformation mechanics cannot decide which interpretation should prevail because each competing treatment may be technically feasible. The unresolved variable is whose judgment controls. Without that basis, a mapping can be changed or challenged without a clear rule for deciding whether the prior treatment or the revised one is authoritative. The dispute then becomes part of the interoperability process because the meaning carried by the result depends on the unresolved decision.
That problem is visible in research-use LOINC mappings from seven Korean university hospitals. A cross-institutional evaluation found that 18% of codes mapped to frequently requested test names were inconsistent. Only two institutions had standard operating procedures for LOINC mapping, while disputed classifications in the evaluation were resolved through consensus among clinical pathologists and terminology experts. The study does not assign formal approval authority, but it shows that technically implemented mappings can still require a separate semantic judgment to determine which treatment the evaluation accepts. [2]
Semantic change can also occur without any failure in data transfer. If one representation is treated as equivalent to another, the judgment may collapse meanings that should remain separate. A different determination may preserve a distinction that an earlier mapping removed. The technical movement of information can remain intact across both treatments even as the accepted meaning changes. Oversight therefore cannot infer semantic stability from successful execution alone.
Oversight needs decision traceability
Review must be able to distinguish implementation evidence from authority evidence. A system record can show that a mapping was applied, and a review step can confirm that the transformation matched its specification. Neither fact identifies who had the authority to approve the equivalence itself unless that decision is separately attributable. [3] When the accepted result is later questioned, the absence of that attribution prevents a clean determination of whether the issue arose from faulty execution or from an unsettled semantic judgment.
The same distinction matters when responsibility is assigned. If a disputed result is traced only to the person or system that performed the mapping, responsibility for semantic meaning may be assigned to the wrong operational function. The implementer may have executed an approved decision, or may have acted where approval authority was undefined. Those conditions are materially different. Leadership cannot distinguish them from technical completion records alone. It needs traceability from the implemented treatment back to the authority whose determination made that treatment acceptable.
This traceability also affects how leaders interpret apparent change. An increase in disputed mappings does not by itself show that semantic conditions have changed. It may reflect more effective identification or review of an equivalence already in use. A change in the accepted equivalence determination is different because it alters the semantic basis of interoperability. Separating those situations is necessary when leaders assess whether they are seeing a change in meaning or a change in how effectively disagreement is detected.
The unresolved control point is visible wherever interoperability depends on an equivalence decision that can be executed before its authority is clear. Technical completion can show that a transformation works, but it cannot establish whose semantic judgment is entitled to govern. Until that authority is identifiable and the implemented treatment can be traced back to it, the interoperable result remains open to challenge even when the underlying mechanics perform exactly as intended.
Sources:
[1] Health Level Seven International (HL7). FHIR R5 ConceptMap. FHIR specification, version 5.0.0 (R5), Standards for Trial Use. HL7 technical specification
[2] Korean Academy of Medical Sciences. Gaps and Similarities in Research Use LOINC Codes Utilized in Korean University Hospitals: Towards Semantic Interoperability for Patient Care. Peer-reviewed original research article, published online October 18, 2024; journal issue January 6, 2025. Peer-reviewed journal article
[3] NHS England. Mapping creation, assurance, and governance. SNOMED CT PaLM Mapping Best Practice implementation guidance, last edited May 22, 2025. NHS England implementation guidance