What are We Measuring in Anatomy Education? An Exploratory Mapping from Anatomical Learning to Clinical Competence

Abstract

Anatomical knowledge is a necessary foundation for clinical practice, but clinical competence develops progressively through its integration with other foundational sciences and supervised clinical experience. The extent to which anatomy education research evaluates downstream application of anatomical learning remains unclear. In this Perspective, supported by an exploratory metadata-based mapping, we examined whether reported outcomes are concentrated at proximal educational levels or extend to clinical performance and healthcare impact. A Scopus search conducted on 2 June 2026 retrieved 19,956 records. After duplicate removal, restriction to original articles, and application of predefined anatomy-education filters, 8088 articles were retained. Titles, abstracts, and author keywords were classified using conservative hierarchical rules into perception/descriptive outcomes, anatomical knowledge, applied anatomical skills, clinical reasoning/professional behavior, patient/healthcare impact, or unclear/not classified. Outcomes related to clinical reasoning/professional behavior and patient/healthcare impact, the two highest translational categories, were identified in 3.20% of the refined corpus and 4.24% of classified records. Signals corresponding to applied anatomical skills, clinical reasoning/professional behavior, or patient/healthcare impact increased over time, whereas the odds of outcomes specifically related to clinical reasoning/professional behavior or patient/healthcare impact did not increase significantly. These findings do not indicate that anatomy education fails to contribute to clinical competence. Rather, they suggest that the literature more often documents proximal or applied outcomes than direct downstream transfer. Outcome selection should therefore be aligned with the stated purpose and curricular stage of each intervention, while future longitudinal and clinically integrated studies examine anatomy’s contributory role within the broader development of competence.

Anatomy Education and the Promise of Clinical Competence

Anatomy remains one of the foundational sciences of medical and health professions education.¹ Its value is not limited to memorizing structures, regions, or organ systems; rather, anatomical knowledge is expected to support clinical examination, diagnostic interpretation, procedural orientation, surgical reasoning, imaging literacy, and patient safety.^1–3 In this sense, anatomy education carries a translational promise. What is learned in the anatomy classroom, laboratory, or digital environment should ultimately help future professionals think, decide, and act more safely in clinical contexts.^2,4

This expectation has become increasingly important as medical curricula have shifted toward integration, competency-based education, reduced contact hours for traditional basic sciences, and greater use of digital, simulation-based, and clinically oriented teaching strategies.^5–7 Competency-based medical education is an outcomes-oriented approach in which progression is organized around the demonstration of abilities rather than time alone.⁸ Within this approach, anatomical competence may be understood as the ability to apply anatomical knowledge appropriately within professional and clinical contexts, rather than merely to recall structures.⁹ Anatomy-specific competency work has similarly emphasized clinically relevant application, alignment between outcomes and assessment, and collaboration with clinicians when defining essential anatomical knowledge.^9–11

Four related constructs should therefore be distinguished. Anatomy is foundational knowledge; an anatomy education intervention is an instructional exposure; anatomical competence is the context-appropriate application of anatomy; and eventual clinical performance is an integrated outcome produced through multiple foundational sciences, clinical knowledge, supervised experience, and workplace conditions. Our argument is not that anatomy education alone should produce or explain clinical competence. Rather, when a study makes claims about clinical transfer, the selected outcomes should be capable of supporting those claims.

Yet, a key question remains insufficiently addressed. What outcomes does anatomy education research actually measure, and how closely do those outcomes correspond to the claims made for a given intervention? Educational innovation alone is not equivalent to clinical transfer.⁴ A new model, simulation tool, dissection strategy, or digital platform may appropriately aim to improve satisfaction, engagement, knowledge, or practical skill, especially during foundational training. The methodological concern arises only when proximal outcomes are interpreted as direct evidence of improved clinical reasoning, professional behavior, or patient care.^4,12

The Basic-to-Clinical Translational Gap in Anatomy Education

In this Perspective, the basic-to-clinical translational gap refers to a possible misalignment between outcomes commonly assessed in anatomy education studies and the downstream clinical competencies that anatomical education is expected to contribute to, rather than independently produce.

This gap does not imply that knowledge-based or perception-based studies are unimportant. Early outcomes such as learner satisfaction, perceived usefulness, anatomical knowledge, and skill acquisition may be appropriate and necessary, particularly in preclinical or foundational courses. They become insufficient only when they are used to support claims about clinical transfer that was not directly evaluated. The central issue is therefore coherence between the intervention’s stated purpose, the learner’s curricular stage, the outcome selected, and the conclusion drawn.^4,13

This distinction is consistent with established frameworks in medical education assessment. Miller’s pyramid conceptualizes clinical competence as a progression from “knows” and “knows how” to “shows how” and “does”, emphasizing that knowledge is the base of clinical performance.¹² Kirkpatrick-type models distinguish learner reaction, learning, behavior, and results, although their use in medical education requires caution and contextual interpretation.^4,13 Entrustable professional activities extend attention at the workplace “does” level by linking observable professional tasks to decisions about the supervision a learner requires.¹⁴ These frameworks collectively show that satisfaction or knowledge is not equivalent to clinical behavior or healthcare impact, while also recognizing that lower-level outcomes remain legitimate when they match the educational purpose.

In anatomy education, this distinction is especially important. A student may correctly identify the branches of a nerve plexus in an examination but still require physiology, pathology, clinical exposure, and supervised practice before interpreting a neurological deficit. A learner may benefit from a three-dimensional model without the intervention being designed to demonstrate patient-level effects. Similarly, improved procedural confidence does not necessarily equal safer procedural performance.^12,15 Anatomy education research therefore needs clearer differentiation between learning anatomy, applying anatomy in practical or simulated contexts, and evaluating anatomy’s contribution to integrated clinical performance.

A Hierarchical Framework for Translational Outcomes

To examine this issue in a structured way, we used a five-level hierarchical framework for classifying outcomes in anatomy education research, informed by established models of clinical competence, educational evaluation, and translational outcomes in medical education^4,12,16,17 (Figure 1). The framework is not intended to replace Kirkpatrick, Miller, Moore, or EPA-based assessment. Kirkpatrick focuses on the effects of educational programs, Miller organizes levels of competence and performance, and EPAs operationalize trustworthy performance in workplace activities. Our framework serves a narrower descriptive purpose, to classify the translational distance between the outcome identifiable in an anatomy education record and downstream clinical practice. It does not assess study quality, determine whether one level is intrinsically superior, or attribute clinical performance exclusively to anatomy.

Figure 1.

Hierarchical framework for classifying outcomes in anatomy education research. The framework organizes measured outcomes according to their translational distance from clinical practice. Levels 0–1 represent proximal outcomes, level 2 applied anatomical skills, levels 3–4 higher translational depth, and unclear/not classified records contain insufficient metadata for reliable assignment. The framework describes outcome type and does not rank methodological quality or imply that higher levels are appropriate for every curricular stage.

Level 0: Perception/descriptive outcomes: These include satisfaction, attitudes, perceived usefulness, preferences, self-confidence, motivation, and descriptive reports of curricular experiences. These outcomes capture learner or educator response but do not directly demonstrate learning or performance.

Level 1: Anatomical knowledge: These include written tests, quizzes, multiple-choice questions, structure identification, knowledge retention, academic performance, or examination scores. This level assesses whether students learned anatomical content.

Level 2: Applied anatomical skills: These include surface anatomy, imaging interpretation, ultrasound training, dissection skills, simulation-based procedural performance, spatial ability, three-dimensional modeling, surgical simulation, or other applied tasks. This level evaluates whether learners can use anatomy in practical or simulated contexts.

Level 3: Clinical reasoning/professional behavior: These include objectively assessed diagnostic reasoning, clinical decision-making, clinical competence, Objective Structured Clinical Examination (OSCE)-type performance, workplace-based assessment, transfer to clinical practice, or performance in clinical cases. This level captures outcomes in which anatomical knowledge is evaluated as one contributor to integrated reasoning or professional behavior.

Level 4: Patient/healthcare impact: These include patient safety, diagnostic or procedural errors, complication rates, adverse events, quality of care, or healthcare outcomes explicitly evaluated in relation to an anatomy-based educational or training intervention. Mere mention of patient benefit, clinical relevance, or complications was not sufficient for this level.

For analytic purposes, Levels 0 and 1 were considered low translational depth, Level 2 intermediate translational depth, and Levels 3 and 4 high translational depth. This framework does not rank studies by quality; rather, it classifies the translational distance between the measured outcome and clinical practice.^4,13

Exploratory Scopus-Based Metadata Mapping

To support this Perspective, we conducted an exploratory metadata-based mapping of Scopus-indexed anatomy education literature. The empirical component was designed to identify broad outcome-reporting patterns and generate hypotheses; it was not designed as a systematic or scoping review of full-text evidence. It therefore did not include full-text eligibility assessment, critical appraisal, duplicate study selection, or evidence synthesis. Preferred Reporting Items for Systematic reviews and Meta-Analyses extension for Scoping Reviews (PRISMA-ScR) and Joanna Briggs Institute (JBI) scoping-review methodology were not adopted as reporting frameworks because those procedures would imply a different review design.^18,19 We nevertheless expanded the reporting of the search strategy, filtering process, classification rules, record flow, and methodological limitations. A future scoping or systematic review with full-text assessment would be required to confirm study-level classifications and exact prevalence estimates.

The search was performed in Scopus on 2 June 2026 and was constructed from two Medical Subject Headings (MeSH)-derived conceptual domains:²⁰ one related to anatomy and anatomical sciences and another related to medical education. The anatomy-related domain included terms such as “Anatom*”, “Visible Human Project”, “Osteolog*”, “Embryolog*”, “Teratolog*”, “Histolog*”, and “Neuroanatom*”. The medical education domain included “Medical Education*”, “Internship and Residency”, “Medical Internship*”, “Internship”, “House Staff”, “Residenc*”, “Medical Residenc*”, “Teaching Round*”, “Clinical Round*”, “Morning Report*”, “Education”, “Educational Activit*”, “Literacy Program*”, “Workshop*”, “Training Program*”, “Teaching”, “Computer User Training*”, “Educational Model*”, “Problem-Based Learning*”, “Remedial Teaching*”, “Self-Directed Learning as Topic”, and “Study Guides as Topic”. The two domains were combined using the Boolean operator AND. The search was restricted to records published from 2000 onwards, records in final publication stage, and document types classified by Scopus as articles or reviews. The complete search syntax and eligibility restrictions are provided in the Supplementary Material.

The strategy was intentionally broad to capture the continuum of anatomy-related education across undergraduate, postgraduate, residency, simulation, remedial, and professional-training contexts. This breadth was appropriate for exploratory mapping but introduced curricular heterogeneity, because the intended role of anatomy and the feasibility of evaluating clinical transfer differ substantially across stages of training.

The initial search retrieved 19,956 records. Duplicate removal by Electronic Identifier left 19,944 unique records. Reviews were excluded from outcome classification, leaving 17,441 original articles. Predefined phrase, proximity, and noise-exclusion rules were then applied to identify a refined anatomy education subcorpus of 8088 articles; 9353 articles did not meet these refinement rules. The record-identification and classification pathway is shown in Figure 2. This broad filtering logic is related to, but analytically distinct from, our previous exploratory study of dissection, simulation, and hybrid teaching modalities, which addressed a different corpus, classification target, and research question.²¹

Figure 2.

Flow of record identification, refinement, and metadata-based outcome classification. The Scopus search retrieved 19,956 records; 19,944 unique records remained after duplicate removal, 17,441 original articles remained after excluding reviews, and 8088 articles met the refined anatomy education criteria. The final corpus was classified as level 0 (n=1763), level 1 (n=1240), level 2 (n=2852), level 3 (n=140), level 4 (n=119), or unclear/not classified (n=1974).

Corpus processing was implemented in R. Titles, abstracts, author keywords, and index keywords were standardized for filtering. The outcome classifier used titles, abstracts, and author keywords. Missing text fields were replaced by empty strings for rule execution, but no missing content was imputed. In the refined corpus, all records contained a title and abstract; author keywords were absent in 1132 records (14.00%), and index keywords were absent in 1954 records (24.16%).

The classifier used predefined term dictionaries for perception, anatomical knowledge, applied anatomical skills, clinical reasoning/professional behavior, and patient/healthcare outcomes. Contextual exclusions reduced common false positives, including “confidence interval”, “teaching hospital”, and non-educational uses of training terminology. A hierarchical rule assigned the highest qualifying outcome level identified in the available metadata, while patient-level terms required an explicit educational context. Records without sufficiently explicit measured-outcome evidence were retained as unclear/not classified rather than removed. This category therefore indicates insufficient metadata for reliable assignment, not the absence of an educational outcome.

Outcome-level assignments recorded in the refined dataset were used for the final statistical analyses. These analyses included descriptive distributions, temporal summaries, and logistic regression models. A conservative model treated unclear records as non-high-translational, whereas a sensitivity model was restricted to records assigned to one of the five outcome levels. Because 2026 was incomplete at the search date, temporal analyses were restricted to publications from 2000 to 2025.

The raw Scopus export, refined outcome-classified dataset, classification rules, and analytic code are identified in the Data sharing statement and accompanying Supplementary Material.

What the Snapshot Shows

The refined corpus showed a strong concentration of anatomy education research in low and intermediate translational levels. Among 8088 articles, 2852 articles were classified as Level 2 applied anatomical skills, representing 35.26% of the corpus. Level 0 perception/descriptive outcomes accounted for 1763 articles, or 21.80%, and Level 1 anatomical knowledge accounted for 1240 articles, or 15.33%. A substantial proportion of articles, 1974 records or 24.41%, remained unclear or not classifiable based on metadata alone (Figure 3).

Figure 3.

Outcome levels and translational depth in anatomy education research. (a) Overall distribution across outcome levels. (b) Distribution by translational depth. (c) Proportional distribution across publication periods from 2000 to 2025. (d) Annual number of articles by translational depth. Temporal analyses excluded 2026 because the year was incomplete at the search date. The figure shows increasing applied outcome signals and a declining proportion of unclear records, while level 3–4 signals remained comparatively uncommon.

By contrast, high-translational outcomes were uncommon. Level 3 clinical reasoning/professional behavior accounted for 140 articles, or 1.73%, and Level 4 patient/healthcare impact accounted for 119 articles, or 1.47%. Together, high-translational outcomes represented 259 articles, corresponding to 3.20% of the full refined corpus and 4.24% of outcome-classified articles.

The distribution by translational depth further illustrates the pattern (Figure 3). Low translational depth outcomes represented 37.13% of the corpus, intermediate translational depth 35.26%, unclear/not classified records 24.41%, and high translational depth 3.20%. Thus, even when applied anatomical skills were common, records with explicit metadata signals of clinical reasoning, professional behavior, or patient/healthcare impact were comparatively uncommon.

Temporal analysis provided additional insight. In a conservative logistic regression model treating unclear records as non-high-translational, the odds of Level 3–4 outcome signals did not significantly increase over time (OR per decade: 1.06; 95% CI: 0.86–1.32; p=0.580). A sensitivity model restricted to classified records similarly showed no significant increase (OR per decade: 0.88; 95% CI: 0.71–1.10; p=0.259). By contrast, the probability of applied or clinical outcome signals, defined as Levels 2–4, increased substantially over time (OR per decade: 1.53; 95% CI: 1.42–1.66; p<0.001). Figure 3 also shows a proportional decline in unclear records and growth in Level 2 across publication periods. These patterns indicate a positive shift toward more applied outcome reporting, while direct assessment of clinical reasoning, workplace behavior, or patient-related impact remained comparatively uncommon. The observed trend cannot be attributed causally to any single guideline or curricular reform.

Translational Coherence as a Future Direction

These findings suggest that anatomy education research would benefit from greater alignment between the stated purpose of an intervention, its curricular position, the outcome selected, and the claims made from that outcome.^16,22,23 Clinical competence is not the product of anatomy alone; it emerges through the interaction of multiple foundational sciences, clinical knowledge, supervised experience, learner development, and workplace context. Accordingly, the framework should be used to describe what was measured, not to demand that every anatomy intervention reach the highest level.

Anatomy education studies should make explicit the proposed pathway from educational exposure to the outcome that is feasible and appropriate for the learners’ stage.^2,16,24 A radiological anatomy course for early learners may appropriately evaluate image interpretation or spatial understanding, whereas a clinically integrated intervention may also assess case-based reasoning, OSCE performance, or workplace application. A procedural anatomy study should distinguish confidence, simulated performance, supervised clinical behavior, and patient safety. The aim is not to require the same endpoint across settings, but to ensure that the measured endpoint supports the claim being made.

This does not mean that every anatomy education study must measure patient outcomes or direct clinical transfer. Such outcomes may be infeasible, inappropriate, or temporally premature in foundational courses, and medical education scholars have cautioned against treating distal outcomes as universally superior endpoints.²⁴ A balanced evidence ecosystem should include perception, knowledge, applied skills, reasoning, behavior, and healthcare impact, with each level interpreted according to the intervention’s purpose and curricular context.^4,16,24

Recommendations for Future Research

First, anatomy education studies should specify their intended outcome level and curricular stage before selecting measures.^12,16,24 If the goal is learner engagement, perception-based outcomes may be appropriate. If the goal is anatomical learning, knowledge or identification assessments are appropriate. Claims of applied skill should be supported by practical or simulated performance, while claims of clinical transfer require case-based reasoning, OSCE-type assessment, workplace observation, or another clinically integrated measure.^25–27

Second, anatomy education researchers should distinguish clinical framing from clinical measurement. A teaching intervention may be described as clinically relevant, but unless clinical reasoning, behavior, or patient-related outcomes are measured, the study should not be interpreted as evidence of clinical transfer.^4,12,24

Third, future studies should develop and validate anatomy-specific measures that connect foundational knowledge to its context-appropriate application.^9,28–30 These may include image interpretation, lesion localization, surface anatomy, procedural planning, anatomical variation, or integrated clinical cases. EPAs and workplace-based assessments may be useful at later stages, but they evaluate integrated professional activities and should not be treated as outcomes attributable to anatomy alone.¹⁴

Fourth, longitudinal and cross-curricular designs are needed when the research question concerns transfer beyond the anatomy course. Follow-up during clerkships, residency, integrated examinations, or workplace-based assessment can examine whether anatomical learning is retained and used alongside physiology, pathology, pharmacology, and clinical experience. These designs will require collaboration between anatomy educators, clinicians, assessment specialists, and curriculum leaders.^11,12,16,24

Fifth, future evidence synthesis should become more systematic.^24,31,32 Full scoping or systematic reviews should combine transparent searches with duplicate screening, full-text outcome assessment, curriculum-stage extraction, independent coding, and inter-rater reliability. The present metadata-based mapping is a hypothesis-generating starting point rather than a substitute for that work.^18,19

Limitations of the Exploratory Analysis

This Perspective has important limitations. The empirical snapshot was based on Scopus-indexed publications and therefore does not capture all anatomy education scholarship, including non-indexed journals, conference proceedings, institutional reports, dissertations, or local educational outputs.³³ The intentionally broad search incorporated heterogeneous curricular settings, including foundational courses, simulation, residency, remedial teaching, and professional training. Because curriculum stage was not available in a standardized form across records, we could not determine whether higher-level outcomes were concentrated in later clinical training. This limits comparisons across the corpus and should be addressed through full-text study-level review.

Outcome classification was based on titles, abstracts, and author keywords rather than full texts. Metadata may omit secondary outcomes, curriculum context, or details needed to distinguish clinical framing from measured clinical transfer. Records classified as unclear should therefore not be interpreted as lacking educational outcomes; they lacked sufficient metadata for reliable classification. Retaining them made this limitation visible, and the classified-only sensitivity analysis reduced dependence on how they were handled. Because classification was based on bibliographic metadata without full-text validation or independent adjudication, the exact proportions should be interpreted cautiously.

Despite these limitations, the mapping identifies a pattern worthy of further study: anatomy education research appears to have become more applied over time, while direct metadata signals of clinical reasoning, professional behavior, or patient-related impact remain comparatively uncommon. This pattern should guide more rigorous full-text synthesis rather than be interpreted as evidence that anatomy education itself is ineffective.

Conclusion

Anatomical knowledge is a necessary foundation for clinical practice, and anatomy education contributes to competence within a broader developmental pathway. Our exploratory Scopus-based mapping suggests that anatomy education research more often reports proximal or applied outcomes than direct downstream clinical or patient-related outcomes. This finding does not imply that all anatomy studies should measure clinical transfer, nor that anatomy alone determines clinical performance. It indicates that outcome selection and interpretation should be aligned with the intervention’s purpose, learner stage, and strength of the clinical claim. Future full-text, longitudinal, and cross-curricular research should examine when and how anatomical learning contributes to reasoning, performance, and patient care.

Funding Statement

Internal funds for strengthening research for APC payment were sourced from the Universidad Simón Bolívar, Vicerrectoría de Investigación, Extensión e Innovación, Barranquilla, Colombia.

Data Sharing Statement

The data and analytical materials supporting this Perspective are openly available in Zenodo (https://doi.org/10.5281/zenodo.21786679). The single consolidated Supplementary Material submitted with this revision contains the complete Scopus search strategy, descriptions of the deposited datasets, variable definitions, classification rules, a missing-metadata summary, and a description of the analytical workflow. The raw Scopus export, refined outcome-classified dataset, and R scripts are available in the repository.

Disclosure

The authors declare they have no competing interests in this work.