Intra- and interobserver agreement in the interpretation of navicular bones on radiographs and computed tomography scans.
Authors: Groth, May, Weaver, Weller
Journal: Equine veterinary journal
Summary
# Editorial Summary: Navicular Radiography and CT Interpretation Consistency Despite published standardised criteria for evaluating navicular disease on radiographs, Groth and colleagues questioned whether these guidelines actually produced reliable, reproducible results across different veterinarians—a fundamental requirement for any diagnostic tool used in clinical decision-making. Three observers independently assessed radiographs and CT scans of 60 cadaver navicular bones using established classification criteria, with one observer reassessing 30 studies to measure both within-observer consistency (intraobserver agreement) and between-observer consistency (interobserver agreement), using kappa statistics to quantify reliability. Whilst observers expressed greater confidence in their pathological assessments on CT compared to radiographs, this confidence was not reflected in improved agreement—interobserver reliability for both modalities only reached 'fair' levels, and intraobserver agreement remained merely 'moderate', with considerable variation in how lesions were actually classified. These findings have significant implications for practice: diagnostic criteria that lack robust agreement cannot reliably guide prognosis or treatment decisions, and practitioners should exercise caution in placing excessive diagnostic weight on radiographic or CT findings without correlating them carefully with clinical signs and additional diagnostic information. The authors' identification of more precise evaluation criteria represents a step towards improving consistency, but further refinement of navicular assessment protocols remains essential before imaging studies alone can be confidently used to support high-stakes management decisions.
Read the full abstract on PubMed
Practical Takeaways
- •Radiographic and CT findings for navicular evaluation should be interpreted cautiously, as observers show only fair to moderate agreement—consider obtaining second opinions before making treatment decisions based on imaging alone
- •CT imaging provides greater diagnostic certainty than radiography, but this confidence does not guarantee superior agreement between observers, so clinical signs and lameness localization remain critical
- •Standardized evaluation criteria exist, but their application varies considerably; ensure your veterinary diagnostician is using consistent criteria and ideally correlating imaging findings with physical examination findings
Key Findings
- •Interobserver agreement for radiographic evaluation was fair (mean kappa), while CT evaluation also showed fair agreement despite expectation of superiority
- •Intraobserver agreement was moderate for both radiographic and CT evaluation, indicating inconsistency even within individual observers
- •Pathological changes were identified with greater certainty on CT scans compared to radiographs, but this did not translate to improved diagnostic agreement
- •Considerable variation in classification of navicular lesions between and within observers challenges the reliability of these imaging modalities for diagnostic and prognostic purposes