Reliability and interobserver variability of a grading system of ventricular distension in dogs

Adriana Czerwik; Martin Jürgen Schmidt; Agnieszka Olszewska; Steven Hinz; Kathrin Büttner; Daniela Farke

doi:10.3389/fvets.2023.1271545

Reliability and interobserver variability of a grading system of ventricular distension in dogs

Front Vet Sci. 2023 Nov 23:10:1271545. doi: 10.3389/fvets.2023.1271545. eCollection 2023.

Authors

Adriana Czerwik¹, Martin Jürgen Schmidt¹, Agnieszka Olszewska¹, Steven Hinz², Kathrin Büttner³, Daniela Farke¹

Affiliations

¹ Department of Veterinary Clinical Sciences, Small Animal Clinic, Neurosurgery, Neuroradiology and Clinical Neurology, Justus Liebig University, Giessen, Germany.
² Small Animal Veterinary Practice Westpark, Wettenberg, Germany.
³ Unit for Biomathematics and Data Processing, Faculty of Veterinary Medicine, Justus Liebig University, Giessen, Germany.

Abstract

Introduction: Internal hydrocephalus is the most common malformation of the central nervous system in dogs. Although the grades of ventricular distension have importance for long-term prognosis, there is no standard classification scheme describing the grade of the ventricular distension in dogs.

Materials and methods: Magnetic resonance imaging (MRI) scans from 147 dogs of various breed, sex, skull conformation, and weight were reviewed retrospectively and blinded between three observers. Based on objectively assessable morphologic characteristics, the lateral cerebral ventricles were graded as normal, minimally, mildly, moderately, severely enlarged or end stage (grade 0 to grade 5), respectively. Evans' index or the ventricle brain index was also measured in all animals. Interobserver agreement between a very experienced, experienced, and unexperienced person was evaluated by the Spearman coefficient and kappa tests. Additionally, correlation to the ventricle brain index was determined using the Spearman coefficient and F-tests.

Results: The Spearman correlation coefficient reached a very strong correlation (r = 0.97) between the experienced and very experienced observer and a strong correlation (r = 0.91) between the very experienced and unexperienced observer. The kappa value revealed excellent interobserver agreement between the very experienced and experienced observers (weighted kappa 0.91) and moderate between the very experienced and unexperienced observers (weighted kappa 0.75). The ventricular-brain index correlated (r = 0.94, Spearman coefficient test) with the grading system, indicating that a more elevated ratio was related to a more advanced degree of ventricular enlargement. The interobserver agreement with regard to the grade between the neurologist in training and a board-certified neurologist was excellent and between the board-certified neurologist and general practitioner achieved lower values.

Conclusion: The presented MRI-based grading of ventricular enlargement is a reliable and functional method for an objective grading of the ventricular system in dogs. Some experience in MRI and brain anatomy is needed for interpretation and grading.

Keywords: VPS; canine hydrocephalus; grading system; magnetic resonance imaging; ventriculomegaly.

Grants and funding

The author(s) declare that no financial support was received for the research, authorship, and/or publication of this article.