Format

Send to

Choose Destination
See comment in PubMed Commons below
Int J Oral Maxillofac Surg. 2008 Dec;37(12):1080-8. doi: 10.1016/j.ijom.2008.06.008. Epub 2008 Jul 30.

A comprehensive classification of mandibular fractures: a preliminary agreement validation study.

Author information

1
Institute of Radiology, Zofingen Hospital AG and Hightech Research Center University of Basel, Switzerland. carlos.buitrago@spitalzofingen.ch

Abstract

This study evaluates a comprehensive classification system for mandibular fractures based on imaging analysis. The AO/ASIF scheme, defining three fracture types (A, B, C), three groups within each type (e.g. A1, A2, A3) and three subgroups within each group (e.g. A1.1, A1.2, A1.3) with increasing severity from A1.1 (lowest) to C3.3 (highest) was used. The mandible is divided into two vertical units (I and V), two lateral horizontal units (II and IV) and one central unit (III) comprising the symphyseal and parasymphyseal region. Type A fractures are non-displaced, type B are displaced and type C are multifragmentary/defect injuries. Groups and subgroups are further defined in the classification system. Two classification sessions using semi-automatic software with 7 and 9 surgeons were performed to evaluate 100 fracture cases in the first session and 50 in the second. Inter-observer reliability and individual rater's accuracy were evaluated by kappa coefficient and latent class analysis, respectively. The analysis of inter-observer agreement for the detailed coding showed kappa coefficients around 0.50 with higher agreement among raters in the vertical units. This system allows standardization of documentation of mandibular fractures, although improvement in the definition of categories and their application is required.

PMID:
18672348
DOI:
10.1016/j.ijom.2008.06.008
[Indexed for MEDLINE]
PubMed Commons home

PubMed Commons

0 comments
How to join PubMed Commons

    Supplemental Content

    Full text links

    Icon for Elsevier Science
    Loading ...
    Support Center