header advert
Orthopaedic Proceedings Logo

Receive monthly Table of Contents alerts from Orthopaedic Proceedings

Comprehensive article alerts can be set up and managed through your account settings

View my account settings

Visit Orthopaedic Proceedings at:

Loading...

Loading...

Full Access

General Orthopaedics

A review of orthopaedic classifications; are they justified in their use?

British Orthopaedic Association 2012 Annual Congress



Abstract

Introduction

Classification systems are used throughout Trauma and Orthopaedic (T&O) surgery, designed to be used for communication, planning treatment options, predicting outcomes and research purposes. As a result the majority of T&O knowledge is based upon such systems with most of the published literature using classifications. Therefore we wanted to investigate the basis for the classification culture in our specialty by reviewing Orthopaedic classifications and the literature to assess whether the classifications had been independently validated.

Methods

185 published classification systems within T&O were selected. The original publication for each classification system was reviewed to assess whether any validation process had been performed. Each paper was reviewed to see if any intra-observer or inter-observer error was reported. A PubMed search was then conducted for each classification system to assess whether any independent validation had been performed. Any measurement of validation and error was recorded.

Results

Four of the 185 classifications (2.1%) had a validation process described in the initial paper that introduced that classification to the literature. 54 (29.1%) of the classifications had a related study that independently assessed the classification for validity. Of these 54, only 10 (18.5%) demonstrated either an intra-observer or inter-observer error that is described as excellent (kappa score >0.8). Only 2 classification systems of the 54 (3.7%) were shown to have both intra-observer and inter-observer errors as excellent, meaning only 2 of the 185 classification systems reviewed (1.1%) have been shown to be highly reproducible.

Conclusion

Over 70% of classification systems in T&O have never been independently validated and assessed for intra-observer and inter-observer error. Of those that have, only 2 are excellent. Such a finding raises questions about the use of classification systems within T&O and queries the use of classification systems in the literature as part of evidence based medicine.