Abstract
The management of proximal humeral fractures is determined by fragment displacement. Intra-operative and radiological classification systems have been proposed, namely AO and the Neer classifications to assess therapeutic options.
This study evaluates the usefulness of these classification systems by the creation of physical 3D models using a rapid prototyping technique avoiding the problems associated with 3D illusions on a 2D screen.
Seven consecutive patients with complex fractures of the proximal humerus were investigated using the data from multi-sliced spiral CT scans. Fractures associated with dislocation were excluded. The data from these CTs was segmented to reveal the anatomy of interest and converted to a stereolithographic format from which the physical models could be made of the proximal humerus via a laser guided filament deposition process. Further manipulation with software allowed angulation and displacements of fragments to be measured.
Inter-observer agreement: All models were assessed by three surgeons. A consultant with a special interest in shoulder surgery, a fellowship trained surgeon and a senior house officer in basic surgical training. Independent assessment of the fractures from the models was made using the Neer and AO classifications. In only 1 incidence did all 3 observers agree on the classification, in 5 incidences only two observers agreed and on 8 occasions none of the observers agreed. Indeed there were 9 occasions that at least one observer thought the fracture pattern could not be applied to a classification.
Fracture Patterns: Observation of the individual models together with measurements of angulation and displacement by further software analysis, demonstrated major subtypes namely valgus and varus angulation with minimal displacement of the greater tuberosity. Appreciation of the integrity of the medial hinge and buckling could be made in relation to the different fracture patterns.
This study highlights concerns on the validity of current classification systems. It also questions whether the existing systems reflect the pathophysiological subtypes of these fractures allowing comparison of surgical results in order to evaluate treatment options.
These abstracts were prepared by Mr Cormac Kelly. Correspondence should be addressed to him c/o British Orthopaedic Association, Royal College of Surgeons, 35-43 Lincoln’s Inn Fields, London WC2A 3PN.
