Open Access Open Badges Research article

Indications for computed tomography (CT-) diagnostics in proximal humeral fractures: a comparative study of plain radiography and computed tomography

Christian Bahrs1, Bernd Rolauffs1*, Norbert P Südkamp2, Hagen Schmal2, Christoph Eingartner4, Klaus Dietz3, Philippe L Pereira5, Kuno Weise1, Erich Lingenfelter6 and Peter Helwig2

Author Affiliations

1 Klinik für Unfall- und Wiederherstellungschirurgie, BG-Unfallklinik Tübingen, Eberhard-Karls-Universität Tübingen, Schnarrenbergstr. 95, 72076 Tübingen, Germany

2 Department für Orthopädie und Traumatologie, Albert-Ludwigs-Universität Freiburg, Hugstetter Str. 55, 79106 Freiburg, Germany

3 Institut für Medizinische Biometrie, Eberhard-Karls-Universität Tübingen, Westbahnhofstr. 55, 72070 Tübingen, Germany

4 Caritas-Krankenhaus Bad Mergentheim, Klinik für Unfall- und Wiederherstellungschirurgie, Uhlandstr. 7 97980 Bad Mergentheim, Germany

5 Radiologische Klinik, Abt. für Diagnostische und Interventionelle Radiologie Universitätsklinikum Tübingen, Hoppe-Seyler-Straße 3, 72076 Tübingen, Germany

6 Northland Bone and Joint Institute, 2750 Clay Edwards Dr. 304, Kansas City, MO 64116, USA

For all author emails, please log on.

BMC Musculoskeletal Disorders 2009, 10:33  doi:10.1186/1471-2474-10-33

Published: 2 April 2009



Precise indications for computed tomography (CT) in proximal humeral fractures are not established. The purpose of this study was a comparison of conventional radiographic views with different CT reconstructions with 2 D and 3 D imaging to establish indications for additional CT diagnostics depending on the fractured parts.


In a prospective diagnostic study in two level 1 trauma centers, 44 patients with proximal humeral fractures were diagnosed with conventional X-rays (22 AP + axillary views, 22 AP + scapular Y-views) and CT (multi-planar reconstruction (MPR) and maximum intensity projection (MIP)) with 2 D and 3 D imaging. 3 observers assessed the technical image quality, the assessment of the relevant anatomical structures (2-sample-t-test) and the percentage of the osseous overlap of the proximal humerus (Welch-test) using a scoring system. The quality of the different diagnostic methods was assessed according to the number of fractured parts (Bonferroni-Holm adjustment).


There was significantly more overlap of the fractured region on the scapular Y-views (mean 71.5%, range 45–90%) than on axillary views (mean 56.2%, range 10.5–100%). CT-diagnostics allowed a significantly better assessment of the relevant structures than conventional diagnostics (p < 0.05) independently of the fracture severity (two-, three-, and four-part fractures).


Conventional X-rays with AP view and a high-quality axillary view are useful for primary diagnostics of the fracture and often but not always show a clear presentation of the relevant bony structures such as both tuberosities, the glenoid and humeral head. CT with thin slices technology and additional 3 D imaging provides always a clear presentation of the fractured region. Clinically, a CT should be performed – independently of the number of fractured parts – when the proximal humerus and the shoulder joint are not presented with sufficient X-ray-quality to establish a treatment plan.