Introduction: The purpose of this study was to evaluate the impact of volume rendering 3D computed tomography reconstructions on the inter- and intraobserver reliability of the OTA/AO and Neer classifications in the assessment of proximal humerus fractures.

Material and Methods: Four observers with different levels of clinical training classified forty proximal humerus fractures according to the OTA/AO and Neer classifications. Three rounds of evaluation were performed and compared. First, fractures were classified on the basis of plain radiographs alone. Then, four weeks later, the combination of plain radiographs and computed tomography scans with conventional 3D SSD reconstructions was evaluated. Finally, four weeks later, the combination of plain radiographs, computed tomography scans, and 3D volume rendering reconstructions was assessed. These readings were repeated in a newly randomized order after an interval of twelve weeks to evaluate intraobserver reliability.

Results: Interobserver reliability for the AO/ASIF classification showed good interobserver reliability with plain radiographs (k=0,65) and two-dimensional CT scans with conventional three-dimensional (SSD) reconstructions (k=0,71). Interobserver reliability improved to excellent when the fractures were classified on the basis of 3D volume rendering reconstructions scans (k=0,84).

Intraobserver reliability of the OTA/AO classification was good with plain radiographs (k=0,70) and improved to excellent after adding three-dimensional SSD reconstructions (k=0,80) and three-dimensional VR reconstructions (k=0,88).

Interobserver reliability of the Neer classification was poor with plain radiographs (k=0,39) and moderate with two-dimensional CT scans and conventional three-dimensional (SSD) reconstructions (k=0,56) and improved to good with the addition of 3D VR scans (k=0,74). Intraobserver reliability for was poor with plain radiographs (k=0,34), good with three-dimensional SSD reconstructions (k=0,61), and excellent with three-dimensional VR reconstructions (k=0,80).

Conclusion: In this study, three-dimensional volume rendering computed tomography improved the inter- and intraobserver reliability of the AO/OTA and the Neer classifications in the assessment of proximal humerus fractures. In the opinion of the authors, 3D volume rendering CT-scans are a helpful tool for preoperative planning and classification of fractures of the proximal humerus.

We evaluated the impact of stereo-visualisation of three-dimensional volume-rendering CT datasets on the inter- and intraobserver reliability assessed by kappa values on the AO/OTA and Neer classifications in the assessment of proximal humeral fractures. Four independent observers classified 40 fractures according to the AO/OTA and Neer classifications using plain radiographs, two-dimensional CT scans and with stereo-visualised three-dimensional volume-rendering reconstructions. Both classification systems showed moderate interobserver reliability with plain radiographs and two-dimensional CT scans. Three-dimensional volume-rendered CT scans improved the interobserver reliability of both systems to good. Intraobserver reliability was moderate for both classifications when assessed by plain radiographs. Stereo visualisation of three-dimensional volume rendering improved intraobserver reliability to good for the AO/OTA method and to excellent for the Neer classification.

These data support our opinion that stereo visualisation of three-dimensional volume-rendering datasets is of value when analysing and classifying complex fractures of the proximal humerus.