Introduction: Prostheses radiolucent lines are currently used to evaluate the components fixation. The objective of this study is to determine concordance and reproducibility of humeral stem radiolucent lines evaluation.

Material and method: Five observers evaluated 64 x-ray belonging to 32 pairs (1 antero-posterior view and 1 outlet view) obtained from 16 shoulder prostheses. 16 x-ray pairs were obtained immediately after surgery and 16 one year after surgery. Evaluation in four degrees of radiolucent width for each of the 7 zones that the humeral component was divided for. Evaluation of the component-cement interface and the cement-bone interface. Each observer made two evaluations of the 64 x-ray separated in 6 weeks. Statistics : index kappa with quadratic weighting.

Results: Intra-observer results: mean kappa index for component-cement interface: 0,3274. mean kappa index for cement-bone interface: 0,5269. Inter-observer results: mean kappa index for component-cement interface: 0,1242. mean kappa index for cement-bone interface: 0,2478. Evaluation of 2 pairs of x-ray of the each prostheses taken in a period of 1 year: component-cement interface: mean of 91,67% of plausible results, cement-bone interface: mean of 80,2% of plausible Results:

Conclusions:

- low kappa index of reproducibility (0,3274–0,5269) of humeral component radiolucent lines evaluation for component-cement interface as well as for cement-bone interface.