Significant differences between local reporting and central assessment of radiological complications in a prospective, multicenter study about locking plate fixation of proximal humerus fractures.

Abstract

To compare radiological complication reporting outcomes undertaken by an independent review board and the responsible on-site study personnel in a multicenter study about locking plate fixation of proximal humeral fractures. prospective, multicenter study SETTING:: Level I trauma centers PATIENTS:: One-hundred and fifty patients with a radiological confirmed, closed, displaced proximal humeral fracture fixed with a locking plate and aged between fifty and ninety years. Re-evaluation of all radiological data according to pre-defined criteria after finished collection of all study radiographs by an independent review board. Differences in outcomes between the review board and the sites assessment were tested with a paired t-test. Inter-rater agreements between the central review board and sites assessments were estimated by means of kappa statistics. The review board revealed significant more radiological complications than the sites assessment (p = 0.006), except for the complication 'head necrosis'. The inter-rater agreement was slight to moderate in all calculated categories. To prevent underreporting of radiological complications by on-site assessment, the implementation of a complication review board, using pre-defined criteria, is recommended for clinical studies that focus on radiological complications.