Glenohumeral arthritis as a risk factor for proximal humerus nonunion.

BACKGROUND

Proximal humerus fractures are common injuries and nonsurgical treatment has proven to yield good to excellent clinical results. A small percentage of these fractures go on to delayed or nonunion and the incidence and risk factors for this complication are poorly understood. We hypothesize that adjacent joint stiffness of the glenohumeral joint might lead to an increased rate of nonunion for proximal humerus fractures.

METHODS

A search of the entire Medicare database from 2005 to 2011 was performed to identify 38,754 patients who sustained a proximal humerus fracture including 13,802 with co-existing ipsilateral shoulder osteoarthritis (OA) and 24,952 with co-existing diagnosis of rheumatoid arthritis (RA). A cohort of 301,987 patients served as a control. Medical co-morbidities and fracture complications were compared between the cohorts.

RESULTS

The incidence of delayed union at 3 and 6 months for OA and RA groups were significantly increased compared to control at 0.79% and 1.74%, and 0.67% and 1.86%, respectively (p < 0.001). Nonunion rates were also significantly increased (p < 0.001) in the OA and RA cohorts at 9 months and 1 year with incidences of 2.39%, 2.89% and 2.59% and 3.08%, Respectively. The incidence of nonunion in the control cohort at the same time points was 1.13% and 1.35%.

CONCLUSIONS

The coexistence of shoulder OA or a diagnosis of RA nearly doubled in the incidence of proximal humerus nonunion. This is the first study investigating this association in the shoulder and should be considered along with traditional nonunion risk factors in the treatment algorithm.