Deep infection and fracture healing in immediate and delayed locked intramedullary nailing for open femoral fractures.

Fifty-nine patients with 61 open femoral fractures were treated with immediate locked intramedullary (IM) nailing (group 1; n=15), delayed IM nailing following nonoperative treatment (group 2; n=42), and delayed IM nailing following external fixation (group 3; n=7). Sixteen fractures were Gustilo type I, 28 were type II, 7 were type IIIA, 6 were type IIIB, and 4 were type IIIC open fractures. Four (6.6%) deep infections occurred. Significant differences existed in the deep infection rate (DIR) between types I and II and all type III fractures (2.3% for types I and II versus 17.6% for type III). The deep infection rate did not differ significantly among the nailing groups (13.3%, 2.6%, and 15.3% for groups 1, 2, and 3, respectively), nor did the deep infection rate correlate with the degree of fracture comminution, the existence of polytrauma or polyskeletal trauma, or preexistence of superficial or pin-site infections. Seven (11.7%) of these fractures resulted in nonunion, excluding one secondary amputation; the nonunion rate correlated with fracture location. There were no significant differences in the mean fracture healing times between any of the nailing groups. These results suggest that IM nailing for the treatment of type III open femoral fractures should be considered carefully, regardless of whether it is performed immediately or delayed.