[Retrospective analysis of healing problems after reamed and unreamed nailing of femoral shaft fractures].

168 fractures of the femoral shaft treated by intramedullary nailing were analyzed retrospectively. From 1986-1992 116 fractures had been treated with the reamed AO universal nail (RFN) and from 1993-1996 52 fractures with the AO unreamed femoral nail (UFN). In 24% of the RFN-group and in 2% of the UFN-group (p < 0.0001) open reduction of the fracture had been necessary. The time to radiological consolidation was similar in both groups (18.1 weeks +/- 6.1 vs. 18.3 weeks +/- 5.7, [mean +/- SD]). Delayed unions were less frequent in the RFN-group than in the UFN-group (3% vs. 13%, p = 0.01). Non-unions occurred in the RFN-group in 4%, in the UFN-group in 8%, the difference is not statistically significant (p = 0.46). Fractures with impaired consolidation (delayed-unions and non-unions) in the RFN group were distributed randomly along the femoral diaphysis, whereas all 11 fractures with retarded healing in the UFN group were short transverse or oblique fractures localized immediately distal to the femoral isthmus. We believe that there is mainly a mechanical reason for this phenomen, in addition to fracture type and fracture localization the (insufficient) length of the unreamed nails might have impaired stability further. The different factors should be investigated in larger series. As a consequence we now treat transverse and short oblique fractures of diaphyseal femoral fractures distal to the femoral isthmus with a RFN whereas in other types and localizations of diaphyseal femoral fractures we continue to use the UFN with special attention to maximal nail diameter and length.