Classification systems for open fractures help the surgeon to follow guidelines for treatment, to predict the prognosis, and to allow comparison of results. The systems of Gustilo and Anderson and of Oestern and Tscherne are most widely used. Although both systems have undergone several revisions, the crucial factors have not changed. They deal with the size of the wound, level of contamination, extent of soft tissue injury, and comminution of bone.

In recent years additional classification systems have been created to classify severe open fractures (type III), mainly of the lower extremity. The Mangled Extremity Severity Score (MESS) became the most practicable score for establishing a dividing line between possible functional limb salvage and the need for primary amputation.

Management: The principle of surgical debridement of all necrotic tissues has to be followed. Nowadays, soft tissue coverage and restoration of lost bone can be achieved secondarily by different means. However, the method of primary skeletal stabilization has a high impact on the final outcome after open diaphyseal fractures.

Upper extremity: Most open fractures of the humerus and forearm can be stabilized sufficiently with plates. Because of the good soft tissue coverage of the humerus and proximal forearm and the good blood supply of this region, coverage of implants can usually be achieved in cases with vast soft tissue destruction and severe bone comminution. External fixation with the option of primary shortening and secondary bone transport is a good alternative in cases of humeral fracture. Indications for intramedullary nailing are limited to minor open fractures that do not require radial nerve exploration. At the distal forearm, the thin soft tissue layer and the necessity of two approaches often make coverage of plates impossible. External fixation is the method of choice in these cases.

Lower extremity: Femoral fractures are a domain for intramedullary nailing. The indications for nailing are restricted more by systemic factors rather than by the extent of soft tissue injury. The advantages of intramedullary nailing are based on the closed surgical procedure that leaves the actual fracture site untouched. Static interlocking ensures axial and rotational stability and warrants early functional treatment and weight-bearing mobilisation. External fixation is indicated as emergency treatment, and plating should be restricted to condylar and supracondylar fractures.

Tibial fractures leaving 7 cm intact bone proximally and 5 cm distally can be sufficiently stabilized by intramedullary locking nails. In cases of open fractures, small diameter nails can be inserted in the unreamed technique without deterioration of the endosteal vascular supply. This method has better reported results concerning time to union, axial alignment, joint function, and infection rates in comparison with the use of external fixation devices. However, external fixation is an adequate method, especially in cases with extreme proximal and distal fractures. If insufficient stabilization or delayed union with the use of unreamed nails or external fixators occurs, reamed nailing can be performed in a second step with good results. Plating should be restricted to tibial head and pylon fractures.

A multicentre trial of four Level One trauma centres retrospectively analysed complications and odds for complications in complex open and closed tibial fractures stabilised by unreamed, small diameter nails.

467 tibial fractures were included in the study. There were 52 proximal fractures (11.1%), 219 mid-shaft fractures (46.9%), and 196 distal fractures (42%). Breakdown into different AO/OTA groups revealed 135 type A fractures (28.9%), 216 type B fractures (46.3%), and 116 type C fractures (24.8%). 265 were closed fractures (56.7%) and 202 were open fractures (48 Gustilo grade I (10.3%), 80 grade II (17.1%), and 74 grade III (15.9%).

Analysis revealed five (1.1%) deep infections (with a 5.4% rate of deep infections in Gustilo grade III open fractures), 43 delayed unions (9.2%), and twelve (2.6%) non-unions. Compartment syndromes occurred in 62 cases (13.3%), screw fatigue in 47 cases (10%), and fatigue failure of the tibial nail in three cases (0.6%). Fracture distraction of more than 3 mm should not be tolerated when stabilizing tibial fractures with unreamed, small diameter nails as this increases the odds to acquire delayed union by twelve times (p < 0.001), and the odds to acquire non union by four times (p = 0.057).

There was a significant increase of complications in the group of grade III open fractures (p < 0.001), AO/OTA type C fractures (p = 0.002), and to a lesser extent in distal fractures. However, the rate of severe complications resulting in major morbidity was low compared to other methods of stabilisation in these severe fractures.