Background Distal humeral fracture is a severe damage around the elbow joint, and is often seen in young adults. It accounts for 2% of all adult fractures and about 50% of all humerus fractures. It′s one of the fractures that is difficult to deal with. The types of distal humeral fracture are divergent. Distal humerus fractures are often comminuted which make operative reduction difficult. Secondary loss of reduction and elbow ankylosis are common postoperative complications. All these difficulties make the distal humerus fracture one of the unresolved problems in fracture treatment. This study is to evaluate the clinical outcome of complex distal humeral fractures treated by operation.Methods (1) General data: twenty-four cases of operative treated distal humerus fractures in author′s hospital from January 2004 to December 2013 were included in this study. There were 15 males and 9 females, aging from 17 to 73, averaged 41. AO/OTA Classification: A3: 9 cases; B1, B2: 6 cases; C3: 9 cases. Two cases were combined with nerve injury. Two cases had histories of high blood pressure and diabetes. (2) Operative method: The patient was placed in the supine position, and the elbow to be operated on was positioned at 90°of abduction and supported on a lucent operating table. A pneumatic tourniquet was placed as proximally as possible on the arm. With the elbow flexed at about 60°, the first incision was made about 7 cm proximal to the tip of the medial epicondyle. In the initial cases, the ulnar nerve was isolated, released from the ulnar nerve groove, and protected carefully. In later cases, the nerve was only exposed. The medial and anteromedial side of the distal humerus was exposed through the opening between the brachial muscle and the medial intermuscular septum. The common origin of the flexor muscles was partially dissected and reflected distally, leaving a 5-mm strut to be re-sutured in situ at completion of surgery. The anterior capsule was incised. The articular surface of the trochlea was then exposed. A second incision was begun approximately 8 cm proximal to the lateral epicondyle. The space between the triceps posteriorly, the origins of the extensor carpi radialis longus and the brachioradialis anteriorly, and the anterior side of the distal articular surface were exposed. The space between the anconeus and the extensor carpi ulnaris was opened, and the most distal articular surface of the capitulum and the lateral part of the trochlea was exposed. The elbow was then flexed about 80°, and the biceps and brachial–bronchial muscles were retracted anteriorly. Any hematoma among the fragments was debrided, and the number and displacement of articular fragments were identified. The main medial articular fragment, usually associated with the metaphyseal fragment, was first reduced to the medial column and temporally fixed with K-wires. Definitive fixation with a reconstruction plate (usually 6 holes) could be completed if the metaphyseal fragment was anatomically reduced. Displaced small articular fragments were reduced to the main lateral fragment and fixed with 0.8 K-wires. The main lateral articular fragment was then reduced medially to the medial articular fragment and proximally to the lateral column and maintained temporarily with K-wires. The reduction in the articular surfaces was then checked under direct vision and using a C-arm. Any step or gap between the lateral and the medial articular fragment was abolished by abduction or adduction of the elbow and compression with forceps while keeping the medial fragment in situ. Simultaneous adjustment of the lateral column was also performed. If the articular fracture was anatomically reduced, a 1.25-mm guide wire was then inserted into the trochlea from the lateral condyle, passed through the fracture and then to the medial condyle, parallel with the distal articular surface and located in the bone between the olecranal fossa and the articular surface as confirmed by C-arm. A 4.0-mm cannulated screw was then inserted along the guide wire. As described above, the medial column could be definitely fixed with a plate if anatomical reduction was achieved. In most cases, the plate was positioned on the anteromedial side of the distal humerus. The distal end of the plate should not extend beyond the medial epicondyle and should be fixed to the bone with 2-3 screws according to the location of the fracture line. The best option was to insert the most distal two screws into the medial trochlea. If the fracture line was too low to be fixed with a plate, a tension band wire or screw fixation was used. The reconstruction plate for fixing the lateral column was carefully contoured, allowing the proximal end to be placed on the anterolateral side, and the distal end with the two most distal holes placed on the lateral side of the distal humerus. At least two screws were used to fix the plate to the lateral articular fragment, with one long screw implanted from lateral to medial side and parallel to the articular surface. Inserting the most proximal screws of the lateral and medial plates at the same level should be avoided. The reduction in the articular surface and the length of the screws were checked by C-arm. No excessive movement of the fracture fragments under the motion of the elbow was confirmed under direct vision. The dissected common origin of the flexor muscles was repaired. (3) Tips and tricks during operation: the nerves and blood vessels should be carefully protected during operation. Yi Jiangying et al reported that the anteversion of distal humerus and carrying angle of upper limb should be well reconstructed. For type C3, the first thing is to reduce intercondylar fragments, changing intercondylar fracture to supracondylar fracture, then restore the lateral column of distal humerus, in the end is to focus on the recovery of the trochlea articular surface. (4) Postoperative treatment: Antibiotics were routinely used in 3-5 d. The drainage tube was removed in 48-72 h. The stitches would be removed after two weeks. Plaster cast or hinged splint was properly applied to protect the elbow according to the classification of the fracture and the actual situation of patients. Early functional exercise was conducted. After a week or so, patients would be encouraged to do elbow flexion and extension. Rehabilitation protocol should be individualized according to fracture type and patients status. The intensity of rehabilitation also should be increased gradually. Proper upper limb weight bearing was allowed after 6-8 weeks.Results All 24 patients were successfully operated. Operation time varied from 55 to 270 minutes, and averaged in 143 min. Blood loss was ranged from 50 to 400 ml, and averaged in 183 ml. All 24 patients achieved Stage I healing. No swelling, effusion, or infection was observed. Postoperative follow-up was 3-6 months (averaged 4.5 months). Callus formation was observed in fracture end. No internal fixation loosening, myositis ossifications, malunion, delayed union or nonunion was observed in follow-ups. The outcome was evaluated according to Mayo Clinic Elbow Score. Good and excellent rate was 62.5% at 3 months postoperatively.Conclusions Classifying the distal humeral fractures using the imaging data is important for choosing appropriate surgery method. The satisfied reduction, rigid internal fixation and early exercise are critical for the functional recovery of the elbow.