The humeral shaft fracture is a common type of fractures in adults, accounting for about 2%-3% of all fractures in the body. The humeral shaft fractures with severe displacement and combined injury need surgical treatments. In recent decades, with the development of internal fixation technology and fixation material design, more choices have been made for surgical methods, and controversy has also appeared. At present, the two most commonly used surgical methods are closed reduction intramedullary nail fixation and open reduction plate internal fixation. Both methods have their own advantages and disadvantages. The advantage of antegrade intramedullary internal fixation is minimal invasive, thus the blood supply around the periosteum and the fracture is well protected, and the fracture end is less stimulated; the disadvantage is that the rotator cuff would be cut when the nail is inserted, and patients with limited shoulder activity and the shoulder pain were reported. The advantages of the plate fixation are that the anatomical reduction can be obtained under direct vision, and the fracture can be firmly and stably fixed; the disadvantages are that the trauma is large with more blood supply damage , and there is a risk of radial nerve injury. At present, the advantages and disadvantages of the two surgical methods are always debated. The previous RCT studies have relatively low level of evidence, and several meta-analytical studies have different conclusions, thus there is no clear conclusion now.

1. General information.Inclusion criteria: Patients with humeral shaft fractures who underwent internal fixation surgery in the Department of Trauma and Orthopaedics of Peking University People's Hospital from August 2013 to October 2017. Exclusion criteria: Open fracture, pathological fracture, old fracture, radial nerve injury, severe multiple injury, abnormal function of shoulder and elbow joint of the affected limb before the injury.A total of 56 patients met the standard, including 35 males and 21 females, aged 18-70 years, with an average age of 42 years. There were 11 cases of traffic accidents, 36 cases of falls, and 9 cases of falls from height. The fractures were classified by the AO clasification: 19 cases were type A, 24 cases were type B, and 13 cases were type C. Locking compression plate fixations were performed in 31 cases (AO company or Biomed company) , and antegrade magnetic navigated interlocking intramedullary nails were in 25 cases (Smith and Nephew Company) . 2. Surgical methods:All patients underwent general anesthesia with tracheal intubation combined with brachial plexus block analgesia. (1) antegrade magnetic navigated interlocking intramedullary nail: The patient took the beach chair position, a vertical straight incision was made on the lateral side of the shoulder for about 4 to 5 cm, the guide needle was inserted at 1 cm medial to the greater tuberosity of the humerus and 0.5 cm behind the biceps groove, parallel to the axial direction of the medullary cavity. Reduction was achieved under X-ray, then the intramedullary nail was inserted, the locking screws were guided magnetically. (2) Locking compression plate: The anterior lateral incision was made in the proximal and middle fractures, and the posterior incision was taken in the distal fracture to find and protect the radial nerve; the fracture was cleared and aligned, and then the Kirschner wires were temporarily fixed. The prebent AO or Biomet locking compression plate (LCP) was placed, and then the cortical bone screws and the locking screws were inserted to fix the fracture depend on the situation.3. Postoperative treatment:The postoperative forearm sling suspension protection lasted 3 weeks. From the second day after surgery, the patients could perform passive shoulder and elbow joint function exercises within the allowable range of pain. Active shoulder and elbow joint exercises were performed 3 weeks after surgery. The affected limb should avoid carrying heavy objects within 2 months. The X-ray was reviewed monthly for 4 months after surgery to understand the fracture healing. X-ray examination was performed once every 3 months after the fourth month, and once every 6 months after 1 year. The operation time, postoperative complications, fracture healing time, and range of motion of the shoulder and elbow joints were recorded. The shoulder and elbow function scores were evaluated 1 year after surgery. The shoulder joints were scored by the University of California, Los Angeles (UCLA) based on pain (10 points) , activity (10 points) , the range of forward flexion (5 points) , forward flexion strength (5 points) and patient satisfaction (5 points) . The score was divided into excellent (34-35 points) , good (29-33 points) , and poor (<29 points) . The elbow joint function used the Mayo score, which was based on pain (45 points) , range of joint activity (20 points) , degree of stability (10 points) , and daily function (25 points) . It was divided into excellent (90 points or more) , Good (89-75 points) , fair (60-74 points) , poor (<60 points) .

All the 56 patients were followed up for 10 to 26 months, with an average of (18.7±3.6) months. The operation time was about 45-110 minutes and the average operation time was (74±14) minutes. Two cases of superficial wound infection occurred, and finally recovered well after debridement and dressing change. One case of intraoperative iatrogenic radial nerve injury was treated with nerve electrical stimulation combined with neurotrophic drugs. The wrist and thumb extension functions were completely restored 6 weeks after operation. The fractures reached clinical union at 8-24 weeks, with an average of (13±4) weeks. There was no delayed union or nonunion. The union rate reached 100%. The shoulder abduction was 95°-170°, with an average of (151±11) °; and forward flexion was 100°-175°, with an average of (153±12) °. The elbow joint extension was 0°-10°, with an average of (5±2) °; the flexion was 122°-145°, with an average of (131±5) °. In the follow-up of all patients, the shoulder function was evaluated using the UCLA score: 53 cases were excellent and 3 cases were good. According to the Mayo score, the elbow joint functions were excellent in 55 cases and was good in 1 case.Comparing the intramedullary nail group with the LCP group, there was no significant difference in gender ratio, age, operation time, length of hospital stays, and fracture types between the two groups, so the two groups were comparable. There was no infection or nerve injury in the intramedullary nail group. There were 2 cases of superficial wound infection and 1 case of nerve injury in the LCP group, but there was no significant difference between the groups. The acromion impingement ratio of intramedullary nail group was significantly higher than that of LCP group. The range of shoulder flexion and abduction activity was significantly lower than that of LCP group. There was no significant difference in the other prognosis indicators.

Locking compression plate fixation group had a higher incidence of wound infection and a higher incidence of radial nerve injury, but there was no significant difference compared with the intramedullary nail fixation group. The incidence of postoperative acromion impact ratio was higher in patients with intramedullary nail, and the mobility of the shoulder joint was also significantly decreased. Therefore, this study found that the use of intramedullary nails and LCP had their own advantages and disadvantages, the overall safety and effectiveness of both were high. All patients could obtain excellent shoulder and elbow joint function. Therefore, the surgeon should take the strength and avoid weakness when selecting the fixation method, for example the surgeon should minimize the damage to the rotator cuff when using the intramedullary nail and improve the surgical technique to properly handle the intramedullary needle cap to avoid acromion impact. Large-scale prospective randomized controlled trials are still needed to validate the result.