Accounting for 6% of all fractures, clavicle fractures are common in clinic with the middle 1/3 fracture as the most common type. The options for treatment include conservative and surgical strategies. The effect of conservative treatment via manipulative reduction is often unsatisfactory (especially for comminuted fractures) . For the middle 1/3 clavicle fractures, the displacement of proximal end caused by the upward and posterior tractions of sternocleidomastoid muscle may result in instability and nonunion of fracture. These risk factors increase the chance of redisplacement, which restrict early activity and affect joint function simultaneously. With in-depth study of the injury and therapeutic effect of clavicle fractures and patients' improved expectation for life quality, more patients require surgical treatment. Those with comminuted and shortened fractures are often treated with plate fixation. The surgical treatment reduces complications including nonunion, shoulder deformity, pain, functional impairment and neurovascular injury, et al., and greatly improves the clinical therapeutic effect of clavicular fracture. However, the excessive dissection of surrounding tissue required by conventional plate fixation increases the chance of nonunion. Moreover, the large incision on skin not only leads to the formation of large scar but also causes supraclavicular nerve damage and numbness in the corresponding area. The minimally invasive percutaneous plate osteosynthesis (MIPPO) technique can nicely solve some of the common drawbacks of traditional surgery.

Ⅰ. General information: From January 2012 to December 2014, 86 patients (52 males and 34 females) with mid-shaft clavicle fractures were treated in the first affiliated hospital of dalian medical university. The age ranged from 14 to 70 years old with an average of 40 years. All cases were closed fractures, which included 48 cases of traffic accident and 38 cases of fall injury. 36 cases were affected by left side, and 50 cases were affect by right side. Among the patients, there were 14 cases of combined rib fractures and 3 cases of scapular fractures (minor fractures) . The time from injury to admission ranged from 1 to 10 days with an average of 5 days. The operation indications: fractures of total displacement, comminuted fractures and clavicle shortening of over 2 cm. Locking compression plate was used in all patients who were randomly divided into Conventional ORIF group (Group A, 41 cases) and MIPPO group (group B, 45 cases) based on the different placements of internal fixators. Ⅱ.Inclusive and exclusive criteria: Inclusive criteria: (1) Unilateral displaced mid-shaft clavicle fractures of adult (Robison type II) ; (2) Combined rib or scapular fractures without the affection of shoulder function assessment. Exclusive criteria: (1) Open fractures; (2) Combination of other severe fractures that affect shoulder function assessment; (3) Combined brachial plexus injury or other disease that affects upper extremity function; (4) Pathological fractures; (5) Patients who are lost for follow-up. Ⅲ. Surgical methods: (1) Conventional ORIF group (group A) : After general anesthesia, the patient was placed in semi-sitting position (beach chair position) . The fracture ends and acromion were both marked. The incision was made along the surface of clavicle and was extended toward acromion. The soft tissue was separated and exposed to reveal periosteum and fracture fragments with particular attention to comminuted or displaced disc-shaped fractures. Comminuted fractures were treated with lag screw for fixation. The fracture fragments were converted into 3-part or 2-part fractures. Finally, anatomical locking compression plate was used for fixation with 3 screws on each side. (2) MIPPO technique group (group B) : After general anesthesia, the patient was placed in semi-sitting position (beach chair position) . The situation of fracture was confirmed, and the fracture ends and acromion were both marked to design minimally invasive incisions on the distal and proximal clavicle. 2-3 cm arc incisions were made from the center of fracture toward both sides along the longitudinal axis of clavicle. The skin, subcutaneous tissue and deep fascia were cut open, and the subperiosteum was slightly dissected for fracture reduction. The periosteum at fracture ends was preserved to the greatest extent, and the hematoma was removed to expose the fracture ends. Crossing clavicle downward from superomedial to inferolateral, the supraclavicular nerve should be identified and protected during operation. The fractures were reduced by manipulation or percutaneous poking with Kirschner wire under fluoroscopy. With satisfactory linear and positional alignments, the small fragments were fixed with lag screw. If necessary, Kirschner wire should be adopted for temporarily fixation and maintenance of the position of clavicle. A subcutaneous tunnel was established along the incision edge, through which the appropriate anatomical locking compression plate was selected and inserted towards the distal end of clavicle. With the position of distal incision determined, A 1-2 cm of distal incision was made. Then, the plate was removed and inserted again through the distal incision to the proximal end. With the position of proximal incision determined, a 1-2 cm proximal incision was made as well. As the plate was fully inserted, both the proximal and distal ends were checked to make sure their attachment with clavicle. The plate position was properly adjusted as needed, and 3 screws were placed through each of the two small incisions. The reduction was further observed under fluoroscopy.Ⅳ. Postoperative treatment: Postoperative antibiotic was given to prevent infection within 24 hours. After adequate drainage, the stiches were taken out 7-10 days later. The affected arm was suspended for 2 weeks with triangular bandage, and passive training was initiated on the 2nd postoperative day. Passive should joint exercise was carried out within 2 weeks, and no-weight bearing active exercise started 2 weeks later. The callus was observed during regular reexamination, and normal activities were gradually restored based on the growth.Ⅴ. Outcome Measurements: (1) Surgical situations and postoperative follow-up scores: the operative time and hospital stay of both groups were recorded, and the wound healing was observed as well. After discharge, the patients were required to be followed up every two weeks till fracture healing. The condition of affected limb was observed under fluoroscopy, and the patients were followed up every 3 months after fracture healing. The criteria of fracture healing include continuous callus formation and the disappearance of fracture line revealed by radiographic examination, no tenderness in the clinical examination of fracture ends, and pain free in active movement and weight-bearing activities. The shoulder joint functions including pain, anatomical location and activity were assessed by Neer score. The criteria of Neer score: 90-100 points as excellent; 80-89 points as good; 70-79 points as moderate; less than 70 points as poor. (2) Complication rate and patients’ subjective satisfaction: the common postoperative complications were compared using statistics, including infection, delayed union, hypertrophic scar, supraclavicular nerve damage, et al., and in the meanwhile, the subjective satisfaction (satisfaction or dissatisfaction) with the operation of patients was asked as well. Ⅵ. Statistical analysis: The SPSS 23.0 software was used for data processing. According to the data distribution characteristics, the enumeration data were analyzed using the t test and the measurement data were analyzed using the χ² test. A P value <0.05 was regarded as statistically significant.

(1) Comparison of surgical situations and postoperative follow-up scores: The surgical situations were compared between the two groups, including operation time (min) , hospital stay (d) , fracture healing time (w) and Neer score. The operation time (approximately 1 hour) , the hospital stays (5-7 days) and the fracture healing time (11-14 weeks) were the same for both groups. Also, the Neer scores for both groups were excellent without significant statistical difference (P>0.05) . The patients' shoulder joint functions including abduction, internal rotation and external rotation were all well performed. (2) Comparison of complication rate and patients’ subjective satisfaction: Both groups have no infection or delayed union occurred, and the comparison was not statistically significant. However, there were differences in the comparison of hypertrophic scar, supraclavicular nerve damage and subjective satisfaction. There were 12 cases of hypertrophic scar in group A and 0 case in group B, and the comparison had statistical significance (χ² =15.307, P<0.001) . There were 20 cases of numbness in the dominated region of supraclavicular nerve in group A and only 4 cases in group B, and the comparison had statistical significance (χ² 16.969, P<0.001) . There were 26 cases of subjective incision satisfaction in group A and 39 cases in group B, and the comparison had statistical significance (χ² =6.285, P<0.012) .

MIPPO technique has been used for the treatment of mid-shaft clavicle fractures. The surgical incision is small, and the internal fixation is reliable. These not only embody the advantage of plate fixation but also minimize the complications of hypertrophic scar and supraclavicular nerve injury. Thus, it provides good conditions for early functional exercise and is worth of clinical application.