The established treatment of pediatric and adolescent clavicle fractures is nonoperative, allowing the fracture to form callous and heal in situ, regardless of whether there is significant displacement (Fig. 2) (Bae et al. 2013). Well-established indications for nonoperative management include non-displaced or minimally displaced fractures, defined as displacement less than 1.5–2 cm. Conversely, open fractures and fractures with associated neurovascular injury or limb at risk should all proceed directly to operative intervention (Table 1).

Nonoperative treatment of clavicle fractures is achieved by immobilizing the child’s shoulder girdle, classically with a sling. A figure of eight dressing or shoulder immobilizer can alternatively be used. Immobilization in neonates who have sustained a clavicle fracture during the birthing process can be achieved by performing a swathe technique, such as placing cast padding followed by an ACE bandage around the torso and arm.

Follow-up radiographs are obtained at 4-week intervals until fracture union is achieved. The child is allowed to resume normal activities as tolerated once union is accomplished. Calder et al. have proposed that follow-up radiographs are unnecessary in the pediatric population due to the near universal fracture-healing rate of children (Calder et al. 2002). However, in children wishing to return to sports, radiographs should be obtained until union is established in order to decrease the refracture risk.

Even though pediatric clavicle fractures have a relatively high incidence and the majority of pediatric clavicle fractures are treated nonoperatively, there is little data regarding the outcomes of these injuries. Union rates from 95 % to 100 % have been reported with nonoperative treatment (Grassi et al. 2001; Khan et al. 2009; Vander Have et al. 2010). Most non-displaced fractures unite by 4–8 weeks post-injury, while displaced fractures achieve union in approximately 10 weeks (Vander Have et al. 2010).

The majority of patients obtain excellent outcomes and are capable of returning to their activities without restrictions. A small percentage of patients with significant fracture displacement who are treated nonoperatively may have subjective complaints of pain with prolonged activity, easy fatigability, axillary pain, or drooping shoulders with bony prominence (Vander Have et al. 2010). Bae et al. evaluated 16 patients with displaced (>2 cm) mid-diaphyseal clavicle fractures treated nonoperatively. All fractures achieved union with no meaningful loss of shoulder motion or abduction-adduction strength by isokinetic testing. The majority of patients had low DASH and pain Visual Analog Scores (VAS) that were very low, with means of 4.9 and 1.6, respectively. Only one patient required a corrective osteotomy (Bae et al. 2013). The authors concluded that routine surgical fixation for displaced, nonsegmental clavicle fractures may not be justified based on concerns regarding shoulder motion and strength alone. Further investigations are needed to determine the risk factors and causes of pain and functional compromise in the minority of pediatric patients with symptomatic malunions. In contrast, adult studies have shown that patients with significantly displaced midshaft fractures that are treated nonoperatively have significantly worse Constant Shoulder Scores and DASH scores, higher rates of nonunion, longer times to union, and more symptomatic malunions than those who have undergone plate fixation (Canadian Orthopedic Trauma Society 2007).

Absolute indications for operative management of pediatric and adolescent clavicle fractures include open fractures, comminuted fractures in which the central fragment is markedly displaced (Fig. 3), and fractures associated with neurovascular injury. Floating shoulder injuries and fractures associated with polytrauma are considered relative indications for surgical treatment by some authors. A floating shoulder injury created by a midshaft clavicle fracture and glenoid neck fracture may be managed by open reduction and internal fixation (ORIF) of the clavicle exclusively, as ligamentotaxis via the coracoclavicular ligament will reduce the accompanying glenoid fracture (Bahk et al. 2009).

Significantly displaced fractures in adults that are managed nonoperatively have demonstrated healing with a malunion that can trigger changes in shoulder mechanics, including pain with overhead activities, decreased strength, and decreased endurance (Hill et al. 1997; McKee et al. 2003). The benefit of operative fixation versus nonoperative treatment of displaced midshaft clavicle fractures has been the subject of several studies. One recent meta-analysis of randomized clinical trials comparing operative and nonoperative treatment in adults revealed a significantly higher nonunion and symptomatic malunion rate in the nonoperative group. Furthermore, patients managed with operative interventions demonstrated earlier functional return, but it is unclear whether this data is applicable to adolescents (McKee et al. 2012). It is clear, however, that young children, especially younger than age 8, have the potential to remodel foreshortened, displaced fractures.

Similar to any procedure that utilizes medical implants, it is essential to have the necessary hardware available prior to proceeding to the operating room (Table 2). Several options exist for treating pediatric and adolescent clavicle fractures, including anatomically designed clavicle plates, standard non-locking and locking plates, and intramedullary devices such as pins, wires, screws, and elastic nails.

Plate fixation requires one to determine if the location of the plate will be anteroinferior or superior. The benefit of using anteroinferior plates is the ability to drill in a posterosuperior direction so that the drill is not directed toward the surrounding neurovascular structures. Moreover, the plate is less prominent in this location. Superior placement of the plate is technically easier and provides strong resistance of the biomechanical forces acting to displace the fracture.

Compared to plate fixation, intramedullary fixation provides the potential benefits of requiring less soft tissue stripping at the fracture site, smaller skin incisions leading to better cosmesis, decreased risk of hardware irritation, easier hardware removal, and less bony weakness after plate removal (Fig. 4). However, intramedullary fixation provides inferior resistance to torsional forces when compared to plating, which can result in fracture of the intramedullary implant. Additionally, the intramedullary device has the potential to migrate, which raises major concerns and has thus limited usage of these devices.

Positioning options during ORIF or intramedullary fixation of clavicle fractures include the beach chair position with or without a Mayfield head positioner or having the patient supine. With either position, a bump is placed behind the scapula to assist in reducing the fracture.

ORIF is performed via a direct approach to the clavicle by creating a skin incision that follows Langer’s lines (Table 3). Skin incisions placed in locations not directly overlying the planned plate location on the clavicle can be done in order to avoid wound complications and to improve cosmesis (Coupe et al. 2005). Following the incision, electrocautery is utilized to divide the platysma, fascia, and periosteum in line with the initial skin incision. Throughout this process, it is imperative to identify and protect the cutaneous supraclavicular nerves which may number as many as 3–4. Subperiosteal dissection is then performed to expose the fracture site while ensuring that soft tissue attachments to any malrotated or segmental fracture fragments are preserved.

Intramedullary fixation is accomplished via an approach that utilizes a small incision over the fracture site to expose only the ends of fracture fragments (Table 4). In order to place the intramedullary device in an antegrade manner, an additional percutaneous incision is positioned over the superolateral region of the clavicle.

Once the fracture site and fragments are exposed, the fracture is reduced using bone-holding forceps. A separate interfragmentary screw or mini-fragment plate fixation can be used in segmental fractures to reduce the fracture from three parts to two. The fracture is then anatomically reduced and clamped, ensuring that areas of comminution are accounted for. Either an anatomic clavicular plate or a small pelvic reconstruction plate is then contoured to permit rigid internal fixation. The plate is subsequently applied in the preferred position, and the reduction, screw placement, and length are assessed with a combination of direct visualization and fluoroscopic imaging in multiple planes. Subsequently, the supraclavicular nerves are protected and the periosteum is closed. Layered closure with absorbable suture, including meticulous skin closure, is then performed to decrease the risk of wound complications and allow for the best possible cosmesis. Lastly, the patient is placed in a sling or shoulder immobilizer.

Intramedullary fixation is performed by exposing the fracture ends and then drilling the distal fragment in a retrograde direction through the intramedullary canal to exit the posterior-lateral cortex. The medial segment is then drilled while ensuring that there is no violation of the anterior medial cortex. The device is then placed retrograde through the canal to exit the posterior-lateral hole followed by the skin. Subsequently, the fracture is reduced and the intramedullary device is advanced antegrade across the fracture site. Several devices have mechanisms that may now be employed in order to prevent migration of the device and to permit fracture compression.

The majority of studies regarding the management of pediatric and adolescent midshaft clavicle fractures are retrospective and involve preadolescents and adolescents. Mehlman et al. retrospectively reviewed 24 children, with a mean age of 12 years, who underwent operative treatment of completely displaced clavicle shaft fractures. This series reported zero nonunions or infections, and 21 of the 24 patients were able to return to unrestricted sports activity. Three complications were reported, including two patients with scar sensitivity and one patient who developed a transient ulnar nerve neurapraxia. Hardware removal was performed on an elective basis for all patients (Mehlman et al. 2009).

Namdari et al. also performed a retrospective review of 14 skeletally immature patients who underwent ORIF for displaced midshaft clavicle fractures. There were no nonunions reported, but eight patients had numbness about the surgical site. Four patients required hardware removal (Namdari et al. 2011).

The only comparative study to date evaluating nonoperative versus operative management of midshaft clavicle fractures in adolescents was performed by Vander Have et al. In this retrospective review of 43 fractures, 25 were treated nonoperatively and 17 were treated operatively. Neither group reported a nonunion although 5 symptomatic malunions occurred in the nonoperative group, 4 of which were managed with corrective osteotomy. All complications within the operative group were associated with hardware prominence. Return to full activities occurred faster in the operative group, by approximately 4 weeks, when compared to the nonoperative group (Vander Have et al. 2010).

Although the Vander Have study reported a high rate (20 %) of symptomatic malunion in the nonoperative group, with many patients requiring corrective osteotomy, Bae et al. have recently reported that the majority of significantly displaced (>2 cm) diaphyseal clavicle fractures treated nonoperatively result in an asymptomatic malunion that does not require corrective osteotomy. Of the 16 fractures included, all progressed to malunion with only 1 patient requiring a corrective osteotomy. The mean DASH score was low at 4.9 and the mean pain VAS was 1.6. There was no significant loss of strength or motion reported (Bae et al. 2013; Vander Have et al. 2010).