18.3.1 Malunions

Malunion may result from a superiorly displaced or externally rotated greater tuberosity, medialization of the lesser tuberosity, varus or valgus neck-shaft angle or a combination of these factors. The described mechanisms and applied loads can result in bone, joint or combined bone and joint malunion, and they can consequently alter the articular function.

The misalignment and/or remodelling of the humeral head results in joint incongruity, which does not provide optimal mechanical conditions; this can lead to stiffness due to the retraction of the capsule and ligaments or impingement of the surrounding structures during shoulder motion and may alter rotator cuff tension. Three situations may be at the origin of the malunion after ORIF as described by Duparc et al. [24]: (1) problems with the initial reduction, (2) problems with the fixation leading to secondary displacement (3) and problems with the protection/stability leading to secondary displacement.

The main symptoms of malunion are pain and limited joint range of motion. The pain must be characterized as precisely as possible. Any isolated signs of subacromial impingement or long head of biceps involvement can direct the therapeutic decision. Examiner should take in mind that malunion at different sites can lead to different pathological scenarios. A patient with malunion of the greater tuberosity may exhibit perceived weakness caused by a shortened functional offset of the posterosuperior rotator cuff. Anterior instability could be also referred by a patient presenting with greater tuberosity malunion because of the posterior abutment. Malunion of the lesser tuberosity may result in weakness in internal rotation. Alterations in the intertubercular groove, involving both tuberosities, are often associated with exceptionally painful, chronic biceps tendinopathies. Electromyographic studies should be performed if a nerve dysfunction is suspected. Radiographic evaluation includes three views of the shoulder: a true AP view in the scapular plane, an axillary view and a scapular lateral view. Most malunions will be detectable in these images. Although displacement of the articular surface or tuberosity is evident on plain radiography, the relationship between the fragments may not be completely well delineated. Compared with plain radiography, CT scan provides more details: bony malunion, articular incongruity and the degree of tuberosity displacement are more clearly visualized [25]. MRI is recommended for detecting osteonecrosis and evaluating soft tissue structures, including the rotator cuff, long head of the biceps tendon and labrum. Proximal humeral malunions have been traditionally classified according to Beredjiklian [26]. This classification evaluates radiographical aspects, focusing on different anatomical sites disruption. Type I malunion includes misalignment of the greater or lesser tuberosity greater than 1 cm from the anatomic position. Type II is distinguished by articular surface incongruity, and type III involves articular surface malalignment with malunion of the tuberosities (>1 cm) and humeral head relative to the shaft.

Boileau et al. [27] proposed another classification system that included the sequelae of displaced PHFs and the implications for surgical management with reverse shoulder arthroplasty (RSA), dividing them in two categories. Category 1 refers to intracapsular injuries and the sequelae of impacted fractures. This category comprises type I, characterized by humeral head necrosis or impaction, and type II with chronic dislocations or fracture-dislocations. In this category, implantation of RSA for the treatment of malunion will not require greater tuberosity osteotomy and provide good and predictable results. Category 2 comprehends the sequelae of extracapsular fractures, dividing them in malunion of the surgical neck (type III) and in severe malunion of the tuberosity (type IV). In this category, tuberosity osteotomy during RSA implantation will be mandatory; however in types III and IV, surgical procedure is characterized by poor and unpredictable results.

Based on these classification systems, the various cases of malunion can be grouped in the following manner: (1) bone malunion (tuberosity, inter-tuberosity, subtuberosity), (2) joint malunion (with or without associated humeral head osteonecrosis) and (3) combined bone and joint malunion (with or without associated humeral head osteonecrosis). Orthopaedic surgeons should accurately evaluate all the features of the patient affected by proximal humeral malunion after ORIF to adequately manage this complication. Not only the fracture characteristics but also the patient’s general condition, functional request and postsurgical compliance should be evaluated. In patients with low activity levels, tolerable pain, significant comorbidities that preclude surgical intervention and in those who are unable to comply with rehabilitation and/or who are willing to accept some loss of shoulder function, nonsurgical management of proximal humerus malunion has been found to provide acceptable results [28–30]. Physical therapy (with NSAIDs and occasional cortisone intra-articular injections) may be useful in low demanding patients to progressively strengthen shoulder musculature and maximize ROM. For patients with persistent dysfunction or pain secondary to malunion after ORIF, surgical intervention may improve quality of life. Patients should be accurately informed that improved function rather than restoration of function is the goal of the surgical treatment. The technique used to address malunion is guided by the existing deformity. Surgical options are broadly divided into two categories: humeral head-preserving or humeral head-sacrificing techniques [26]. In the setting of a preserved articular surface and intact blood supply to the humeral head, the use of a head-preserving technique is indicated. Arthroscopic soft tissue releases and rotator cuff retensioning for tuberosity malunion [31–33], open osteotomies and removal of bony protuberances [34, 35] are the main head-preserving procedures. Head-sacrificing techniques are applied in the setting of glenohumeral joint incongruity or degeneration and the development of osteonecrosis. These techniques include hemiarthroplasty, anatomical reduction and RSA [36–39]. Adequate treatment of soft tissue dysfunction should be mandatory in all the aforementioned techniques, to appropriately address specific post-traumatic articular restrictions. Soft tissue management may include release of the subdeltoid and subacromial spaces, rotator interval and subscapularis.

18.3.2 Non-unions

The prevalence of non-union in proximal humeral fractures after ORIF is 1.7%, although it increases to 8% in those cases with metaphyseal comminution and to 10–13% if more than one-third of the surgical neck is involved [22, 40, 41]. Several factors such as soft tissue interposition, severe displacement, poor anatomical reduction and early mobilization have been reported to promote non-union. Non-union in patients who received ORIF after proximal humeral fractures represents a mechanical problem related to internal fixation, with a strong influence of biological factors and comorbidities. This complication, therefore, should be considered as a multifactorial problem [42, 43]. Mechanical outcomes of internal fixation for proximal humeral fractures are influenced by soft tissue interposition, extensive comminution, hanging arm casts, poor surgical technique or any combination thereof. Recognized predisposing biological factors, contributing to non-union development, include personal pre-existing pathological conditions and living habits. Pre-existing personal pathological conditions promoting non-union include advanced age, female sex, osteoporosis, diabetes mellitus and neurovascular dysfunctions [42, 44]. Age over 60 must be considered at high risk of non-union, particularly in females because of hormonal imbalances after menopause [23]. Uncontrolled diabetes and vascular and neurotrophic problems have been reported to influence the non-union development decreasing the formation of collagen and cells involved in formation and maturation of bone callus. Persons who smoke are at 5.5 times higher risk than non-smokers for developing non-union [29]. Living habits as diet, smoking, alcoholism, as well as drugs taken for pain control after surgery, usually non-steroidal anti-inflammatory drugs (NSAIDs), or non-adherence of the patient in postoperative management (such as temporary bracing in selected cases) can predispose the complication onset. The time to surgery has been demonstrated to not influence the risk for non-union, independent of fracture type [45]. Although not all patients with humeral non-unions are clinically symptomatic, those presenting with symptoms are typically severely disabled by pain and loss of motion.

18.3.3 Avascular Necrosis of the Humeral Head

One of the main complications after ORIF in proximal humeral fractures (PHFs) is avascular necrosis (AVN) of the humeral head. It represents a devastating event with serious sequelae. The incidence of AVN varies from 3 to 90% [46, 47] throughout the literature. Traditionally this complication has been related with displacement and the number of parts created by the fracture lines (as described by Neer). The humeral head articular surface is characterized by a tenuous blood supply [46, 48, 49]. The arcuate branch of the anterior humeral circumflex artery provides a significant proportion of the flow to the humeral head articular surface in a retrograde fashion. Recent studies have attempted to use deltoid-splitting or minimally invasive approaches with the belief that less soft tissue disruption in proximity to the humeral head would preserve its blood supply. Surgical approach has been demonstrated affecting the AVN incidence in patients affected by proximal humeral fractures. Comparing the minimally invasive approach to the deltopectoral approach, Liu et al. [50] reported lower rate of AVN in the latter group. The authors believed that the minimally invasive approach decreased soft tissue stripping and preserved the blood supply around the proximal humerus. Hertel presented a useful classification to predict development of humeral head ischaemia according to fracture pattern. Good predictors were the short metaphyseal head fragment extension (<8 mm), the integrity of the medial hinge and the basic fracture pattern [51]. Following first observation, the same author few years later described that initial ischaemia after intracapsular fracture of the humeral head did not necessarily lead to the development of avascular necrosis [52]. On the other side, avascular necrosis may occur unexpectedly in initially perfused heads as a long-term complication. Patients who develop this complication can face significant long-term disability and few reliable surgical options for treatment. Burrus reported that TSA in patients with humeral head AVN is associated with significantly increased rates of numerous postoperative complications compared to patients without a diagnosis of AVN, including infection, dislocation, revision arthroplasty, stiffness, periprosthetic fracture and medical complications [53]. Archer et al. [54] recently investigated the relationship between the timing of PHF fixation and rates of AVN. A temporal relationship between time to ORIF of displaced femoral neck fractures and development of AVN has been established. It follows that the same could be true for the proximal humerus, and therefore a shorter time to ORIF may correlate with a smaller incidence of AVN in these fractures. No correlation between time to surgery, either early (less than 72 h) or late (greater than 72 h), and incidence of AVN was identified in the population studied. Xia et al. [55] evaluated the effectiveness of virtual planning for ORIF of proximal humeral fractures. Computerized preoperative planning facilitated ORIF and showed good results for patients with complex proximal humeral fracture, representing a favourable option with reported lower rate of AVN. Hardware selection can also affect the AVN incidence. CFR-PEEK plates proved as reliable as metallic plates in the treatment of proximal humeral fractures. The advantages of these new devices include a better visualization of fracture reduction during intraoperative fluoroscopic assessment and easy hardware removal due to the absence of screw-plate cold fusion [56]. Schliemann et al. [57] reported a lower incidence of AVN in patients treated with their CFR-PEEK implant compared to conventional locking plate, with a minimum follow-up of 2 years.

18.3.4 Heterotopic Ossifications

Heterotopic ossification (HO) in the shoulder represents a rare consequence after ORIF of the proximal humeral fractures, with a reported incidence ranging from 0 to 10% [58]. It consists in formation of lamellar bone in non-osseous tissues such as muscles, nerves and connective tissue. Despite periarticular ossification after shoulder surgery has been reported since the nineteenth century, the underlying pathogenetic processes are not yet fully understood. HO formation is presumed to result from inappropriate differentiation of pluripotent mesenchymal cells into osteoblasts; however the definitive pathophysiologic causal factors remain uncertain [59]. Bone morphogenic protein (BMP2) has been shown to promote this differentiation interacting with the Wnt/b-catenin in osteoblasts [60]. Differentiation usually occurs 16 hours after surgery and peaks at around 32 h postoperatively. Heterotopic ossification is typically asymptomatic and detected only as an incidental finding on radiograms, usually 4–6 weeks after surgery. When symptomatic, it most commonly causes decreased range of motion at the affected joint, and in most severe cases complete bony ankylosis may occur. HO has been reported to be related with symptoms as local tenderness and pain, and if located superficially, there may be symptoms as localized warmth, mild oedema and erythema [61]. Recognized risk factors for developing HO include male gender, osteoarthritis, duration and complexity of the surgery and previous personal history of HO at a particular anatomical site [62]. The decision to provide prophylactic treatment must balance a patient’s risk of heterotopic bone formation against the potential risks of preventive treatment. The two primary prophylactic modalities are radiation therapy (RT) and non-steroidal anti-inflammatory drugs (NSAIDs). However, both treatment options have disadvantages. For NSAIDs therapy, prolonged bleeding time, gastrointestinal side effects and an increase in non-union of associated fractures have been observed. A well-known risk of NSAID-related therapy is related to patient compliance: up to 37% of the patients that used NSAIDs to prevent HO had to cease these medications because of serious side effects [63–65]. When using radiation therapy, the potential risk of cancer and infertility should be considered in addition to higher costs, including transportation of patients to the radiation department [66, 67]. NSAIDs administration and subsequent prostaglandin inhibition effect (in particular prostaglandin-E2) have been shown to significantly reduce the incidence of HO, specifically indomethacin [68, 69]. Indomethacin is commonly used for prophylaxis, given its ease of administration and low cost. It is typically given over a period of 5–6 weeks at 25 mg three times per day. Other NSAIDs like ibuprofen, tenoxicam, naproxen, flurbiprofen, ketorolac and diclofenac are proven to be effective [70, 71]. COX-2 blockers have been proposed as a reasonable alternative to NSAIDs prophylaxis to prevent HO. Seven-day treatment with etoricoxib 90 mg once daily represents a promising treatment option, associated with fewer gastrointestinal side effects than non-selective NSAIDs [72]. A further advantage of COX-2 inhibitors compared with non-selective NSAIDs is that COX-2, which do not interfere with platelet function, can reduce perioperative blood loss [73, 74]. In the past, diphosphonates were also used for prophylaxis. These agents were largely abandoned after they were found to only prevent mineralization of the ectopic bone matrix [75]. The appearance of periarticular ossifications of the shoulder after surgery seems to be related to a minor clinical impact, rarely painful and with little influence in joint function [76]. Severe cases with major functional deficits should and can be prevented by a fast and atraumatic operation technique [77]. It’s difficult to accurately delineate the real amount of HO in shoulder function impairment, due to its usual association with other complications. Malunion or non-union of the humeral head after ORIF is frequently associated with HO, being reasonably more influent to determine articular dysfunction.

18.4.1 Screw Loosening and Pull-Out

Introduction of locking plates for proximal humeral fractures has significantly reduced problems related to epiphyseal screws mobilization. Screw loosening has led in the past to multiple anatomopathological scenarios, ranging from absence of any symptoms to very severe consequences as lesions to adjacent structures such as lungs and vessels. Tingart [81] investigated the three-dimensional trabecular bone mineral density (BMD) in the humeral head and determined the effects of trabecular BMD on the pull-out strength of cancellous screws. Five regions of interest (ROI) were defined in the humeral head (superior-anterior, superior-posterior, central, inferior-anterior, and inferior-posterior). The trabecular BMD of each region was determined by use of peripheral quantitative computed tomography. Cancellous screws were inserted in each ROI and cyclically loaded. The superior-anterior ROI had a lower trabecular BMD than all other ROIs (P < 0.001). The central ROI had a higher trabecular BMD than the inferior-anterior ROI (P < 0.01), whereas no differences were found between the inferior-posterior, superior-posterior and central ROIs. Pull-out strength was lower in the superior-anterior ROI compared with all other ROIs (P < 0.01). The trabecular BMD and pull-out strength were significantly correlated (P < 0.01). The author concludes that placement of screws in regions with a higher trabecular BMD may help to prevent implant loosening and may also improve patient outcome. Proximal humerus fracture fixation can be problematic because of osteoporosis making it difficult to achieve stable implant anchorage in poor bone stock even when using locking plates. This may cause implant failure requiring revision surgery. Krappinger et al. [42] pointed out the relevance of preoperative assessment of local bone mineral density. Local bone mineral density, restoration of the medial column, non-anatomic reduction, and age were significant predictors of fixation failure (p < 0.01). Osteosynthesis of osteoporotic fractures with non-reconstructable comminution of the medial column is prone to failure. Several strategies have been proposed to improve stability in proximal humeral fractures treated with ORIF. In patients with impaired bone mineral density, cement augmentation either directly to the head prior to screw insertion or via cannulated and perforated screws has been demonstrated to be a valid option to decrease the risk of varus impaction [82]. Cement augmentation of particular screws of a locking plate in the regions of low bone quality has been demonstrated to be effective in improving stability in a proximal humerus fracture model [83]. Locked plating of proximal humeral fractures with trauma cement augmentation of humeral head screws showed similar clinical outcomes but reduced the rate of early implant-related complications compared to locked plating without additional cement augmentation [84].

18.4.2 Implant Failure

Failure rate should be influenced by the different locking plate design including overall profile, manufacturing, material and screw configurations. Although these variations are small, they are likely to have some impact upon fixation failure. Specifically, implant stiffness has a direct effect upon the bone-implant interface [85]. Under cyclic loading, rigid implants lead to early loosening and failure of the bone-implant interface presumably due to the mechanical mismatch of the bone and the implant [86]. Less rigid and smaller-dimensioned implants, although potentially ‘poorer’ in terms of the early stability that they offer, exhibit lower peak stresses at the bone-implant interface compared with more rigid and oversized osteosynthetic devices and may be better suited to the treatment of osteoporotic fractures where screw cut-out is a significant problem. There is ongoing work to produce proximal humeral locking plates that have an elastic modulus that is more similar to that of human bone while still maintaining implant strength. Lab-based research investigating carbon fibre-reinforced polyetheretherketone (PEEK) is extremely promising [57], and interesting results from high-quality multicentric clinical studies have been recently published on this topic [56, 57]. Where medial column stability is suboptimal, fracture stability can be improved using a medial supporting screw(s) that is inserted into the inferior most portion of the humeral head, by using an endosteal implant, through impaction of the shaft into the humeral head fragment to restore load transfer through the ‘new’ calcar or by inserting an intramedullary fibular strut graft. The literature demonstrates that a medial support screw(s) enhances the primary stability of locking plate fixation in the majority of fixations and therefore should be used in all cases where technically feasible to support the medial column. The accurate placement of the calcar screws within the bottom 25% of the humeral head has also been shown to decrease the risk of fixation failure [87]. A minimum of five screws should be inserted into the humeral head aiming if possible for the central, inferior-posterior and superior-posterior regions [85]. Anatomical reduction is the aim, but achieving inherent bone to bone mechanical stability is questionably more important especially if an anatomically acceptable and inherently stable fracture configuration can be obtained. Varus mal-reduction and lack of medial column support are high predictors of failure and should be avoided at all cost. Finally, in addition to obtaining optimal fracture reduction and/or fracture stability, there is good evidence that a medial column support screw should be used routinely.

18.4.3 Malpositioning

Adequate intraoperative plate positioning frequently represents a challenge for the surgeon, in particular in 3- and 4-part fractures. Subacromial primary impingement can be the result of poor intraoperative plate positioning, while secondary has been related to sequelae of humeral head collapse. Impingement rate after ORIF ranges from 4.8% [88] to 18.5% [89] in literature. ORIF represent a good surgical option for severely displaced proximal humerus fractures, but functional impairment related to the hardware malpositioning can persist. Acklin et al. [90] analysed the functional outcome after locking plate removal in proximal humerus fractures, showing statistically significant improvement of the Constant score after implant removal. Major concerns against hardware removals are very high complication rates of 20–47% [91]. Particularly in locked compression plates, the most frequently observed complications were jammed screws (11% risk) and damaged recess in which the screwdriver turned freely [92]. A longer time in situ contributed to the complication rate. A proper surgical technique is mandatory for avoiding this specific complication: increased attention to plate placement and preventing varus collapse are the methods surgeons are using to decrease this complication [93].

18.4.4 Articular Perforation

In a systematic review of the literature in 2011, Sproul et al. [88] looked at 12 studies with a total of 514 patients and found that screw perforation occurred in 8% of patients and was the most common cause for re-operation. More recently in 2013, in a cohort of 121 patients, Jost et al. [18] found secondary screw cut-out in 57% of patients.

Screw penetration can be primary, due to the screws being placed too close to the articular surface or indeed perforating the articular surface intraoperatively, leading to patient morbidity from screw impingement upon the glenoid, chondrolysis and the need for further surgery especially if the prominent screws involve the major articular component of the humeral head.

Standard intraoperative images may miss nearly half of screw penetrations, and it is recommended that a combination of four projections (axial view with 30 degrees abduction and anterior-posterior views in internal rotation, neutral and external rotation) have 100% sensitivity for identifying screw perforation [94]. Secondary penetration occurs due to loss of fracture reduction and head fragment subsidence. Brunner et al. reported 35 screw penetrations (22 primary and 13 secondary) in a cohort of 158 patients [95]. In both types of penetration, surgical technique is invariably the main culprit. With proximal humerus locking plates, the locking of the threaded screw heads within the plate provides increased axial and angular stability. However, if there is head collapse post fixation, the screws are unable to back out, and therefore the screws penetrate through the head. Clinical studies for locking plates report a significant number of complications due to the perforation of screws through the humeral head. One potential solution is to use polyaxial screws in them. This has been named the second-generation locking technology as it allows the screw direction to be adjusted before locking, as opposed to the conventional locking systems where screw angles are predefined and therefore monoaxial. Egol et al. [96] performed a retrospective study on patients with acute traumatic fracture of the proximal humerus with metaphyseal defect treated with open reduction-internal fixation with a locked plate. Metaphyseal defects were treated with 1 of 3 strategies: no augmentation, augmentation with cancellous chips or augmentation with calcium phosphate cement. Findings of joint penetration were significant among patients treated with plates and screws alone versus those augmented with calcium phosphate (P = 0.02) and for those augmented with cancellous chips versus those augmented with calcium phosphate (P = 0.009). Augmentation with calcium phosphate cement in the treatment of proximal humeral fractures with locked plates decreased fracture settling and significantly decreased intra-articular screw penetration.

18.5.1 Neurovascular Injuries

Surgical management of proximal humeral fractures is graved by a recognized risk for axillary nerve injury [97]. Neurologic lesions during ORIF for proximal humeral fractures are often related to surgical approach. The deltoid-splitting approach is characterized by a higher risk of axillary nerve lesion during preparation, because the anterior branch of the axillary nerve perpendicularly crosses the incision. Wu et al. [98] compared electrophysiological results between two groups operated using deltopectoral approach or deltoid-splitting approach at 3 months of follow-up. They found that 12.5% in the deltopectoral group and 25% in the deltoid-splitting group showed signs of reinnervation or denervation of the deltoid muscle. Another cohort study by Gavaskar et al. [99] prospectively analysed 50 patients with proximal humeral fractures who underwent open reduction and internal plate fixation using the extended deltoid-splitting approach. Electrophysiological findings showed three temporary and one permanent axillary nerve lesion. The permanent lesion did not correlate with clinical findings because the patient did not show any sensory or motor deficiencies. Some authors proposed intraoperative trick to reduce nervous lesions: they indicated the nerve with the index finger in the subdeltoid bursa, and its course was marked on the skin. Additionally, they used a five-hole plate that was inserted with its tip contacting the bone, and screws were fixed in the three distal holes, far away from the axillary nerve [100–102]. Vascular sequelae can also affect outcomes after ORIF in proximal humeral fractures. The anterolateral acromial approach, which uses the anterior deltoid raphe and axillary nerve protection, has recently been advocated as a minimally invasive technique. Splitting the anterior deltoid raphe from the acromion distally allowed direct access to the lateral plating zone of the proximal humerus. The bare spot in this region may be a safe area for plate application. These findings may be of particular importance if the vascular supply to the humeral head has already been partially compromised by preceding trauma. This direct approach to the lateral bare spot on the proximal humerus may minimize iatrogenic vascular injury when treating these fractures [103]. The reported incidence of venous thromboembolism (VTE) after ORIF for proximal humeral fracture was 0.82% [104]. Diabetes mellitus, rheumatoid arthritis and previous ischaemic heart disease were identified as major risk factors for VTE in patients [105, 106].

18.5.2 Infections

The reported rate of infections after ORIF for proximal humerus fractures ranges from 0 to 8% [107]. Recently published studies, comparing the intramedullary nailing for PHF with ORIF, reported a higher rate of infections after the second technique [108]. Common causative organisms are coagulase negative Staphylococcus species and P. acnes. P. acnes has been implicated as a common cause of deep shoulder infections [109]. The length of surgery also increased the risk of infection (OR, 1.009; P ¼ 0.05). Blonna et al. [110] highlighted that preoperative skin preparation with chlorhexidine gluconate, length of surgery and type of prophylactic antibiotic play an important role in the rate of acute deep infections after surgical treatment for proximal humeral fractures. They recommend, if not contraindicated for referred patient’s conditions, preparing the shoulder with chlorhexidine gluconate and avoiding the use of first-generation cephalosporin in favour of a more effective prophylactic therapy such as third-generation cephalosporin or vancomycin to reduce the postoperative infection rate. Preoperative axillary hair removal is a common preoperative standard care, considered as a method for infection prevention. Marecek et al. [111] clipped one randomly selected axilla in 85 healthy male volunteers with commercially available surgical clippers. Aerobic and anaerobic culture specimens were taken from the clipped and unclipped axillae. Each shoulder was then prepared with 2% chlorhexidine gluconate and 70% isopropyl alcohol. Repeated culture specimens were then taken from both axillae. There was no difference in the burden of P. acnes between the clipped and unclipped axillae before or after surgical preparation (P = 0.109, P = 0.344, respectively). There was a significantly greater bacterial burden in the clipped shoulder compared with the unclipped shoulder before preparation (P < 0.001) but not after preparation (P = 0.285). There was a significant reduction in total bacterial load and P. acnes load for both axillae after surgical preparation (P < 0.001 for all). Removal of axillary hair showed no effect on the burden of P. acnes in the axilla. Clipped axillae had a higher total bacterial burden. A 2% chlorhexidine gluconate surgical preparation was proved to be effective at removal of all bacteria and specifically P. acnes from the axilla. The diagnosis of acute deep infection after the surgical fixation of proximal humeral fractures represents a challenge, and the diagnosis can be easily missed or delayed. Incisional erythema and drainage are findings common to all patients, and constitutional symptoms are rarely present. Laboratory data, such as leukocyte count, erythrocyte sedimentation rate and C-reactive protein, can assist orthopaedic surgeon with the diagnosis; however, they are by no means definitive [109]. The presence of a deep infection after ORIF can negatively affect the fracture healing process, determining septic non-unions. In this unlucky event, surgical management is mandatory for most patients, requiring combined surgical and medical procedures for tissue debridement and infection eradication. The standard treatment of acute deep infection includes serial surgical debridements and intravenous organism-specific antibiotics. Surgical debridement should be complete, removing all necrotic, non-viable and fibrinous debris. Fixation hardware should remain in situ to stabilize the fracture until healing occurs. Grossly loose hardware with a poorly stabilized fracture should be removed and revised. Patients should be informed that the results of treatment of acute deep infection after surgical fixation of proximal humeral fractures are beset with high complication rates, poor functional outcome and a notably high non-union rate [109].