GENERAL COMPLICATIONS

The incidence of infection after shoulder arthroscopy is low. This may be due to the result of the copious amount of irrigation that is constantly run through the shoulder during the procedure, as well as the minimal exposure of deep tissues to the environment. D’Angelo and Ogilvie-Harris⁶ reported on 9 cases of septic arthritis after arthroscopy, with an incidence of 0.23%. Based on their studies, they recommended routine antibiotic prophylaxis to prevent these potentially serious complications. Bigliani and coworkers⁷ reviewed the literature in 1991 and noted infection rates from 0.04% to 3.4%. Other reviews have noted similar infectious rates.^2,4 Brislin and colleagues⁸ published a recent study on 263 patients who had undergone arthroscopic rotator cuff repair; they noted only a single infection in this group, with an incidence of 0.38%. Athwal and associates,⁹ in 2007, reported on 39 cases of deep infection after rotator cuff repair. They noted that long-term functional limitations in these patients are not uncommon. They emphasized recognizing Propionibacterium acnes as the most likely causative organism and the need for at least 7 days of culture to identify this organism. P. acnes has also been implicated as a potential causative factor for the onset of chondrolysis after shoulder arthroscopy and therefore should be treated aggressively. Regardless of the true incidence, infection after shoulder arthroscopy requires a high index of suspicion with prompt management once deep infection is recognized.

General anesthetic complications germane to any surgical procedure also apply to shoulder arthroscopy. In addition, fatal air embolism¹⁰ and pneumothorax,¹¹ possibly related to the use of CO₂, pulmonary edema,¹² and pneumomediastinum¹³ have all been reported during shoulder arthroscopic procedures. It is important to educate patients about these risks for them to understand that minimally invasive procedures can still be associated with significant complications.

Of special relevance to shoulder arthroscopy is the increasing use of interscalene anesthesia for intraoperative pain management and postoperative pain control (Fig. 31-1). Although purported to be a safe and effective means for pain control, reports of seizures, cardiovascular collapse, and severe respiratory distress have been noted. Additionally, neurologic complications in the form of transient and even permanent brachioplexopathies are a known risk of regional brachial plexus blockade.^14,15 Lenters and coworkers¹⁶ have reviewed the experience of a single medical center over a 15-year period and supplemented this data with records from the national American Society of Anesthesiology Closed Claims Database. From the hospital experience, they reported a total of 41 major complications following interscalene block, with 14 of them still present, affecting comfort and function at the most recent follow-up. From the database, they noted 19 permanent complications and 4 deaths attributable to interscalene brachial plexus block. It appears that fewer complications arise as anesthesiologists become more experienced with the techniques and with the use of newer technologies, including ultrasound-guided blocks. However, complications can and do still occur. It is important that the anesthesiologists performing these procedures have significant experience and understand the risks, and that they communicate these to the patients. These procedures must be performed with the patient awake. Ultrasound guidance to direct needle placement is expected to decrease the risk of interscalene blocks.

FIGURE 31-1 Needle placement for interscalene regional anesthesia prior to positioning for shoulder arthroscopy. Proximity to vascular structures and apex of lung should be noted.

Deep venous thrombosis (DVT) of the lower extremity after shoulder arthroscopy has rarely been noted in the literature.^17,18 Although there have only been a few reported cases of life-threatening pulmonary embolism (PE), the prevalence of these complications are likely much more common than once thought. Standard precautions, such as using mechanical compression devices on the lower extremities during the procedure and early mobilization, may minimize the risk. Currently, there are no guidelines in place suggesting that the use of routine chemoprophylaxis is indicated after shoulder arthroscopy.

COMPLICATIONS SPECIFIC TO SHOULDER ARTHROSCOPY

Complications resulting from the two prevalent patient positions used during arthroscopy do occur. In the lateral decubitus position (Fig. 31-2), traction neuropraxia has been reported to occur up to 10% of the time.¹⁹ In 1987, Klein and colleagues²⁰ placed strain gauges on the brachial plexus of cadavers and measured the strain on the plexus in the lateral decubitus position with varying amounts of flexion and abduction. They found that increasing flexion and abduction decreased strain on the brachial plexus. Stanish and Peterson²¹ have reviewed the literature on neurologic injuries after shoulder arthroscopy and suggested that careful patient positioning, minimal traction to distract the joint (no more than 7 kg), limiting the length of the procedure, and accurate portal positioning could minimize neurovascular complications.

FIGURE 31-2 Lateral decubitus position for shoulder arthroscopy. The patient is positioned with a beanbag, with the right arm in longitudinal suspension.

The beach chair position (Fig. 31-3) was described in 1988 by Skyhar and associates.²² Its advantages include easier airway access, placing the anatomy in the standard upright position, the ability to move the arm into different positions intraoperatively and decreased risk of traction-related neuropraxia. Several reports of catastrophic complications have been reported recently with the use of the beach chair position. Pohl and Cullen²³ reported on four cases of shoulder surgery in the beach chair position that resulted in death in one patient and severe brain damage in three others. Stroke, brain death, loss of vision, and ophthalmoplegia have also been described.^24,25 These complications are thought to be attributable to errors in blood pressure reference points leading to decreased cerebral perfusion. In 2008, Papadonikolakis and coworkers²⁶ reviewed the literature and suggested using invasive blood pressure monitoring or placing a blood pressure cuff around the contra lateral brachium (at the level of the heart) to avoid erroneous pressure numbers. Currently, it appears that the choice of position should be left up to the surgeon. Recognition of the potential complications of the preferred position and taking preventative measures can minimize the occurrence of a significant untoward event.

FIGURE 31-3 Beach chair position for shoulder arthroscopy. The right arm is attached to a hydraulic positioning device that allows the arm to be manipulated during the arthroscopic procedure.

With the increasing complexity of arthroscopic procedures about the shoulder, the need to establish portals that allow for optimal access and precision placement of implants has also increased. Although vascular complications associated with portal placement remain rare,^4,27 direct neurologic injury remains a concern. Segmuller and colleagues²⁸ noted direct cutaneous nerve injury in 7% of their 304 patients undergoing shoulder arthroscopy. Most of these injuries involved a cutaneous branch of the axillary nerve at the site of the established lateral portal. This concern over direct nerve injury underscores the importance of accurate portal placement. In 2004, Lo and associates²⁹ studied the neurovascular anatomy in proximity to the most commonly used arthroscopic portals established via an outside-in approach. They noted that with the exception of the cephalic vein, all the neurovascular structures were more than 20 mm away from all the portals evaluated; it was concluded that the standard and accessory arthroscopic portals are safe when established in an outside-in fashion. Other considerations such as careful palpation of bony landmarks, accurate knowledge of shoulder anatomy, accounting for fluid extravasation, and the use of cannulas during portal placement may enhance safety during shoulder arthroscopy.

Iatrogenic tendon injury has been described in the literature. Norwood and Fowler³⁰ described this complication in 1989. The port of Wilmington portal is an accessory portal used primarily for repair of posterior and posterior-superior labral tears. Proper placement of the portal leads to penetration of the muscular portion of the supraspinatus. However, if improperly placed, injury to the tendinous portion of the supraspinatus can occur. It appears that accurate portal placement and a thorough understanding of shoulder anatomy can minimize this complication.

Fractures of the acromion, clavicle, and humerus can occur during shoulder arthroscopy. Fracture of the acromion has been extensively described in the literature.^4,5,31 This is usually a result of overzealous and inaccurate resection of the acromial spur, often in the setting of a patient with a thin acromion preoperatively (Fig. 31-4). Simple precautions can be taken to prevent this complication. Measurement of the thickness of the acromion on preoperative radiographs and attention to the depth of resection using the known diameters of the arthroscopic tools as a reference are helpful (Fig. 31-5). Additionally, adequate visualization and exposure of the anterior acromion to prevent beginning the resection too far posteriorly, as well as working from anterior-inferior to superior, will further minimize the chance of excessive resection. Fractures can be treated with screw fixation, tension band fixation, or excision of the remaining bone and repair of the deltoid attachment. Clavicle fractures can occur as a result of the surgeon becoming disoriented during subacromial surgery. At this point, the surgeon may mistakenly resect through the shaft of the clavicle instead of the acromioclavicular (AC) joint because of straying too medially during the procedure. Humeral fractures are also a concern during shoulder arthroscopy. Humeral neck or shaft fractures can occur as a result of excessive manipulation of a stiff shoulder (Fig. 31-6). Greater tuberosity fractures may follow improper instrument penetration (Fig. 31-7) or implant placement. Paying attention to the amount of force imparted during manipulation and careful attention to instrument placement and topography of the tuberosity can minimize the risk of these complications. When such a fracture occurs, it can be treated conservatively or by reduction and internal fixation, depending on the stability of the fracture.

FIGURE 31-4 Outlet view radiograph depicting an acromial fracture postoperatively after shoulder arthroscopy and subacromial decompression.

FIGURE 31-5 Schematic representation of the planes of resection for an acromioplasty procedure. The red lines demonstrate incorrect resection planes leading to inadequate and excessive acromial resection. The green line demonstrates a correct resection plane leading to adequate decompression without excessive resection.

FIGURE 31-6 AP radiograph demonstrating a greater tuberosity fracture (white arrow) as a result of overly aggressive manipulation of a stiff right shoulder under anesthesia.

FIGURE 31-7 Arthroscopic photograph of a right shoulder viewed from the posterior portal. Shown is a humeral osteochondral fracture caused by improper blunt trochar placement when attempting to establish the posterior portal.

Stiffness is probably the most common complication of any surgery about the shoulder. It can be a significant cause of morbidity, loss of function, and continued disability. Significant restriction of motion has been proposed to occur, ranging from 2.8%⁵ to 15%³ following shoulder arthroscopy. Stiffness is manifested clinically by decreased active and passive range of motion, often with a very profound loss of external rotation. This is often accompanied by a period of worsening pain after the postsurgical pain has abated. The physical examination demonstrates restricted glenohumeral motion, with a disproportionate amount of motion coming from scapulothoracic rotation. It is important to differentiate postoperative stiffness from restricted motion secondary to complex regional pain syndrome (CRPS). In the latter, physical examination findings of hyperesthesia, profound edema, excessive warmth, and neurologic dysfunction often accompany the lack of motion. If CRPS is suspected, a full workup, usually with the aid of a pain specialist, is required to manage the condition. Postoperative stiffness can be minimized, with early range of motion (ROM) and emphasis on scapular stabilization to encourage smooth glenohumeral motion.

With respect to stiffness, rotator cuff and instability surgery represent challenges for the surgeon. Too early and aggressive motion may predispose the repair to failure, whereas too conservative a rehabilitation program could increase the likelihood that arthrofibrosis may occur. In my experience, stiffness is best prevented with extensive patient education, very detailed physical therapy protocols, and frequent follow-up appointments to identify those patients who may be prone to developing stiffness. Understanding the difference between passive, active-assisted, and active ROM as well as the limits of a safe ROM are essential for adequate compliance. Additionally, frequent communication between the surgeon and physical therapist to formulate a unified rehabilitation plan is invaluable.

Correctly identifying those patients who are stiff preoperatively is as important as identifying postoperative patients who may develop stiffness. Often, pain restricts a patient’s ROM in the clinical setting, However, if the clinician fails to diagnose early adhesive capsulitis in this setting or if it is, for example, mistakenly attributed to impingement syndrome, the wrong course of treatment may be prescribed. Erroneous diagnosis in this setting may lead to unnecessary surgery and poor postoperative outcomes.

Despite the best efforts on the part of surgeon and patient, a small subset of postoperative shoulders will become stiff and not respond to conservative management. In this group, closed manipulation and/or arthroscopic release is highly effective to restore motion and relieve pain.^32,33 Currently, there is no consensus regarding the timing of manipulation or release in this population, with proponents of early (3 month) and late (more than 6 months) very prevalent.

Chondrolysis is a rare but devastating process that has been identified to occur after shoulder arthroscopy (Fig. 31-8). Although no specific cause has been elicited, several potential contributing factors have been explored. Reports of chondrolysis have been made following the use of an intra-articular pain pump.^34,35 Uncertainty exists about its contribution to articular cartilage damage. Continuous intra-articular infusion pumps have been studied in the knee and to the best of my knowledge, there have been no reports of chondrolysis. Pain pumps often infuse bupivacaine, a substance that has been implicated as chondrotoxic in various animal and in vitro models.^34–36 However, a 2008 study by Dragoo and coworkers³⁷ reported that exposure of cultured chondrocytes to bupivacaine up to a maximum of 48 hours was not significantly chondrotoxic, whereas more prolonged exposure led to increased chondrocyte death. Additionally, bupivacaine injections into the glenohumeral and knee joints are commonly performed in the orthopedically in clinical and operating room settings without significant observed rates of chondrolysis, lending more evidence that significant cartilage damage may not occur after a single exposure to intra-articular bupivacaine.

FIGURE 31-8 Arthroscopic photograph of a right shoulder viewed from the posterior portal. Significant chondrolysis of the humeral articular surface developed after an index arthroscopic shoulder procedure. This patient continued to have pronounced pain and disability after the index procedure, prompting re-exploration.

Several reports have associated chondrolysis with the use of a thermal probe for capsulorrhaphy or capsular release.^38–40 In 2009, Good and colleagues⁴¹ studied temperatures in the glenohumeral joints of cadavers using a thermal device in different flow states. They noted that the temperature in the joint rose above 45° C in all flow states but that higher flow states led to quicker dispersion of the heat than lower flow states. Irrigant temperatures above 45° C have been shown to result in chondrocyte death. They concluded that the use of a thermal probe might lead to temperatures high enough to cause chondrocyte death and that maintaining adequate fluid pump flow rates may help minimize exposure to high temperatures and protect the articular cartilage.

Many other factors have also been implicated in the development of chondrolysis, including infection, bioabsorbable implants, mechanical imbalance, and repetitive and single traumatic events. At present, there is insufficient evidence to suggest that there is any one predisposing cause for the development of chondrolysis; research to identify these complex interactions and their contributions to this disease process is ongoing.

Increasing numbers and types of implants used to stabilize and repair structures about the shoulder are becoming available. These implants have had instrument systems developed to aid their placement arthroscopically. Instruments can break during attempted implantation of an anchor, leading to incarcerated foreign bodies within the glenohumeral joint (Fig. 31-9).

FIGURE 31-9 Arthroscopic photograph of a left shoulder in the lateral decubitus position viewed from the anterior-superior portal. Shown is a broken drill bit incarcerated into the anterior glenoid rim during anterior instability surgery.

The surgeon has the choice to use metal, bioabsorbable, bioinert, or even recently developed biocompatible implants. Metal implants have the advantage of being able to be followed on routine radiography to confirm placement. Disadvantages include artifact creation on subsequent MRI evaluation, loosening, and migration (Fig. 31-10).⁴² To address these concerns, bioabsorbable implants have been developed and extensively used in place of metal implants. These implants have been scrutinized for complications, including loose body creation as a result of partial anchor resorption and osteolysis stemming from the biologic response as the anchor resorbs (Fig. 31-11).^43,44 Reported complications with the use of bioabsorbable implants are very rare, especially with newer generation implants with more reliable degradation profiles and sturdier suture-eyelet constructs.⁴⁵ Bioinert and biocompatible implants have also been developed to address concerns about the degradation and osteolysis profile of bioabsorbable implants. When these radiolucent implants are used, if they are inserted in a proud manner or become loose, they may cause significant articular cartilage injury as loose bodies before the cause is recognized (Fig. 31-12). One should harbor a high index of suspicion for this complication in the setting of a patient with mechanical symptoms and pain after implantation of radiolucent anchors.

FIGURE 31-10 AP radiograph depicting anchor migration and pullout after rotator cuff repair.

FIGURE 31-11 Arthroscopic photograph of a left shoulder in the lateral ducubitus position viewed from a lateral portal. The greater tuberosity is pictured, with extensive osteolysis resulting from a reaction to a bioabsorbable implant.

FIGURE 31-12 Arthroscopic photographs of a right shoulder in the lateral decubitus position viewed from the posterior portal. A, Glenohumeral joint with the humeral articular surface on the right, with anchor pullout as a result of rotator cuff failure. B, Subacromial space with additional anchor failure. The implants in these images are bioabsorbable. The patient complained of mechanical symptoms and increasing pain after surgery.

COMPLICATIONS ASSOCIATED WITH SPECIFIC PROCEDURES

Subacromial Surgery

Overall, subacromial surgery has the lowest rate of complications of all procedures in the shoulder, between 0.76% and 6.5%.^1,5 The most common reported cause of failure of arthroscopic acromioplasty is inadequate resection.⁴⁶ Difficult visualization is a significant factor in most inadequate resections. Visualization can be maximized with extensive bursectomy, meticulous hemostasis, relative hypotensive anesthesia, and adequate hydrostatic pressure. Excessive resection of the acromion should also be avoided (Fig. 31-13). As noted, preoperative planning with careful attention paid to acromial thickness and morphology can minimize its occurrence.

FIGURE 31-13 Outlet radiograph of a right shoulder showing excessive acromial resection. The red lines outline the edges of the resected acromion and distal clavicle.

Finally, postoperative AC symptoms can occur when the AC joint is violated during acromioplasty. Whether to leave the AC joint undisturbed, coplane the AC joint, or resect the distal clavicle when violation and inferior osteophytes are encountered remains controversial.^46,47 Biomechanical studies have demonstrated increased superior and anterior translation of the clavicle in coplaned specimens⁴⁸ and AC joint instability symptoms have been reported clinically after coplaning. However, additional studies have noted no increased incidence of AC joint symptoms, compromised clinical results, or additional surgery at an average follow-up of up to 6 years after coplaning the AC joint.⁴⁹

Rotator Cuff Repair

Arthroscopic rotator cuff repair has gained in popularity recently, with subjective and objective results similar to miniopen rotator cuff repair.⁵⁰ Failure of the rotator cuff repair remains a concern regardless of the technique employed. In 2007, Cole and associates⁵¹ evaluated cuff repair integrity in 49 shoulders at a minimum of 2 years follow-up using MRI and noted that 22% of repairs had a recurrent tear. Brislin and coworkers⁸ reviewed 263 repairs in 2007 and noted only one reoperation for repair failure. No follow-up imaging was used to identify a retear. Although cuff integrity is not universal following rotator cuff repair in asymptomatic shoulders, it is clearly associated with a better outcome.⁵² Several recent studies have demonstrated that the tear size at the time of repair and age are critical factors in assessing the risk of failure.^51,53 Additionally, other causes for failure can prove to be to the result of concomitant pathology not addressed or identified, including glenohumeral arthritis, impingement, AC joint pain, biceps problems, and even cervical-based pain origins.

During rotator cuff repair, visualization and arm positioning are key to avoid placing anchors too medially into the articular margin (Fig. 31-14). Visualization is improved with careful hemostasis, attention to intraoperative blood pressure, and extensive bursectomy. When placing suture anchors into the medial aspect of the greater tuberosity, the arm should be adducted as much as possible to avoid a shallow approach to the tuberosity and potential penetration of the humeral head. An incision just off the lateral acromion will also improve the trajectory of the anchor and decrease the risk of articular injury.

FIGURE 31-14 Arthroscopic photograph of a left shoulder in the lateral decubitus position viewed from the posterior portal. Shown is medial placement of a rotator cuff anchor resulting in damage to the articular margin of the humerus. The suture limbs of the anchor are shown by the black arrow.

Instability Repair

Recurrence after arthroscopic instability surgery remains a relatively common complication. Shaffer and Tibone⁴⁶ reviewed the literature on rates of recurrence and noted a 16% to 33% rate with staple capsulorrhaphy, 0% to 60% with transglenoid suturing, 0% to 37% with tack stabilization, and 0% to 30% with suture anchor repair. Additionally, each of these techniques has its own unique complications associated with it. Stabilization using a resorbable tack has been reported to demonstrate the rare occurrence of an exuberant synovitis caused by the degradation products of the tack.^54,55 Posterior to anterior pin placement with tacks can place the apex of the lung at risk with resultant pneumothorax.⁵⁶ Transglenoid suture capsulorrhaphy places the suprascapular nerve in danger (Fig. 31-15). The transglenoid method has also been associated with fistula and cyst formation⁵⁷ and intrathoracic penetration with the guide pin.⁵⁸

FIGURE 31-15 Patient with profound infraspinatus atrophy of the left shoulder secondary to suprascapular nerve injury after an arthroscopic shoulder procedure. Nerve conduction studies confirmed the injury.

Recently, suture anchor capsulorrhaphy has become the gold standard in arthroscopic stabilization. Several studies have claimed that with newer techniques and improved anchors, outcomes of arthroscopic stabilization are approaching those of open technique, with the potential benefit of less morbidity and improved ROM.^59–61 This view is not without some controversy. A meta-analysis performed by Lenters and colleagues⁶² in 2007 noted that arthroscopic suture anchor techniques are associated with significantly higher risks of recurrent instability and dislocation than open methods. However, the arthroscopic repairs were associated with higher Rowe scores than the open methods, providing evidence that function and motion may be better preserved using arthroscopic approaches.

Suture anchor placement during arthroscopic capsulorrhaphy requires a detailed three-dimensional knowledge of glenoid anatomy to prevent anchor malposition (Fig. 31-16). Understanding the slope of the glenoid neck, glenoid version, and angle of approach during anchor placement is critical. Similar to rotator cuff repair, choosing metal or radiolucent implants will determine whether implant position can be followed postoperatively with standard radiography.

FIGURE 31-16 AP radiograph demonstrating improper placement of a suture anchor into the inferior glenoid during instability surgery. The anchor (black arrow) has penetrated the inferior glenoid cortex.

SUMMARY

Shoulder arthroscopy appears to be a safe and effective technique based on a review of the literature. Despite the increasing complexity of procedures being done, complication rates have remained steady. However, serious and even life-threatening complications, although rare, can and do occur. Decisions made before (e.g., choice of position, type of anesthesia), during (e.g., choice of implant, use of thermal devices, portal placement) and after surgery (e.g., rehabilitation, use of pain catheters) can affect the complication profile and rate in an individual practice. A thorough understanding of shoulder anatomy, implant design, and patient risk factors can help minimize complications. Finally, when complications do occur, early recognition and management can minimize the morbidity of these procedures.