The technique of arthroscopic subacromial decompression (ASAD) and distal clavicle resection has become an increasingly common procedure for dealing with impingement and acromioclavicular (AC) joint disease. Many surgeons are now routinely combining these arthroscopic decompression techniques with either mini-open or, more recently, completely arthroscopic repair of rotator cuff tears. As with other open operations that have evolved arthroscopically, the learning curve for these procedures is significant and should not be underestimated.
Complications associated with shoulder arthroscopy in general are low. Small, in his 1986 study , found a complication rate in subacromial space surgery of 0.76%. However, the complication rate with anterior staple capsulorraphy was 5.3%. The rates excluded clinical failures. In his follow-up study in 1988 [2,3], reporting on complications relating to arthroscopy performed by experienced arthroscopists, the complication rate with shoulder arthroscopy was 0.7% overall, again with the highest rate in anterior capsulorraphy.
Curtis et al.  reviewed 711 shoulder arthroscopies and found an overall complication rate of 6%. Of the 43 complications, 19 were secondary to postoperative stiffness, six secondary to transient neurologic symptoms, five associated with wound hematomas, and six associated with bruising. One patient each had problems with hardware removal, reflex sympathetic dystrophy, laceration of the cephalic vein, pulmonary embolus, and painful posterior portal, corneal abrasion, and heterotopic bone. The rate increased from 4.5% for arthroscopic procedures to 8.0% for combined arthroscopic and open surgery (i.e., mini-open cuff repair or stabilization).
Several of the complications associated with shoulder arthroscopy are related to arthroscopic technique in general and are common to other joints that are arthroscopically examined . In a survey sent to arthroscopy association members of the shoulder study group, however, some procedure-specific complications were also identified . This chapter identities some of the difficulties inherent in these arthroscopic techniques and provides some suggestions for precautions and modifications that may help in their avoidance.
For the most pant, neurologic complications have been associated with the lateral decubitus position for shoulder arthroscopy (Fig. 12-1). Neuropraxia involving traction injury to the brachial plexus may be secondary to the traction weight, direction of pull, and duration of the surgery. Five to 10 Ibs of distal traction is usually adequate for the average patient, and 15 Ibs is reserved for larger or well-muscled individuals. Increasing the weight to 20 Ibs or more, as was initially done in shoulder arthroscopy, results in changes in the somatosensory-evoked potentials of the musculocutaneous, median, ulnar, and radial nerves, with the musculocutaneous being the most sensitive at all arm positions and traction weights .
In a cadaver study, Klein et al.  demonstrated the greatest brachial plexus strain with the arm at 70° abduction and 30° of forward flexion. The minimum overall strain was noted at 90° of flexion and 0° of abduction, but this resulted in poor visualization. They recommended positions of 45° of forward flexion and either 0° or 90° of abduction, depending on the intra-articular region of interest.
Dual traction as described by Gross and Fitzgibbons  (see Fig. 12-1) with low distal traction weights on a minimally abducted arm coupled with a laterally directed distraction force appeared to be associated with less compromise to somatosensory-evoked potentials in Pittman et al.’s study . Gross and Fitzgibbons  also recommended rolling the patient back about 25° to 30° to orient the glenoid joint surface parallel to the floor. The rollback position coupled with 15° of additional flexion of the arm puts the direction of pull into Klein et al.’s safer zone . No work has been specifically directed at determining the neurologic effects of the laterally directed distraction force in this set-up, but no reported complications have been associated with this type of traction.
For arthroscopic surgery in the subacromial space, minimal abduction (15° to 25°) and 15° of flexion from the rollback position opens up the space and yet does not put excessive strain on the brachial plexus. Excessive forward flexion of the arm brings the tuberosity into contact with the anterior acromial hook, making exposure difficult.
Careful attention needs to be given to the position of the head, which should be as close to exactly neutral as possible. Any sagging of the head down and away from the operative arm increases the strain on the brachial plexus. Overcompensation and excessive propping of the head away from the “down” arm can result in the opposite brachial strain. Careful padding and wrapping of the traction device at the wrist are necessary to avoid compression injury to the sensory branch of the radial nerve with resultant thumb numbness. Moreover, careful padding of the ulnar and peroneal nerves on the downside is necessary . Time in traction is also a factor, and conversion to an open operation is recommended if the operation is extending past 2 hours or if distention is becoming severe.
The beach-chair position alleviates most of the neurologic concerns already stated , but careful positioning and support of the head are still necessary. A case of hypoglossal nerve injury has been reported with this position . Exposure in the subacromial space, however, may be diminished because of loss of distraction. The dual traction technique is not possible with this position.
General anesthesia provided to the patient in the lateral decubitus position appears safe relative to hypotensive and neurologic problems, so long as proper padding has been established. Interscalene nerve blocks are commonly used for either intraoperative anesthesia or additional postoperative pain relief. A temporary ipsilateral phrenic nerve palsy routinely results from this block but rarely causes pulmonary problems except in patients with preexisting pulmonary insufficiency [13,14].
Esch and Baker , however, reported on two patients requiring ventilatory support after interscalene block for ASAD. The anesthesia literature documents various relatively significant complications with interscalene blocks, including bilateral spread affecting both phrenics; complete spinal, bilateral cervical, and thoracic epidural blockade; prolonged Horner’s syndrome; auditory disturbance; and cardiac arrest [16-22]. Pneumothorax caused by incorrect needle placement has also been reported . Complications associated with interscalene block appear to be more common when the block is performed after induction of a general anesthetic as opposed to when the patient is awake and a nerve stimulator has been used.
Direct nerve injury can be associated with incorrect portal placement.
The traditional posterior portal as described by Andrews et al.  has become the standard position for the initiation of glenohumeral and subacromial arthroscopy. This penetrates the so-called soft spot approximately 1 cm medial and 1 to 2 cm inferior to the posterolateral corner of the acromion. The arthroscope should enter the joint approximately in the interval between the infraspinatus and teres minor muscles. This portal passes through the deltoid, ranging from 2 to 4 cm from the axillary nerve and the posterior humeral circumflex artery, and lies approximately 1 cm lateral to the suprascapular nerve and artery.
Inferior medial migration of this portal as described by Wolf  for the central posterior portal or inferior lateral migration for his modified posterior portal  places these structures at greater risk. Directing a blunted conical trocar toward the coracoid process provides some increased margin of safety for these posterior portals. Sharp trocars should generally be avoided for shoulder arthroscopy because of the increased risk of neurologic and chondral damage.
Several anterior shoulder arthroscopic portals have been described [24,25,27-30]. The anterior superior portals as described by Andrews et al.  and Wolf [25,26] and the superolateral portal described by Laurcncin et al,  are neurologically safe and most useful for subacromial surgery. They are also readily used to provide an anterior viewing portal for glenohumeral work. The central anterior portal in the superior recess above the subscapularis tendon as described by Matthews et al.  also appears to be safe relative to neurovascular structures. As the surgeon moves inferior to the tip of the coracoid process, the risk to neurovascular structures increases. These portals as described by Wolf (anterior-inferior portal) , Resch el al. , and Davidson and Tibonc  (anterior-inferior transubscapular) are more useful for arthroscopic instability surgery and are not generally used for subacromial or rotator cuff work (Fig. 12-2).
External (A) and arthroscopic (B) views of the anterior portals, including the superolateral , anterior-superior , anterior , central anterior , anterior-inferior , and anterior-inferior trans-subscapular [28, 29] portals.
The supraclavicular fossa portal was devised by Neviaser ; it allowed placement of an additional inflow portal at the posterior superior aspect of the joint, as well as access for superior instrumentation. The suprascapular nerve and artery lie deep and on the inferior surface of the supraspinatus muscle, approximately 2 cm medial to the path of the cannula. Too vertical a passage can injure the suprascapular nerve and artery; too lateral a passage can injure the rotator cuff tendon, particularly if the arm is abducted more than 30° . This portal is no longer routinely needed or used, particularly for subacromial work (Fig. 12-3).
Multiple subacromial portals have been described for decompression and for AC joint and rotator cuff surgery. These include the traditional posterior portal with angulation of the arthroscope or instrument superiorly into the subacromial bursa, the posterolateral portal , the central lateral portal , the anterolateral portal, the anterior and posterior AC joint portals , and the accessory high portals for anchor placement in rotator cuff surgery (see Fig. 12-3).
The neurologic structure most at risk with the use of these portals is the axillary nerve, which traverses the underside of the deltoid muscle approximately 3 to 5 cm from the acromial margin. If the surgeon keeps the skin incision less than 5 cm from the acromion and directs the trocar toward the subacromial space, as opposed to directly down through the deltoid muscle, axillary nerve problems should be avoided . These portals are best set up after preliminary placement and direct visualization of an 18-gauge needle to ensure the proper orientation for shaving or anchor placement.
Problems related to hypotension have been associated with arthroscopy done with the patient in the beach-chair position, especially in elderly or hypertensive patients. Adequate fluid replacement and compression leg stockings may be beneficial to avoid premature termination of the procedure or switching to a lateral decubitus position. Although deep venous thrombosis is rare, it has also been reported with shoulder arthroscopy .
Likely, the most common vascular complication with subacromial surgery is bleeding. Because the subacromial area is extensively traversed by veins, frequently inflamed, and not a closed space, bleeding is more troublesome here than at almost any other arthroscopic site. Failure to control bleeding and to maintain visualization and orientation are common sources of complications in subacromial surgery. Use of electrocautery is strongly recommended. Strategies currently used for the control of bleeding include:
Pneumothorax is a known complication associated with interscalene block anesthesia . There have been rare cases of subcutaneous emphysema, pneumomediastinum, and potentially life-threatening tension pneumothorax associated with arthroscopic decompression .
Skin burns have been reported with shoulder arthroscopy if a noninsulated cautery tip is used with conductive solution. This problem can be avoided if the surgeon uses an insulated tip, or newer bipolar devices for ablation and cautery. These tips can be used safely even in conductive solutions such as normal saline or lactated Ringer’s solution.
Sterile water, which was used early on because of its nonconductivity, is injurious to soft tissues. Scattered reports exist of skin and muscle necrosis associated with extremely long procedures using water as an irrigating solution. Glycine (1.5%), used in urologic procedures, and less frequently as an arthroscopic medium, has been associated with transient blindness and is no longer recommended .
Extravasation and distention of the soft tissues with saline or lactated Ringer’s solution may sometimes produce alarming appearances. Studies have shown, however, that the intramuscular pressures rapidly return to normal at the end of the procedure [40,41]. The effect of soft-tissue distention on the nerves surrounding the shoulder has not been well-documented.
The infection rate resulting from arthroscopy in general is very low. Johnson et al.  reported less than one infection in 2000 new arthroscopies when using 2% glutaraldehyde as a sterilizing agent. Only four infection cases with shoulder arthroscopy have been noted in the literature to date [1,15,43].
Glutaraldehyde solution has been commonly used for instrument sterilization, especially in the outpatient setting. However, instruments must be thoroughly rinsed before use. Even trace amounts (such as may be found in arthroscopic rinse baths) can induce a severe synovial inflammatory reaction . Because of this and because of concerns regarding HIV transmission through the use of glutaraldchyde, sterilization is increasingly being performed by automated sterilization units such as the Steris (peracetic acid; Steris Co, Mentor, OH)  or Sterad (gas plasma with hydrogen peroxide;
J & J Medical, Arlington, TX) .
The potential for an equipment failure increases with the complexity of the procedure. Cannulated suture hooks and punches, various suture retrievers, linear punches, and grasping forceps can break off in the joint. Keeping a retrieval instrument, such as the magnetic Golden Retriever suction device (Instrument Makar, Okemos, Ml), handy can considerably simplify recapture of metallic pieces.
Arthroscopic Subacromial Decompression
Complications of ASAD include 1) variable bone resection, 2) deltoid detachment, 3) heterotopic bone, and 4) residual coracoacromial ligament (CAL) snapping.
Variable Bone Resection
Variable hone resection is probably the most common complication of ASAD. Both inadequate decompression and excessive resection have been reported. Wolf  reviewed 35 patients with failed previous arthroscopic surgery of the shoulder. Of these patients, 60% failed because of previous inadequate ASAD; 20 of 21 had complete recovery after further ASAD . Matthews et al. [27) and Esch  have reported on both acromial and clavicular fractures secondary to excessive resection.
Inaccurate decompression is usually secondary to inadequate preoperative planning with or without poor visualization and orientation during the procedure.
Outlet and axillary views are key to evaluating the acromion. The outlet view is used to determine the shape of the acromion (type I, II, or III) and the overall thickness [49,50]. Rockwood and Lyons  have described a modified anterior shoulder view that, although helpful in making the diagnosis of impingement, is not beneficial in terms of preoperative planning. On the outlet view, two lines are drawn on the undersurface of the acromion – the first from the front tip of the acromion to the posterior edge, and the second along the posterior half of the under-surface of the acromion extending out anteriorly. The distance between these two lines at the anterior margin approximates the amount of undersurface anterior bone that will be resected (Fig. 12-4).
The axillary view is used to determine the shape of the acromion (cobra-shaped vs. square-tipped), as well as to determine whether there is any “anterior acromial protuberance”  anterior to the level of the AC joint (Fig. 12-5). This approximates the amount of bone that will be removed anteriorly in addition to the amount of bone that will be taken from the undersurface.
After these measurements have been determined, the two-portal technique of acromioplasty makes it relatively simple to reproduce this resection. It is difficult to obtain a flat acromion when visualizing only from the posterior portal because of the amount of curving away of the acromion from the arthroscope . The acromion may appear flat from medial to lateral and front to back, but may still have a considerable anterior-to-posterior concavity when later viewed from the lateral portal. Placing the arthroscope laterally and then bringing the shaver forward from the posterior portal using the posterior half of the acromion as a “culling block”  helps ensure a straight, flat cut in the sagittal plane. This technique reliably converts a type II or type III acromion to a type I flat surface as demonstrated on postoperative radiographs (Fig. 12-6). The surgeon should be sure to subsequently replace the arthroscope posteriorly and to confirm flatness of the acromion in the medial to lateral plane. The anterior lateral acromial comer is often difficult to visualize from the lateral portal. Good visualization from both portals assures a flat, smooth surface.
(A) Preoperative templating for arthroscopic subacromial decompression (ASAD) on outlet- view radiograph. Dolled line indicates correct line of bone resection.
(B) Postoperative radiograph of ASAD.
The surgeon should beware of the thin curved type acromion (type C) . If, on the outlet view, a very thin or curved acromion is found, the cutting block line on the undersurface of the posterior half of the acromion may actually exit from the superior aspect of the acromion, taking off too much anterior bone (Fig, 12-7A) . In such cases, the cutting block technique would be inappropriate. In this situation, the original resection technique as described by Ellinan  is more applicable, that is, removing only a small anterior hook and not producing a type I flat acromion (Fig. I2-7B).
Preoperative (A) and postoperative (B) radiographs of decompression on the thin, curved acromion. Dotted line in (A) indicates excessive bone resection with cutting-block technique. Postoperative radiograph demonstrates more conservative anterior resection with significant increase in anterior acromial humeral distance.
This finding is usually secondary to excessive bleeding (which can be managed as previously outlined), poor localization of the subdeltoid bursa, or inadequate debridement of the subacromial space. Remembering that the bursa is an anterior structure, the surgeon should make every effort when in the subacromial space to direct the arthroscope into what Wuh and Snyder  termed the room with a view. Time and care should be spent at the beginning of the procedure, debriding the bursitis and the thickened periosteum on the undersurface of the anterior half of the acromion. Some of the posterior bursal curtain may need to be resected to clearly visualize the bony architecture. Debridement can be done with a shaver-burr, or a cautery-ablation system, but the surgeon should be sure to stay on the acromial bone and not deviate into the deltoid fibers. Debridement should be performed from the anterior lateral corner of the acromion toward, but not into, the AC joint, and then posteriorly along the lateral edge of the acromion. Spinal needles are used in the anterior lateral corner and the AC joint to gain better clarification of the bony landmarks.
A burr should next be used to resect 3 to 4 mL of bone, again along the anterior margin of the acromion from the anterior lateral corner to the AC joint, then tapering posteriorly along the lateral edge of the anterior half of the acromion. This improves orientation and visualization when the arthroscope is placed in the lateral portal and the shaver is brought in posteriorly. The surgeon should not resect too much anterior bone – only enough to make it easy to delineate the anterior edge of the acromion as seen from the lateral portal. The surgeon should let the shaver from the posterior portal resect most of the bone, coming forward in a smooth, controlled, flat cut (Fig. 12-8)The amount of bone resected is easy to determine by comparing the anterior remaining ledge with the thickness of the burr.
Serial intraoperative views of arthroscopic subacromial decompression of the right shoulder.
A) Conservative anterior resection from the lateral portal with the shaver tip in the coracoacromial ligament.
B) Lateral view of the burr starting forward during cutting block resection.
C) Completed resection with flat acromial undersurface posterior to anterior and intact deltoid fascia.
The posterior portal (through the deltoid muscle) must be at the inferior edge of the acromion, and not further below with soft-tissue interposition, so that the shaver does not angulate superiorly in an excessive or artificial manner (Fig. 12-9). Although the same placement of posterior skin incision is used for both glenohumeral and subacromial arthroscopy, the burr needs to puncture the soft tissues right at the inferior edge of the acromion for successful two-portal cutting-block technique.
Although the same skin incision is utilized for both the glenohumeral and subacromial examinations, the trocar penetrates the deeper soft tissues at different levels, so the shaver can be closely applied to the undersurface of the acromion.
If portal or soft-tissue penetration is too inferior, the burr will angulate too far superiorly and excess resection will occur.
Deltoid detachment results from overly aggressive resection at the anterior aspect of the acromion. If no significant anterior acromial protuberance is seen on the axillary radiograph, then simply flattening the undersurface of the acromion will adequately decompress it.
When a protuberance does exist, it usually is an extension of calcification interiorly into the CAL. Resecting the ligament and the contained bone with subsequent flattening of the acromion will usually eliminate the anterior overhang. Routinely resecting 8 to 10 mm of full-thickness anterior bone (as described for open procedures ) from the lateral portal puts the deltoid attachment at significant risk. This damage cannot subsequently be repaired unless the shoulder is then opened. The surgeon should take a small amount of anterior bone from the lateral portal and the bulk of the bone from the posterior portal using the cutting-block technique, thus teasing the bone off anteriorly from the fascia. Deltoid detachment, either open or closed, is arguably the most devastating complication of shoulder surgery and should be avoided .
This finding has been reported to be associated with both acromioplasty and AC resection [55,56]. In the 10 cases reported by Berg et al. , eight developed recurrent impingement symptoms, with an apparent strong correlation with active spondylitic arthropathy or a profile of hypertrophic pulmonary osteoarthropathy – male, obese, smoker with chronic pulmonary disease. They recommended prophylactic measures (indomethacin or radiotherapy} with these two types of patient groups. The surgeon should also unplug clogged shavers and burrs (or use an accessory gravity drainage portal) to avoid debris (“clouds of snow”).
Continued snapping with abduction and rotation maneuvers from an inadequate resection and a rescarring of the CAL do occur occasionally. After bony decompression has been completed, another 5 to 10 mm of ligament can be resected using a shaver, basket punches, or an ablation device. This is especially appropriate if snapping of the biceps tendon or bursal fold on the CAL is an identifiable preoperative complaint. Partial release of the anterolateral band alone may be curative in some athletes engaged in overhead throwing maneuvers |57]. Care should be taken to avoid excessive release of the lateral extension of the ligament along the lateral edge of the acromion. The ligament and deltoid fascia are intimately connected at this location, and release risks the deltoid attachment [58-].
Resection should not be done in the presence of significant degenerative arthritis of the glenohumeral joint or cuff arthropathy with a massive cuff tear, or if future arthroplasty is contemplated .
Arthroscopic Distal Clavicle Resection
Thorough clinical and radiologic evaluation of the AC joint should be performed prior to decompression. Significant inferior osteophytes should be noted on the anteroposterior view, and narrowing and sclerosis (degenerative joint disease) or widening and cystic changes (osteolysis) should be noted on the anteroposterior and axillary views. Differential injection into the AC joint instead of into the subacromial bursa may be necessary to confirm AC involvement.
Depending on the findings listed above, a decision must be made preoperatively (if possible) regarding the AC joint and distal clavicle. The surgeon must decide whether to 1) leave the AC joint untouched, 2) bevel the distal clavicle, or 3) perform an arthroscopic distal clavicle resection.
Most early descriptions of ASAD recommended routine beveling of the distal clavicle [33,60]. However, this practice destabilizes the AC joint to a certain extent by resecting the’ weaker inferior ligaments. The author has seen two patients in his own practice (and anecdotal reports from others) who, after decompression and beveling, developed AC joint pain necessitating later distal clavicle resection. Prior to surgery, these patients appeared clinically and radiographically to have normal AC joint.
If there are significant inferior osteophytes off the clavicle, then there is likely already compromise of the inferior AC ligaments and enough direct irritation of the underlying cuff to warrant beveling the tip. As the beveling is performed, downward pressure on the clavicle will bring some of the articular surface into view. If it appears significantly arthritic, then a complete resection of 1.0 lo 1.5 cm of clavicle should be performed.
However, in patients with AC joints appearing normal on preoperative clinical and radiographic examinations, the author no longer routinely bevels the clavicle; instead, the medial acromial bone is leased off the capsule and cartilage, much as is done with the deltoid fascia during decompression.
Other complications associated with distal clavicle excision relate to )) heterotopic bone formation, 2) inadequate resection, 3) underlying muscle injury, and 4) excessive bleeding. Incomplete resection of the superior cortical bone during distal clavicle resection is not uncommon. Clear visualization of this area using either a 50° or 70° arthroscope is necessary to remove all the superior bone (Fig. 12-10).
A) View from the posterior portal looking up at the acromioclavicular (AC) joint (arrowhead) with the inferior half of the distal clavicle already resected from the right shoulder.
B) Posterior view of the AC space with the distal clavicle resected and two spinal needles placed externally to measure the distance between the clavicle tip and the medial edge of the acromion (right).
C) Lateral view of the clavicle resection with the posterior and superior ligaments intact (arrow).
If a conical eggshell of bone is left behind, elevation or cross-chest maneuvers by the patient will remain painful. The bone will also serve as a nidus of heterotopic hone formation (Fig. 12-11).
Heterotopic bone formation after distal clavicle resection (right shoulder).
A) Six month follow-up radiograph showing early heterotopic nidus.
B) Two-year postoperative radiograph demonstrating mature bone in the AC interval.
The optimal amount of bone to be removed arthroscopically from the tip of the clavicle remains unresolved. If the superior and posterior AC ligaments are well maintained with the resection, the length of clavicle 10 be removed can be reduced [61 ]. If the superior and posterior ligaments are violated, however, then the remaining tip of the clavicle becomes more unstable, and further resection is needed [62,63«|. Studies suggest that 1 lo 1.5 cm of bone resection would be adequate with use of the arthroscopic technique (Fig. 12-12); 1.5 cm of bone should be resected if the posterior-superior ligaments have been compromised.
Preoperative (A) and postoperative (B) radiographs of the acromioclavicular resection (right shoulder).
Care should be taken to measure the distance between the clavicle and the acromion with two 18-gauge spinal needles placed parallel through the skin from above. This should be performed at both the anterior and posterior aspects of the clavicle (see Fig. 12-10B). It is easy to obtain an uneven gap in resection with more bone removed anteriorly than posteriorly, which should he avoided.
Caution should be taken when using unhooded burrs to resect the tip of the clavicle because it is very simple to wrap up the soft underlying cuff musculature in the instrument. The author prefers to use a well-hooded burr with the open side always either facing up or in, toward the cancellous middle of the clavicle. Suction should be low, just sufficient to clear debris.
Vascularity around the tip of the clavicle and AC joint is plentiful. Cauterization of the fat pad beneath the AC joint before debridement is helpful. t is also beneficial to outline the tip of the clavicle frequently with the cautery device when it is being resected medially because periosteal vessels are numerous.
Arthroscopic Rotator Cuff Repair
The rotator cuff should be thoroughly evaluated arthroscopically both from the articular and bursal sides. Partial tears are usually well-managed with limited debridement and ASAD. Excessive debridement of partial tears can lead to complete rotator cuff tears if caution is not exercised. Evaluation is aided by placement of “suture markers” – an 18-gauge needle placed from a superior position through the cuff and into the joint with a #1 monofilament suture grasped in the joint and brought out through the anterior portal as the needle is removed. This allows the investigator to closely examine the exact area on the bursal surface of the cuff that corresponds to the torn area on the articular side. Nearly complete tears that will not heal with adequate strength should he completed and repaired [64|. Several clinical studies have demonstrated that repair of rotator cuff tears in conjunction with ASAD fares better in the long run when compared with simple debridement and decompression [52,64,65].
Failure of fixation of rotator cuff repair is a common problem with both open and arthroscopic repair. This may be due to mechanical factors, biological factors, or both.
Several factors related to technique can affect the mechanical strength of rotator cuff repair and increase the likelihood of success. Anchor fixation into bone can be improved by roughening the area of the tuberosity widely to increase the surface area for cuff repair, while decorticating lightly. Creation of deep troughs in the soft cancellous bone lead to anchor pull-out (or tunnel breakage during open techniques). St. Pierre et al.  demonstrated good healing in animal studies without the need of a deep trough.
Inserting the anchor at about a 45° angle to the direction of pull (Burkhart’s “deadman’s angle”)  results in increased resistance to pull-out and puts the anchor under the stronger subchondral bone medially. Recent studies also suggest that simple suture may be stronger than mattress sutures in the fixation of tendon to bone in the rotator cuff area .
With the 45° “deadman’s angle” approach to the bone, the suture is pulled at an acute angle, with repetitive tension over the medial lip of the anchor hole. If the edge is too sharp, fraying and breaking of the suture may ensue. If the initial drill has a slight bevel at the stop point or if the hole is subsequently chamfered, the results of this problem may be diminished.
The vascular involvements of rotator cuff repairs in older patients are always in question. Factors that may affect the blood supply to the repair should be kept in mind. Debriding the edges of a rotator cuff back to bleeding tissue stimulates an acute healing response.
Reducing tension on the rotator cuff repair in its early healing phase improves circulation and healing potential. This reduction can be accomplished by fixing small tears where they appear (i.e., more medially than the normal tuberosity attachment). Burkhart has described the principle of “margin convergence”  in the reduction of tension on rotator cuff tissue and repairs. By repairing the larger tears with side-to-side sutures medially and then working laterally to fix the remaining Y-shaped or L-shaped tissue to bone using anchors, tension on the repair and susceptibility to anchor pull-out are reduced. This appears to be a valid principle in the author’s practice.
The larger the tear, the more beneficial an abduction pillow may be to improve blood flow to the “critical zone” and again reduce tension on sutures. Careful monitoring of postoperative rehabilitation is essential. The goal is to protect the repair and to avoid development of excessive scar tissue formation in the subdeltoid space – a “captured shoulder,” as described by Mormino et al. . Passive motion is less stressful on the repair in the early phases. Debate exists as to when to allow active abduction; this decision should be influenced by the size, pattern, and vascularity of the tear, as well as the stability of fixation.
Symptoms that persist despite adequate decompression or distal clavicle excision may be secondary to incorrect diagnosis. Decompression in a patient with secondary impingement from underlying anterior instability may often be unsuccessful until the underlying instability has been addressed. Posterior superior impingement is not likely to respond to anterior decompression. Underlying glenohumeral arthritis in weightlifters may negate the beneficial effects of distal clavicle excision for osteolysis. Suprascapular nerve syndrome must be diagnosed using specific neurologic modifications of standard electromyographic technique. A high index of suspicion for this entity must be maintained. Radicular cervical disease, metastatic carcinoma of the scapular neck, Pancost’s tumors of the lung, and referred pain to the shoulder from visceral structures are also part of the differential diagnosis. Postoperative pain associated with a cracking sensation is most likely secondary to a fracture of the acromion.
Norwood and Fowler  reported on four cases of recurrent symptoms after technically well-performed shoulder arthroscopy, secondary to persistent cuff tears. These appeared to be related to inadequate healing on the articular side of the tendinous portion of the cuff at the posterior portal site. Because larger cannulae are now being used routinely, this problem may become increasingly noted, both anteriorly and posteriorly. If persistent or recurrent pain and weakness occur after arthroscopy, repetition of the arthrogram may be worthwhile. If results are positive, an open repair is likely to be beneficial.
Inadequate Surgical Preparation
Surgical preparation entails not only physician training, preoperative planning, and equipment requirements, but also patient education. Many patients have unrealistic expectations as to the results that arthroscopic surgery can accomplish relative to the shoulder. Education about soft-tissue healing times and scar tissue maturation should help temper unwarranted enthusiasm and activity. Because pain associated with arthroscopic procedures is often reduced, careful monitoring of postoperative activity, especially with rotator cuff repair, is necessary.
Physician preparation is mandatory for successful surgical results. Training at meetings and cadaver laboratories, such as the Orthopedic Learning Center (Chicago, IL), is prudent before attempting new techniques in one’s practice. These procedures are equipment intensive, and a step-wise progression from open to mini-open to arthroscopic technique is recommended.
Arthroscopic subacromial decompression, distal clavicle excision, and rotator cuff repair are demanding operative procedures. It is hoped that diligent preoperative planning and intraoperative attention to the possible complications presented will increase the potential for successful surgical outcomes.
Recently published papers of particular interest have been highlighted as:
* Of Interest
** Of Outstanding Interest