Thermal capsulorrhaphy has been implicated in glenohumeral chondrolysis observed in association with arthroscopic stabilization procedures [32, 44]. Good et al. reported a series of eight patients previously treated with thermal capsulorrhaphy. Six of the eight patients were diagnosed with grade 4 humeral head and glenoid cartilage loss during repeat arthroscopy a mean of 8.2 months following the index procedure [17]. In a cadaveric study investigating the effect of radio-frequency probe use and varying rates of fluid flow on joint fluid temperatures, fluid temperatures were raised above a safe level (defined as below 45 °C based upon basic science data related to temperature-related chondrocyte death) in all testing conditions [28, 54]. Intermittent heating with 100% flow states resulted in the lowest average maximum joint fluid temperature. When the no-flow state was tested in conjunction with intermittent heating, the average maximum joint fluid temperature was 57.75 °C ± 20.07 °C. Moreover, under these same conditions, joint fluid temperatures required an average of 156 ± 38.64 s to return to a safe level [16]. These data underscore the importance of judicious use of radio-frequency energy in the glenohumeral joint and the importance of adequate fluid egress in order to normalize joint fluid temperatures and avoid chondral injury.

An emerging body of basic science and clinical results has also convincingly shown the negative effects of intra-articular bupivacaine pain pumps. Hansen et al. reported on a series of 12 of 19 patients treated with intra-articular bupivacaine pumps in conjunction with shoulder arthroscopy that developed glenohumeral chondrolysis [21]. Moreover, in several animal models, Chu and associates have demonstrated that local anesthetics as a class may be harmful to chondrocytes, providing further evidence against the use of intra-articular anesthetic pumps as an adjunct for controlling postoperative pain [5, 6, 30].

Nonmetallic, bioabsorbable suture anchors for arthroscopic stabilization were developed to provide a reliable means of refixation of avulsed capsuloligamentous tissues to the bone while increasing the likelihood of reintegration of autologous tissues as the implants gradually degrade. Recently, several reports have raised concerns regarding the in vivo behavior of poly-L-lactic acid (PLLA) implants. Specifically, substantial rates of intra-articular anchor debris, synovitis, and high-grade chondral damage have been observed in association with PLLA implants [12, 20, 37]. However, the rate of complications from biodegradable anchors is extremely low, and we believe that they are much safer than metallic anchors, which can cause severe articular cartilage damage when they are left “proud” in the shoulder.

Errors in the placement of suture anchors may manifest as recurrent instability or glenoid rim fracture. As higher rates of repair failure have been observed with repairs using less than three anchors, arthroscopic Bankart repair should involve a minimum of three anchors placed below the 3 o’clock position, with the first anchor placed as inferior on the glenoid as possible (between the 5 o’clock and 6 o’clock positions) [3]. Glenoid rim fracture, the so-called postage stamp fracture owing to its serrated appearance, has also been described [2, 13]. In an analysis of four cases of glenoid rim fracture, Fitsch et al. recommended that suture anchors be inserted at varying angles of medial-lateral and superior-inferior inclination so as to avoid narrow bone bridges oriented in a linear fashion. These authors speculated that such an anchor configuration created a zone of weakness that predisposed the anteroinferior glenoid to fracture. The size of the suture anchor may also play a role in the occurrence of glenoid rim fracture, and it is advisable to use the smallest diameter anchor possible.

Failure of arthroscopic stabilization techniques resulting in recurrent instability, which has been reported to occur at rates ranging from 4 to 19%, can be attributed to a number of factors, including patient selection, failure to identify and treat associated pathology, and/or poor surgical technique [1, 3, 11, 45]. An accurate assessment of glenoid bone loss is critical prior to undertaking an arthroscopic treatment approach as significant bone loss, commonly recognized as greater than 25% of the inferior glenoid diameter, has been associated with higher rates of recurrence [4, 52]. Additionally, the failure to diagnosis and treat lesions oftentimes found in conjunction with Bankart tears—anterolateral labroligamentous periosteal sleeve avulsion (ALPSA), humeral avulsion of the glenohumeral ligaments (HAGL), Hill-Sachs lesions—jeopardizes the durability of any repair and diminishes the likelihood of a favorable clinical outcome. In particular, an emerging body of literature has validated the role of the engaging (“off-track”) Hill-Sachs lesion on clinical outcomes [8]. Shaha et al. evaluated the results of 57 shoulders treated with arthroscopic Bankart repair and observed that 4 out of 49 (8%) patients with “on-track” Hill-Sachs lesions were deemed treatment failures versus 6 out of 8 (75%) patients with “off-track” Hill-Sachs lesions that failed treatment (P < 0.0001). Furthermore, these authors determined the positive predictive value for failure of “off-track” to be 75% [51]. Similarly, Locher et al. recently reported on the incidence and association of “off-track” Hill-Sachs lesions in the setting of recurrent instability among 100 patients treated with arthroscopic stabilization. Of the 100 patients, 88 were found to have “on-track” Hill-Sachs lesions and 12 had “off-track” Hill-Sachs lesions. Five patients (6%) with “on-track” Hill-Sachs lesions required revision surgery, while four patients (33%) with “off-track” Hill-Sachs lesions (odds ratio, 8.3, P = 0.006) [33].