Classifications in Brief: The Instability Severity Index Score for Predicting Recurrent Shoulder Instability After Arthroscopic Bankart Repair

History

Since its emergence in the 1990s, arthroscopic Bankart repair has been one of the most widely used surgical procedures to address shoulder instability [7]. However, recurrent instability is common, with studies reporting that as many as 89% of the patients who undergo the procedure will experience a redislocation [9, 30, 48]. As such, in the 2000s, the focus shifted toward identifying patients at high risk for dislocation who may be more appropriately treated with bone block augmentation procedures such as the open Latarjet, which is known to result in decreased risk of recurrent instability [3, 25, 31, 53], ranging from 2% to 14% [1, 10, 13, 15, 23, 33].

In 2007, Balg and Boileau [2] created the Instability Severity Index (ISI) score to identify patients at high risk of recurrent instability after an isolated arthroscopic Bankart procedure and who therefore might be better served by an open operation. In their study of 131 patients treated with arthroscopic Bankart repair between 1999 and 2002, 14.5% (19 of 131) of patients experienced recurrent instability at a mean follow-up interval of 31 months. Several risk factors for recurrent instability were identified and used to develop the scoring system. These risk factors included: age less than 20 years at the time of surgery, sports involving contact or forced overhead activity, participation in competitive sports, shoulder hyperlaxity, a Hill-Sachs lesion visible on external rotation AP radiograph, and glenoid loss of contour on AP radiograph.

Purpose

The ISI score is a preoperative risk-stratification tool used to estimate the risk of recurrent shoulder instability after arthroscopic Bankart repair [2]. The score was meant to be easily obtainable in the outpatient office setting, using information that is available at the first visit, including a preoperative questionnaire, physical examination, as well as simple, plain radiographs [6]. Surgeons have used the ISI measure to guide the surgical management of anterior shoulder instability [6, 43]. This score has been used to try to improve communication among physicians and has been widely adopted in clinical research. However, because this tool has shown inconsistent predictive ability in the clinical setting, it should be used only with caution.

Description

The ISI scoring system combines six preoperative factors into a 10-point score [51]. The higher the score, the higher the risk of recurrent instability. Two points are scored for age younger than 20 years or for participation in competitive sports. One point is scored for involvement in contact or forced overhead activities, or if shoulder hyperlaxity is present on physical examination. Forced overhead activities refer to sports that include overhead hitting movements or contact sports with a high risk of falls, such as basketball, handball, volleyball, tennis, soccer, downhill skiing, rugby, judo, and karate [12]. Anterior hyperlaxity is defined as external rotation greater than 85° with the arm at the side, whereas inferior hyperlaxity is defined as positive hyperabduction test result [22] in which a side-to-side difference greater than 20° is present. On an external-rotation AP radiograph, two points are scored if a Hill-Sachs lesion is visible or if there is a loss of the normal inferior glenoid contour. After applying this score to their study population, Balg and Boileau [2] found that the mean score for those with recurrence was 5.3, and the mean score for those without recurrence was 2.7. A score of 3 or less was associated with recurrence in 5% of patients, whereas a score of 6 or less was associated with recurrence in 10% of patients. However, when the score was greater than 6, the risk of recurrent instability after arthroscopic stabilization rose to 70%. On this basis, Balg and Boileau [2] advocated that a score of 7 through 10 may serve as an indication for performing an open Latarjet procedure rather than an arthroscopic Bankart repair; however, this is unlikely to be universally agreed upon. According to the original study [2], a stepwise approach was used to determine the best scoring system, but it is unclear as to why a score out of 10 was chosen and why certain factors had more relative weight than others.

Validation

Multiple studies from diverse geographic regions have sought to determine the clinical utility of the ISI scoring system with inconsistent results [8, 14, 16, 35, 40, 44, 45, 47]. Phadnis et al. [40] were among the first to validate the reproducibility of the ISI score. This retrospective study compared patients with recurrent dislocation after arthroscopic stabilization (13.5% [19 of 141] of patients) with those who had no further documented episodes of instability at a mean follow-up of 47 months. The ISI score was applied retrospectively, and a receiver operator characteristic curve was constructed to set a threshold ISI value for considering alternative surgery. Similar to the findings of Balg and Boileau [2], the ISI score in the recurrent dislocation group was higher than that of the group without recurrence (5.1 versus 1.7; p < 0.001). However, a lower ISI threshold for the open Latarjet procedure was found; a 70% risk of recurrent dislocation failure was predicted by an ISI score of 4 or more, as opposed to a 4% risk with an ISI score of less than 4.

The largest series validating the ISI score included 670 patients with a minimum follow-up of 5 years and was performed in Italy [35]. Patients with an ISI score of 4 to 6 had an increased risk of recurrence compared with those with a score of 3 or less (HR = 2.4 [95% confidence interval (CI) 1.4 to 4.3]; p = 0.002], whereas those with an ISI score higher than 6 had an even greater risk (HR = 9.4 [95% CI 5.2 to 17.7]; p < 0.001. Additionally, although 94% of patients with an ISI score of at least 3 remained recurrence-free at 5 years, this percentage dropped to 86% in those with an ISI score of 4 to 6, and to 55% in those with an ISI score of greater than 6. The authors ultimately concluded that arthroscopic stabilization could be considered for patients with an ISI score of 3 or less.

A multicenter study performed in Canada and Switzerland assessed the interobserver reliability of the ISI score in 114 patients with traumatic anterior shoulder instability awaiting shoulder stabilization procedures [44]. In this study, five independent evaluators were asked to score the severity of patients’ instability using the ISI tool. The authors reported an intraclass correlation coefficient of 0.933, indicating the ISI score has high reliability among different observers.

However, other studies have demonstrated less reliable results using the ISI score, particularly in active populations with low baseline risk of recurrent instability [14, 16]. In a study of 217 military service members in the United States [16], there was no difference in ISI score between patients with and without recurrent dislocations after arthroscopic stabilization at a mean follow-up of 42 months. Although young age and participation in contact sports were associated with a higher risk of recurrent dislocation, these associations were not observed with any of the other ISI parameters. Rather, the authors found that parameters not included in the ISI score, such as duration of instability symptoms (> 3 months), percentage of glenoid bone loss (> 14.5%), and Hill-Sachs volume (> 1.3 cm³), served as risk factors for recurrent instability. However, patients with severe bone loss and hyperlaxity were excluded from their analysis, which may explain why the mean ISI score was only 3.6 among study participants, and why the scores did not differ between the two study groups. Similar findings were reported in a separate study of 131 military service members [14]. At a minimum 2-year follow-up interval, recurrent dislocation was observed in 26% (34 of 131) of the cohort after arthroscopic Bankart repair, and no difference in ISI score was found between patients with and without recurrence. Furthermore, no individual ISI domains were independently associated with subsequent recurrent dislocation or revision stabilization. A recent study from Spain reported that the ISI score was unable to anticipate the likelihood of recurrence after arthroscopic Bankart repair in patients with scores of 6 or less [45]. Within this study, 14% (20 of 142) of patients experienced recurrence, and the mean preoperative ISI score was 1.8 in both patients with and without recurrence. These results suggest that the ISI may not be a valid tool to estimate the risk of recurrent anterior instability (subluxation or dislocation) in populations with low preoperative risk of recurrent instability.

Overall, the evidence demonstrates inconsistent associations between ISI scores and recurrent instability after arthroscopic Bankart repair. As such, this clinical tool should not be used in isolation to estimate the risk of recurrent instability after arthroscopic Bankart repair, as other important parameters associated with redislocation risk are not considered in the scoring system. In addition, there appears to be no consensus as to the specific patient population to which this scoring system should be applied.

Limitations

One weakness of this scoring system is the lack of advanced imaging [20]. Although routine radiographs are easier to obtain, they are less sensitive and specific for evaluating glenoid and humeral bone loss than CT or MRI, and they do not quantify the size and specific location of bony defects [5, 27, 29, 42]. This is an important limitation of the ISI scoring system as the extent of glenoid and humeral bone loss is considered one of the key determinants of redislocation after arthroscopic Bankart repair [9-11, 18, 19, 24, 26-28, 46]. In addition, the ISI score may be limited by evaluation bias due to the inherent variability of reproducing patient positioning while obtaining routine radiographs.

There have been recent efforts to include advanced imaging into the ISI score. In 2020, the glenoid track instability management score was introduced to incorporate the glenoid track concept into the ISI [20]. The glenoid track instability management score uses 3D CT as the sole imaging modality to assess on-track (0 points) versus off-track (4 points) Hill-Sachs lesions while maintaining the other scoring parameters of the ISI. In a recent evaluation of these scoring systems in 261 patients, Di Giacomo et al. [20] demonstrated that the glenoid track instability management score reduced the usage of the Latarjet procedure compared with the ISI score. Using a cutoff score of 4 to indicate a Latarjet procedure, 69% of patients in the ISI group would have received a Latarjet procedure compared with only 14% in the glenoid track instability management score group. Additionally, similar improvements in postoperative Western Ontario Shoulder Instability Index and Single Assessment Numerical Evaluation scores were achieved between groups. The authors concluded that advanced imaging should be incorporated into the ISI score to reduce potential overuse of the Latarjet procedure. Aside from the quantification of bipolar bone loss (combined bone defects of the glenoid and humeral head), one must consider additional risk factors for recurrent anterior instability that are not included within the ISI score, such as the duration of instability symptoms [16, 17], the number of previous dislocation events [17, 32, 52], or history of prior shoulder stabilization surgery [21, 37].

Another limitation of the ISI scoring system pertains to the treatment of those with an intermediate risk of recurrent instability; that is, a score between 4 and 6. Currently, there is no widely accepted treatment standard for this group, which might include arthroscopic Bankart repair with or without capsular plication or remplissage, open Bankart repair with or without capsular shift, the Latarjet procedure, or free bone block reconstruction procedures with the use of autograft or allograft [4, 34, 36, 39, 41, 49]. In fact, several recent studies have shown that the addition of the remplissage procedure to an arthroscopic Bankart repair results in equivalent functional outcomes and a similar risk of revision surgery compared with the Latarjet procedure in patients with off-track Hill-Sachs lesions [24, 38, 50]. On the contrary, the treatment of those with a low (ISI score ≤ 3) or high (ISI score > 6) risk of recurrence is more universally accepted given the relative consistency of the data [2, 35, 40].

Conclusion

The ISI score was originally developed at a time when advanced imaging was not consistently performed. Despite its wide adoption in research, this tool should not be used in isolation in the clinical setting to estimate the risk of recurrent instability after arthroscopic Bankart repair, but instead, it should complement other important sources of information derived from advanced imaging and detailed patient history. In our experience, routine collection of the independent variables considered in the ISI score among all patients presenting with anterior shoulder instability is important for counseling and to frame treatment discussions and expectations, but we do not use the combined score given its inconsistent predictive ability and varying cutoff values.