Opinion editorial—first-time shoulder dislocation: My approach

ABOUT THE EXPERT

Eiji Itoi, MD, PhD, is an emeritus professor of Tohoku University, Sendai, Japan. He had a shoulder fellowship year with Dr. Shiro Tabata in 1989. He then became a research fellow with Kai-Nan An, PhD, and Bernard F. Morrey, MD, at Orthopaedic Biomechanics Laboratory, Mayo Clinic, Rochester, Minnesota, in 1990–93 and a visiting scientist at the same institute in 1997–98. After coming back from the United States, Dr. Itoi became a professor and chair of the Department of Orthopaedic Surgery, Akita University School of Medicine in 2001 and moved to Tohuku University in 2006. He is currently a vice president of Tohoku Rosai Hospital.

First, I would like to comment on my treatment strategy for initial dislocators. There are two pathologic entities: a first-time dislocation and a recurrent dislocation. They look similar, but they are quite different from the therapeutic point of view. The first-time dislocation is an acute injury caused by a significant trauma, whereas the recurrent dislocation, a consequence of a first-time dislocation, is a pathologic status known as chronic instability. Once the initial dislocation has been reduced, we never know whether the patient will have a chronic instability (recurrent dislocation) in the future. We know that the recurrence rate is very high among young males involved in vigorous activities. Surgical stabilization provides the best stability to the shoulder among all the treatment options. This is the reason why some surgeons recommend surgical stabilization after an initial dislocation. However, a routine use of surgical stabilization to all the initial dislocators should be avoided. The number needed to treat for surgical stabilization after an initial dislocation is calculated to be 7.0 for the general population and 2.6 for high-risk patients. This means that if you perform surgical stabilization in all the patients after an initial dislocation, you will end up doing unnecessary surgery in 60% to 85% of them. Robinson et al. stated that there was not sufficient evidence to support the routine use of prophylactic repair after all primary dislocations. The first-time dislocators do not have chronic instability yet. Thus surgery for first-time dislocators is a prevention of chronic instability. Preventive surgery seems to be too invasive. There is always a risk of performing unnecessary surgery. On the other hand, recurrent dislocators already have chronic instability. Surgery for recurrent dislocators is a treatment of chronic instability. There is no unnecessary surgery, and the surgery is 100% warranted. A more recent meta-analysis showed that the number needed to treat ranged between 2.0 and 4.7, which made the rate of unnecessary surgery to be 50% to 79%. The authors warned that prevailing practice of low-threshold surgery after a first-time dislocation could be questioned. They also mentioned that waiting to see if a patient develops chronic instability may save half of patients from unnecessary surgery. Thus my treatment strategy is to wait until a second dislocation occurs unless the patient has a large glenoid bone loss and/or an off-track Hill-Sachs lesion. Then comes a question, “Does a second dislocation make the outcome of stabilization surgery worse?” There is no direct answer to this question. However, in one study the authors compared surgery after a single dislocation versus multiple in-season dislocations. Return to play and recurrence rate were 89% and 11%, respectively, for those who underwent surgery after a single dislocation and 90% and 10% for those who underwent surgery after multiple in-season dislocations. There were no differences between the two groups. The authors did not specify how many dislocations they had during the season. However, the same authors reported in another study that the average number of in-season dislocation or subluxation was 2.2. Another study showed that an average number of in-season dislocation or subluxation was 1.4. Based on these studies, it is likely that these patients who returned to play had one or two additional dislocations or subluxations during the season. This study strongly suggests that waiting for a second dislocation does not make the outcome of surgical stabilization worse. Thus we can wait even in patients with high risks. While waiting, the treatment option may change according to the activity levels of the patients and the timing of injury. If it is an in-season athlete, we prefer to prescribe a protective brace and let the patient return to sport because in-season athletes usually want to return to play as soon as possible. After the season, if the shoulder remains stable, no further treatment is necessary. If the shoulder has been unstable, we can perform surgical stabilization at this point. Using this strategy, we can avoid unnecessary surgeries without worsening the outcome. If it is an off-season athlete or nonathlete, the patient is a good candidate for conservative treatment. I prefer to use immobilization in external rotation if the patient agrees. The brace we currently use is Shoulder Brace-ER (Alcare), which keeps the shoulder in 15 degrees of external rotation ( Fig. 43.1 ). ^, We ask the patient to use this brace 24 hours a day, except when taking a shower or changing clothes, for 3 weeks. The compliance rate of this brace was 72%. There are two sizes: M–L (waist size: 65 to 95 cm) and LL–3L (waist size: 85 to 115 cm). The brace is self-attachable, and the elbow can be extended when lying in bed, which makes it easier to use the brace while sleeping.

The outcome of immobilization in external rotation is still controversial. We reported that immobilization in external rotation resulted in relative risk reduction of 38%. ^, Since then, several randomized controlled trials have reported positive results or negative results. The meta-analyses were also controversial. The most recent meta-analyses have shown that immobilization in external rotation is effective in reducing the recurrence rate. ^, All these studies were based on the immobilization position of external rotation in adduction (i.e., the arm at the side). We shed a new light on this position again. We found that external rotation with the arm in 30 degrees of abduction provided a better reduction of the inferior part of the Bankart lesion compared with external rotation in adduction. Using this new position, a recent study from Germany showed the recurrence rate of 20%, which seems to be quite promising. In Japan, a multicenter prospective randomized controlled trial organized by the Japan Shoulder Society is currently going on, comparing the traditional position of adduction-internal rotation versus a new position of 30 degrees of abduction and 30 degrees of external rotation ( Fig. 43.2 ). The best conservative treatment needs to be determined in terms of position and length of immobilization. Furthermore, we have very limited evidence on rehabilitation programs after immobilization. The rehabilitation should be studied together with immobilization to achieve the best performance of conservative treatment.

The only exception not to wait for a second dislocation is a fracture-dislocation case, in which a glenoid fragment is large enough to be detectable on plain radiograph and/or with an off-track Hill-Sachs lesion. In such cases a computed tomography scan is advised to assess the exact size and location of the glenoid fragment. If the fragment is equal to or greater than 25% of the glenoid width, it should be fixed using a screw. If there is no bony fragment or a fragment too small to be fixed is seen on plain radiograph, we can wait until a second dislocation occurs. It has been reported that the prevalence of an off-track Hill-Sachs lesion is 7% of patients with recurrent anterior dislocations, whereas it is 0% in patients after a first-time dislocation. However, a more recent study reported that 13% of first-time dislocators had an off-track Hill-Sachs lesion, and they all showed recurrent dislocation after conservative treatment. Based on these findings, if you clearly see a large Hill-Sachs lesion on plain radiographs, you would do well to take CT scans to make sure that the Hill-Sachs lesion is on track before starting conservative treatment.

I would like to comment on the recent progress in the on-track/off-track concept as well. Bipolar lesion of the shoulder (glenoid bony defect and Hill-Sachs lesion) is common and observed in 33% to 44% of initial dislocators and 62% to 79% of recurrent dislocators. ^, In our series of recurrent dislocators, the prevalence of bipolar lesion was 81%. We used to say that the critical size of the glenoid defect is 25% of the glenoid width. These studies clarified that the safe defect size was less than 17% and the critical defect size was more than 25% ( Fig. 43.3 ). ^, The size between 17% and 25% remains a gray zone.

Recently, a new concept of “subcritical bone loss” has been introduced by Shaha et al. The authors treated military patients using arthroscopic Bankart repair. At 2-year follow-up, they found that the high failure rate was the problem for those with a bone loss of 20% or more of the glenoid width. However, among those with a bone loss less than 20% of the glenoid width, Western Ontario Shoulder Instability Index (WOSI) score was significantly worse in patients with more than 13.5% bone loss than those with less than 13.5%. They concluded that the zone between 13.5% and 20% be called “subcritical bone loss.” We looked at our data of patients from every walk of life, not the military patients as in the previous study. We found that those with a bone loss ranging between 17% and 25% showed significantly lower WOSI score than those with a bone loss less than 17%. Thus, in our series, the bone loss ranging from 17% to 25% of the glenoid width (see Fig. 43.3 ) seems to be equivalent to the subcritical bone loss reported by Shaha et al. The size of the subcritical bone loss may change according to the risks of the patients. Recently, we found that there was a “subcritical zone” in the glenoid track. Patients with a Hill-Sachs lesion located in the most medial one-quarter of the glenoid track showed significantly lower WOSI scores compared with those with a Hill-Sachs lesion located in the remaining three-quarters of the glenoid track. We called the former a peripheral-track lesion and the latter a central-track lesion ( Figs. 43.4 and 43.5 ).