Abstract
Objectives:
Based on findings from our multi-center, double-blinded randomized controlled trial comparing arthroscopic Bankart repair with/without remplissage for the treatment of traumatic recurrent shoulder instability there is a greater risk of postoperative recurrent instability in patients without a remplissage at two years follow-up. The objective of the current study is to determine whether our short-term findings change with medium-term follow-up.
Methods:
This was a double-blinded, randomized clinical trial with two 1:1 parallel groups conducted at two sites. Patients were randomized intra-operatively to an arthroscopic Bankart with remplissage (REMP) or isolated Bankart repair (NO REMP) between 2011 and 2017. Inclusion criteria were patients 14 years or older diagnosed with traumatic anterior shoulder instability with a glenoid defect >15% and the presence of a Hill-Sachs defect (of any size). Patients were contacted by telephone (Spring 2020) and asked standardized questions to determine whether any additional information regarding subluxations, dislocation or additional surgery occurred since their two-year follow-up. “Recurrent instability” was defined as patient-reported dislocation, or two or more episodes of subluxation. Kaplan-Meier survival analyses were performed to assess survival distribution between groups. For the time factor in the survival analysis, the number of months from time of surgery to outcome (either failure or no failure) was based on the time of the long-term follow-up phone call, or from the time of last reported outcome based on clinical or study follow-ups, whichever was the greatest. These were conducted for: 1) recurrent instability and 2) dislocations. Odds ratios with 95% confidence intervals were also calculated.
Results:
Study groups were similar at baseline with respect to age, gender and BMI. Fifty-four patients were randomized to each study group at the time of the original study. , with 52 REMP and 50 NO REMP included in the analyses up to 24-months post-operative. Of those, 36 from each group were available for mid-term follow-up. Mean months (mean) from surgery to last follow-up was 53.8 for REMP and 49.3 for NO REMP. The rate of postoperative recurrent instability at medium term follow-up was 10% (5/52) in the REMP group at an average of 24 months and 30% (15/50) in NO REMP at an average of 19.5 months (p=0.010). The odds ratio of recurrent instability in the NO REMP group relative to the REMP group was 4.029 (1.337-12.135; p=0.010) and the survival curve was significantly different favouring REMP (χ2 = 6.958, P=0.008; Figure 1). With respect to dislocations only, the odds ratio in the NO REMP group relative to REMP was 3.385 (0.999-11.463; p=0.041) and the survival curve was also significantly different favouring the REMP (χ2 = 4.412, p=0.036; Figure 2).
Conclusions:
At medium-term follow-up, patients undergoing a Bankart repair with remplissage have a better rate of survival than those with an isolated Bankart repair.
Survival curve for patients undergoing Bankart and remplissage or bankart only comparing time to recurrence of instability (dislocation and/or recurrent subluxation).
Survival curve for patient undergoing Bankart and remplissage or Bankart only comparing time to dislocation.