Outcomes of reverse shoulder arthroplasty in patients with previous rotator cuff repair: a systematic review and meta-analysis

BACKGROUND: Rotator cuff tears are a common injury encountered by orthopedic surgeons. Reverse shoulder arthroplasty (RSA) has become a treatment option for those with acute tears, as well as those with prior failed rotator cuff repair (RCR). The objective of this study was to determine if there are differences in postoperative outcomes for patients with previous RCR compared to those undergoing primary RSA for rotator cuff tears. METHODS: A comprehensive literature search was conducted using PubMed, EMBASE, Scopus, and Cochrane Library databases. All relevant studies were reviewed by two authors according to inclusion and exclusion criteria defined in the study. Demographics and postoperative outcomes including functional scores, range of motion, pain scores, and complications were recorded. RESULTS: Six level III articles encompassing 2176 shoulders were included in the analysis (846 with previous RCR and 1330 without). Average postoperative American Shoulder and Elbow Surgeons scores were 75.41 in the prior RCR group and 81.61 in the group without prior repair (mean difference [MD] = −5.95, P ≤ .0001 I(2) = 36). Average postoperative Simple Shoulder Test scores were 8.40 in the RCR group and 9.57 in the group without prior RCR (MD = −0.93, P ≤ .0001, I(2) = 66). Average postoperative Constant Score was 62.97 in the RCR group and 66.63 in the group without prior RCR (MD = −5.02, P ≤ .0001, I(2) ≤ 0.001). Average postoperative University of California at Los Angeles Shoulder Scale score was 28.27 in the RCR group and 29.13 in the group without prior RCR (MD = −1.75, P ≤ .001, I(2) ≤ 0.001). Average postoperative visual analog scale pain score was 1.77 in the RCR group and 0.95 in the group without prior RCR (MD = 0.75, P ≤ .0001, I(2) = 35). CONCLUSION: Patients with prior RCR undergoing RSA have worse postoperative functional scores and pain scores than those without prior RCR. However, these differences are below the minimal clinically important difference for each outcome.