Subscapularis repair in reverse shoulder arthroplasty: a systematic review and meta-analysis with a sub-analysis by...

The debate about repairing the subscapularis in the setting of reverse shoulder arthroplasty (rTSA) is ongoing. Some reported benefits of the subscapularis repair are a reduced rate of instability, more internal rotation (IR) improvement, and better functional outcomes. However, with the popularization of lateralized implants, these benefits are becoming more limited. Therefore, this meta-analysis compared outcomes of rTSA depending on whether or not the subscapularis was repaired with a sub-analysis by rTSA design (lateralized vs. medialized) and subscapularis repair technique.

PubMed, Cochrane, and Google Scholar were queried through January 2026. Inclusion criteria consisted of studies comparing the outcomes of rTSA based on whether the subscapularis was repaired or not. Extracted data included adverse events, improvement in patient reported outcome measures, and improvement in range of motion. A sub-analysis was done by whether or not the rTSA was lateralized or medialized, and by subscapularis repair techniques (trans-osseous (TO), tendon to tendon (TT), or both).

Nineteen articles based on 18 cohorts met the inclusion criteria with 1,698 patients in the subscapularis repair group (R) and 1,284 in the nonrepair group (NR). Patients in the R group had a lower rate of instability (odds ratio = 0.26, P < .001), better pain improvement (mean difference = 0.47, P = .01), and better IR improvement (standardized mean difference = 0.17, P = .04). When sub-analyzing by rTSA design the R group in the medialized rTSA subgroup had a lower rate of complications (odds ratio = 0.10, P < .001) and instability (odds ratio = 0.06, P < .001), as well as better American Shoulder and Elbow Surgeons improvement (mean difference = 7.01, P < .001) compared to the NR group. The R group had better pain improvement in both the medialized and lateralized subgroups. Of note, there was no difference in IR improvement between the 2 groups in the lateralized subgroup. When sub-analyzing by subscapularis repair technique, the R group had a lower rate of complications and instability (odds ratio = 0.06, P < .001, both) in the TT + TO subgroup, and better American Shoulder and Elbow Surgeons (mean difference = 5.22, P < .001), and pain improvement (mean difference = 0.62, P < .001) in the TO subgroup, and better IR improvement (standardized mean difference = 0.52, P < .001) in the TT subgroup.

While subscapularis repair is shown to be beneficial in medialized rTSA implants, minimal benefits are reported in lateralized rTSA implants. However, repair techniques were shown to affect outcomes with TO repair resulting in better patient reported outcome measures improvement, and TT in better IR improvement. Future studies are needed to assess different tendon repair techniques in the setting of lateralized rTSA implants.