Rheumatoid arthritis (RA) is a systemic autoimmune condition characterized by widespread joint inflammation, synovitis, and other systemic symptoms [1,2]. While medical management of RA has improved recently with the development of disease-modifying anti-rheumatic drugs (DMARDs), uncontrolled RA remains relevant to shoulder surgeons due to its potential to lead to erosive arthritis of the glenohumeral joint [3]. In addition, the use of corticosteroids in the treatment of RA may lead to avascular necrosis of the humeral head, necessitating treatment [3]. The treatment options for RA patients experiencing RA-related glenohumeral arthritis are largely limited to either anatomic total shoulder arthroplasty (aTSA) or reverse shoulder arthroplasty (RSA). However, performing TSA in patients with RA entails a greater complication profile, such as a heightened risk for infection from immunosuppressive medications [4,5], compromised rotator cuff integrity [1,6], and impaired bone quality from the systemic inflammation and corticosteroid use, making these patients more susceptible to acromial stress fractures [7,8], intraoperative fractures [7], and the sequelae of poor implant fixation [1].

Despite these challenges, a multicenter study by Lévigne et al. [9] demonstrated that RSA is a durable and effective treatment option for RA-related destruction of the glenohumeral joint, with a survivorship of 96% at 7-year follow-up. Conversely, in a systematic review of 279 shoulders, Haleem et al. [1] showed an expectedly lower revision-free survivorship of aTSA in RA patients, 87% at less than 7 years of follow-up. While numerous studies have reported on the outcomes of shoulder arthroplasty in patients with RA, there is a paucity of large-cohort studies that assess the risks of RA in comparison to primary osteoarthritis (OA). Understanding the comprehensive risk profile of RA patients is important for informed decision-making when considering surgical intervention. As such, the aim of this study was to compare both systemic and joint-related postoperative complications in patients undergoing primary TSA with RA versus those with primary OA.

Although several studies have reported on the outcomes of shoulder arthroplasty in patients with RA, there is a paucity of large-cohort studies evaluating the risks relative to patients with primary OA. In this study, we compared both systemic and joint-related complications following TSA in 1 to 1 matched cohorts of patients with RA and OA using a large, up-to-date national database. Within 90 days following TSA, RA patients had significantly higher risk of total complications, wound dehiscence, deep SSI, MI, pneumonia, acute renal failure, UTI, readmission, and mortality. Additionally, RA patients exhibited an increased likelihood of a blood transfusion within 72 hours following surgery. Furthermore, within 5 years following aTSA or RSA, RA patients showed a significantly higher risk of prosthetic dislocation and PJI, and a significantly higher risk of scapular fractures was observed in patients treated with RSA.

Our study found a greater incidence of systemic complications in RA patients compared to OA patients in the first 90 days following TSA. As a systemic autoimmune disease, RA causes damage not only to joints, but also to other tissues and organs [11]. The incidence of cardiovascular disease in RA patients ranges from 30% to 60%, and such disease is the most common cause of death in this population [11,12]. Our study found RA patients to have a two-fold greater incidence of MI and mortality within 90 days after surgery compared to OA patients. This contrasts reported outcomes in the total hip/total knee arthroplasty (THA/TKA) literature, in which no significant difference in MI and mortality rates was found between RA and OA patients [13,14]. In addition, a significantly higher incidence of pneumonia was found in the RA cohort. Respiratory complications can occur in 30%–40% of RA patients and is the second leading cause of death [11,12].

Significantly greater postoperative rates of wound complications including dehiscence, deep SSIs, and readmissions were found in RA patients, which coincides with outcomes reported in the THA/TKA literature [13-17]. The greater rate of aseptic wound complications exhibited in the RA cohort may be attributed to the abundant presence of tumor necrosis factor-alpha, which can inhibit granulation tissue formation and disrupt wound healing [18]. Patients with RA were more likely to receive a blood transfusion within 72 hours postoperatively. Anemia is a common comorbidity in patients with RA, prevalent in 33%–60% of patients, and may explain the greater incidence of blood transfusions in RA patients [19]. Though some cases may result from iron deficiency caused by gastrointestinal bleeding secondary to non-steroidal anti-inflammatory drugs use, the anemia observed in RA patients is more likely a consequence of chronic inflammation and increased production of proinflammatory cytokines [19].

Furthermore, RA patients had a significantly higher risk of PJI and prosthetic dislocation within 5 years following surgery for aTSA or RSA. While superficial SSIs were similar in the two cohorts, RA patients had significantly higher rates of deep SSIs and PJIs postoperatively. Nezwek et al. [20] conducted a retrospective review of 902 RSA patients to explore risk factors for infection within 3 months postoperatively. In their study, patients with RA had a 3.5 times greater risk of developing a PJI compared to non-RA patients [20]. Patients with RA may be susceptible to infections due to immunosuppressive drugs, such as steroids and DMARDs. Therefore, physicians should maintain a heightened index of suspicion of postoperative infection in RA patients who present with persistent unexplained shoulder pain postoperatively [20]. The chronic inflammatory effects of RA may compromise the soft tissue around the shoulder as well, which may predispose patients to higher rates of prosthetic dislocation. Within 5 years following RSA, RA patients demonstrated a more than two-fold greater incidence of scapular fractures. This outcome aligns with previous studies that have demonstrated RA to be a strong predictor of postoperative acromial and scapular spine stress fractures [8,21-25]. RA can cause significant bone loss and reduced bone strength, resulting in a higher risk of fracture [26]. However, no significant difference in the incidence of periprosthetic fractures was found. With increased postoperative joint complications in RA patients, the revision rate was slightly increased compared to OA patients following RSA, but this did not reach statistical significance (4.06% vs. 3.12%, P=0.093).

Reverse and anatomic TSA are both indicated in the setting of end-stage glenohumeral joint arthritis secondary to RA. While RSA has become increasingly favored over aTSA in the treatment of RA patients due to concerns over the integrity of the rotator cuff and managing glenoid bone loss, in our study, the RA-aTSA cohort (5,720 patients) was larger than the RA-RSA cohort size (2,332 patients). Nevertheless, both surgical options are effective in restoring function and reducing pain [1,25,27]. Garcia et al. [27] performed a retrospective study comparing clinical outcomes and complications of aTSA and RSA in 86 patients with inflammatory arthritis, with a mean follow-up of 51.6 months. The authors showed that both aTSA and RSA significantly improved patient-reported outcome measures and range of motion; although, aTSA exhibited significantly greater improvement and final postoperative outcome scores than RSA [27]. While the authors demonstrated no significant difference in complication rates between aTSA and RSA in RA patients [27], our study found that patients who underwent RSA had significantly higher risk of prosthetic dislocation, PJI, and scapular fracture. Additionally, we found that patients who underwent RSA had a higher revision rate compared to those who underwent aTSA. However, this contrasts with findings of a systematic review of 279 shoulders by Lévigne et al. [9] showing RSA to be a durable and effective option with a revision-free survivorship of 96% at a 7-year follow-up, which was greater than the 87% revision-free survivorship reported for aTSA [1].

Our study had several limitations. First, the data were sourced from electronic medical record databases of healthcare organizations, originally intended for clinical practice rather than research. Additionally, the dataset was limited to electronic medical records and lacked chart review data. Hence, certain essential health information may have been underreported, and inaccurate diagnoses and coding of clinical events may have occurred. The use of ICD codes may not consistently classify all relevant diagnoses, procedures, or adverse events, which may have affected the cohorts created and the outcome variables studied. In particular, we were unable to control for rotator cuff health in patients undergoing TSA for primary OA. Despite these limitations, this study is the first to assess and compare postoperative complications after TSA in patients with RA to those in patients with OA.

Following TSA, RA patients should be considered at a higher risk compared to patients with primary OA. RA is associated with significantly higher systemic complication rates within 90 days, including wound dehiscence, deep SSI, MI, pneumonia, acute renal failure, urinary tract infection, blood transfusion, readmission, and mortality. Patients with RA also have an elevated 5-year risk of PJIs and prosthetic dislocations following aTSA and RSA, as well as scapula fractures following RSA in comparison to patients with OA. Knowledge of the comprehensive risk profile of RA patients undergoing TSA is essential for appropriate patient counseling and education. Further studies are necessary to enhance our understanding of the increased risks associated with RA and to explore postoperative management following shoulder arthroplasty in this higher-risk patient population.