Acute complications after reverse total shoulder arthroplasty for treatment of cuff arthropathy versus fracture
Article information
Abstract
Background
Reverse total shoulder arthroplasty (RTSA) has become increasingly popular in recent years, and this trend is expected to continue. However, differences in outcomes of RTSA for fractures compared with other indications are poorly understood. This study aimed to identify the compare the incidences of adverse events during RTSA to treat cuff tear arthropathy (CTA) versus RTSA to treat fractures, as well as identify risk factors for any adverse event.
Methods
Patients who underwent RTSA from 2010 to 2020 in the National Surgical Quality Improvement Program database were included. Matched cohorts were created using nearest-neighbor matching. Independent sample t-tests and chi-square tests were used to determine differences between groups, and binary logistic regression was performed to determine odds ratios and 95% CIs.
Results
In total, 27,607 CTA patients (94.5%) and 1,537 fracture patients (5.3%) underwent RTSA. Adverse events occurred in 1,088 CTA patients (3.9%) and 192 fracture patients (12.4%). Incidences of adverse events, postoperative transfusions, and returns to an operating room were all significantly higher in the fracture cohort compared with CTA patients.
Conclusions
Following surgery, higher rates of adverse events were observed in the CTA patients compared with those in the fracture cohort. These findings improve our understanding of the effectiveness of an increasingly popular surgical intervention due to the use of a large database analysis to identify short-term complications and risk factors.
Level of evidence
III.
INTRODUCTION
Reverse total shoulder arthroplasty (RTSA) has become an increasingly popular and useful procedure in recent decades [1]. Although originally designed for treatment of cuff tear arthropathy (CTA), improvements in implant design [2], surgeon technique, and surgeon experience have expanded the indications for RTSA to a breadth of shoulder pathologies [3]. RTSA offers improved range of motion, joint and muscle loading, and overall joint stability, making it an attractive intervention for complex shoulder disorders, including massive rotator cuff tear (MCT) with or without glenohumeral osteoarthritis, primary osteoarthritis, failed primary shoulder arthroplasty, and proximal humeral fracture (PHF) [4-6].
Considerable research has been conducted, with positive outcomes reported following RTSA across various preoperative diagnoses, including CTA, MCT, and PHF. A comparison of results following RTSA versus shoulder hemiarthroplasty for PHF has been associated with “more favorable clinical outcomes” in terms of functional scores, pain relief, and revision rates for patients treated with RTSA [7]. As a result, RTSA has become a leading treatment option for PHF, with utilization rates increasing by 400% between 2005 and 2012 [3,8]. Despite this growing popularity, few attempts have been made to assess the differences in outcomes of RTSA for PHF compared with other preoperative diagnosis.
While the RTSA may yield improved outcomes for pain and function when treating PHF, this preoperative diagnosis may elicit surgical challenges and postoperative complications compared to that of RTSA for CTA [9]. Patients with PHF typically present with unique demographics, preoperative comorbidities, and functional demands that may influence surgical outcomes following RTSA [10]. The primary aim of this study was to determine if there was a difference in the incidence of adverse events (AEs) in RTSA to treat CTA versus a fracture. Second, this study aimed to identify patient risk factors for any AE among and between patients who underwent RTSA for CTA versus fracture.
METHODS
Patient Population
This was a retrospective comparative analysis of data from the National Surgical Quality Improvement Program (NSQIP), which is managed by the American College of Surgeons. NSQIP data were collected by a certified surgical clinical reviewer at each participating site. Participant use data files (PUFs) from 2010 to 2020 were used in this analysis, including those for patients who underwent surgery from January 2010 to December 2020. The NSQIP uses a systematic sampling process to determine which cases are included in the PUF. All patients were followed for 30 days postoperative.
The inclusion criteria for this analysis were any patients who underwent RTSA to treat CTA or fracture. NSQIP criteria for case exclusion included minors (younger than 18 years; cases secondary to urgent, high-acuity situations such polytraumas that may confound quality assessment; and patients who returned to the operating room due to a complication from a prior procedure. The data for this analysis were cleaned in R Studio version 2023.06.0 (R Foundation for Statistical Computing) to exclude patients who had an operation time or body mass index (BMI) less than or equal to 0 minutes (R studio). Cases were also excluded if functional status, dyspnea status, sex, or American Society of Anesthesiologists (ASA) class were not supplied. If the principal anesthesia technique was “none,” “not reported,” or “other,” the case was excluded.
Variables
The independent variable for this analysis was indication for RTSA (CTA or fracture). The dependent variables were AEs following surgery. The outcomes of interest were death, wound dehiscence, sepsis, pulmonary embolism, renal complication, myocardial infarction, cardiac arrest, stroke, transfusion, deep vein thrombosis, urinary tract infection, pneumonia, intubation issues, surgical site infection, and return to the operating room. AEs included all of these complications.
Intervention
Current procedural terminal code 23472 was used to define patients that underwent RTSA. International Classification of Diseases, Tenth Revision (ICD-10) codes were used to define patients who underwent reverse TSA to treat CTA or fracture (Table 1). All patients were followed for at least 30 days following surgery.
International Classification of Diseases, Tenth Revision (ICD-10) codes
Statistical Plan
Matched cohorts were created using the nearest neighbor method to match patients according to age, BMI, sex, race, diabetes, smoking status, ASA class, hypertension requiring medication, congestive heart failure, chronic obstructive pulmonary disease (COPD), and bleeding disorders. IBM SPSS version 28.0.1.1 (IBM Corp.) was used for statistical analyses. Independent sample t-tests were used to determine if there was a difference between the CTA and fracture groups for continuous variables. Fisher’s exact tests were used where possible to determine if there was a difference between indications for surgery for categorical variables. If a Fisher’s exact test was unable to be performed, a chi-square test was used. Binary logistic regression was performed to determine the odds ratios (ORs) and 95% CIs with any AE as the dependent variable and patient demographics and comorbidities as covariates for the combined matched group. Results were statistically significant at a P-value ≤0.05.
RESULTS
Demographics
Between the years 2010 to 2020, a total of 27,607 patients (94.5%) in the study group underwent RTSA for CTA and 1537 patients (5.3%) underwent RTSA for fracture. Patients who underwent RTSA for CTA had a mean age of 69.0±9.3 years and a mean BMI of 31.2±6.8 kg/m2, and 45.7% were male. Patients receiving RTSA for fracture had a mean age of 72.0±9.1 years and a mean BMI of 30.3±7.4 kg/m2, and 18.4% were male. Current smokers accounted for 10.2% of CTA patients and 12.4% of fracture patients. The most common preoperative comorbidities were non-insulin–dependent diabetes (12.6% and 14.2%) and COPD (6.6% and 7.2%) for both the CTA and fracture patients, respectively. Full demographic information is summarized in Table 2.
Demographic and comorbidity characteristics of patients undergoing RTSA for CTA or fracture
Incidence of AEs
AEs within the 30-day postoperative period occurred in 1,088 CTA patients (3.9%) and 192 fracture patients (12.4%). However, after matching by near-equal patient characteristics among the two cohorts, CTA patients had a higher rate of AEs (4.6%) compared with fracture patients (12.4%). Postoperative transfusion (1.6% and 8.7%, respectively) was the most common AE in both cohorts of patients. A return to the operating room was necessary in 1.3% of CTA patients and 3.1% of fracture patients. Incidences of AEs, postoperative transfusions, and returns to the operating room were all significantly higher among fracture patients compared with CTA patients (all P<0.001) (Table 3).
Incidence of adverse events for patients undergoing RTSA for CTA or fracture
Risk Factors for Any AEs
Following RTSA for CTA, increasing age (OR=1.022, P<0.001), BMI (OR=0.988, P=0.015), insulin-dependent diabetes mellitus (OR=1.447, P=0.004), congestive heart failure (OR=1.773, P=0.025), preoperative transfusion (OR=3.069, P=0.006), history of bleeding disorder (OR=1.945, P<0.001), operative time (OR=1.006, P<0.001), length of stay (OR=1.081, P<0.001), functional status (OR=6.212, P<0.001), COPD (OR=1.284, P=0.022), and history of MI (OR=1.394, P<0.001) were all significant risk factors for any AE, while male sex (OR=0.726, P<0.001) was a protective factor.
For patients who underwent RTSA for fracture, age (OR=1.030, P=0.010), preoperative transfusion (OR=3.179, P=0.015), bleeding disorder (OR=2.367, P=0.004), operative time (OR=1.009, P<0.001), length of stay (OR=1.106, P<0.001), and renal failure (OR=12.094, P=0.040) were all significant risk factors for any AE (Table 4).
Odds of developing any adverse event during surgery as related to patient demographics and comorbidities
DISCUSSION
This study aimed to identify any difference between the incidence of AEs (within a 30-day follow-up window) in patients receiving RTSA to treat CTA and those receiving RTSA to treat fractures. After matching, the overall rates of AEs, operative time, length of stay, postoperative transfusions, and return to operating room were significantly higher in patients undergoing RTSA for fractures, with secondary findings that include identification of independent risk factors for any AE for each cohort. A discussion of these results in the context of the literature and relevant surgical considerations is warranted.
Compared with the CTA cohort, the demographics of the patients undergoing RTSA for traumatic fractures, prior to matching, showed dramatically greater rates of female sex, greater average age, higher ASA classifications, dependent functional status, and multiple comorbidities (Table 2). Taken together, these results suggest that RTSA for trauma selects a patient population with lower levels of baseline health at the time of operation. Likewise, compared with the RTSA for CTA cohort, these trauma patients are often limited by time and circumstances that prevent risk-factor modification and preoperative intervention, given the sudden injury and relative immediacy of surgery.
Other baseline differences may be poorer bone quality in the trauma group compared with the CTA group, as evidenced by risk factors such as advanced age and greater percentage of female sex [8]. This association is also supported by the profile of PHF patients treated by RTSA as established in the literature: low-demand, elderly patients with poor bone stock, fracture dislocation, non-constructible tuberosities, or failed conservative management [8]. However, despite age being a known risk factor, a recent study by Stenquist et al. [11] reported no differences in functional outcomes in two cohorts divided by age (mean age, 64 vs. 78 years) post-RTSA for complex PHF with a 3-year minimum follow-up period. As such, future studies are needed to clarify these trends, associations, and long-term outcomes between baseline health, poor bone quality, and implant survival in RTSA for traumatic fractures [12,13].
Regarding rates of AEs in this trauma cohort, this analysis identified risk factors including age, preoperative transfusion, and bleeding disorders, among others (Table 4). Among the many possible factors contributing to these findings, one consideration is the association of advanced age with higher rates of coagulopathies, blood thinners, and traumatic falls [14]. Taken together, these findings suggest that a greater degree of bleeding from soft tissues at the fracture sight and around the time of surgery necessitates transfusions or creates technical difficulties in surgery to navigate intraoperative bleeding or hematomas. These factors could contribute to the increased operative times and rates of return to operating rooms seen in the trauma cohort.
As the unpredictability of acute injuries such as PHF creates an inability to modify certain risk factors, future studies targeting ideal surgical time parameters and patient comorbidity characteristics would be useful. Surgeons would be wise to continue using their best judgment when considering surgical delays for medical optimization efforts in the setting of painful and morbid fractures for patients. Regarding rates of AEs in the CTA cohort, independent risk factors included age, BMI, insulin-dependent diabetes mellitus, CHF, COPD, history of bleeding disorder, increased operating time, length of stay, functional status, and history of myocardial infarction. Notably, male sex was a protective factor for AEs in this cohort. Given that RTSA for arthropathy is typically an elective surgery, there is time for risk-factor modification and medical optimization; a recent study by Churchill et al. [15] outlined guidelines for such an approach in RTSA patients. Surgeons can use these data in conjunction with the existing literature to educate patients and make preoperative plans in pursuit of optimal outcomes and mitigated risks.
This study is limited due to its retrospective design and lack of causative insight. Further inherent limitations of retrospective database studies include selection bias, lack of surgical details, and human error in the NSQIP data-collection methodology [16]. This study cannot control confounding factors such as surgeon experience, skill, and technique [14,17,18]. This study only reports complications in the database within a 30-day follow-up. As such, it does not report on important surgical complications unique to RTSA surgery, such as instability/dislocation, acromial stress fracture, posttraumatic osteoarthritis, hardware failure, and implant loosening, which are either not in the NSQIP or take place over much longer periods of time [19,20]. Other indications for RTSA were not evaluated or compared outside of the arthropathies and fractures classified as we were limited by procedure and indication coding.
CONCLUSIONS
This study used the NSQIP database to compare short-term complications in patients undergoing RTSA for acute traumatic fractures versus chronic CTA indications. Prior to matching, the trauma cohort had higher rates of female patients, older ages, higher ASA scores, greater functional dependency, and higher rates of multiple comorbidities. RTSA for trauma was also associated with longer operative times and hospital stays, higher transfusion rates, and more frequent returns to the operating room. While limited by its retrospective, database-based design and lack of long-term orthopedic outcomes, the findings of this study should help inform surgeons and patients about the short-term risks associated with RTSA for trauma.
Notes
Author contributions
Conceptualization: JS, CON, CE, BE, JLV. Data curation: CTN, IS, HH. Formal analysis: CTN, IS, HH, Investigation: CTN, JS, HH, CON, CE, BE, JLV. Methodology: CTN, HH, CON. Project administration: JLV. Software: HH. Supervision: JS, CON, CE, BE, JLV. Validation: JS, CON, CE, BE. Visualization: CTN, IS. Writing – original draft: CTN, IS. Writing – review & editing: CTN, IS, JS, HH, CON, CE, BE, JLV. All authors read and agreed to the published version of the manuscript.
Conflict of interest
None.
Funding
None.
Data availability
Contact the corresponding author for data availability.
Acknowledgments
None.