1 Introduction

Nonagenarians, individuals aged 90 years and older, constitute a burgeoning demographic within the United States (U.S.) By the year 2050, it is anticipated that this group will contain an astounding 71 million individuals worldwide, indicating a rapidly growing population. As this group ages, they encounter a number of medical issues, including the imperative need for restoration of shoulder function.¹ Procedures such as reverse shoulder arthroplasty (RSA) have increasingly grown in popularity in the past decade for their effectiveness in pain relief and function-restoring treatment.^1–4 In the U.S., there are more than 60,000 RSAs are performed annually.⁵

It is crucial to note with the rapidly expanding population of nonagenarians in the U.S., this may translate to a higher RSA burden in this population. However, with any surgical intervention, there are inherent risks and potential perioperative complications. Previous studies have reported RSA can lead to surgical site complications and medical complications including death, deep vein thrombosis (DVT), acute coronary syndrome, stroke etc.^6–8 However, the current research on hospital characteristics and perioperative complications after RSA lacks information specific to nonagenarians. Given the higher rates of complications in nonagenarians undergoing total hip arthroplasty (THA), our study aimed to examine complications and associated costs in nonagenarians after RSA.⁹

We hypothesize that nonagenarians will be subject to longer hospitalization, higher care-associated costs, and higher rates of unfavorable perioperative outcomes. The purpose of our study was to analyze the Nationwide Inpatient Sample (NIS) Database to further examine the patient characteristics, demographics, and incidence of postoperative complications in patients aged 90 years or older who underwent RSA.

2 Material and methods

3 Results

The NIS database identified a total of 59,925 patients who underwent RSA between 2016 and 2019. Those patients were categorized into nonagenarians and non-nonagenarians (controls). Nonagenarians were more likely to undergo non-elective surgery had greater hospital length of stay and total charges when compared to controls. Additionally, the control group patients had significantly greater proportions of diabetes without complications and tobacco related disorders in comparison to nonagenarian patients (Table 1).

There were significantly greater proportions of females undergoing RSA in comparison to males in both the nonagenarian and the control group. The control group also had greater proportions of patients with obesity. In each group, patients were stratified by ethnicity including Caucasian, African American, Hispanic, Asian or Pacific Islander, and other, where many of our patients identified as Caucasian (Table 2).

Comparing nonagenarians and control group patients, nonagenarians were more likely to die during hospitalization and were significantly more prone to pneumonia, blood loss anemia, acute renal failure, and likely to require blood transfusions in comparison to control group patients. We found no significant differences between nonagenarians and controls in terms of periprosthetic mechanical complications, periprosthetic dislocations, periprosthetic fractures, surgical site infections, deep vein thrombosis, pulmonary embolisms, or myocardial infarction (MI) (Table 3).

In patients undergoing elective RSA, nonagenarians tended to have longer lengths of stay compared to controls and were more frequently prone to blood loss anemia, PNA and require blood transfusion. In contrast, nonagenarians receiving elective RSA had no significant differences in total charges, wound dehiscence, periprosthetic infections, mechanical complications and fractures, DVT, PE, SSI, MI, and ARF (Table A.2).

In patients undergoing non-elective RSA, nonagenarians had significantly longer lengths of stay compared to controls and more frequently experienced blood loss anemia, required blood transfusion, suffered from ARF, as well as died during hospitalization. We report no other significant differences in complications or total charges for non-elective RSA in nonagenarians and controls (Table A.3).

Nonagenarians who underwent elective RSA had significantly shorter lengths of stay, fewer total charges, lower mortality during hospitalization, less frequent blood loss anemia and blood transfusion, and were less prone to suffering from ARF compared to nonagenarians who underwent non-elective RSA (Table A.4).

4 Discussion

The main findings of this study are that nonagenarians undergoing RSA were subject to longer LOS, greater COC, and greater complication rates compared to their younger counterparts. Nonagenarians are also more likely to require non-elective RSA, which is associated with greater LOS, COC, and complications.

The longer LOS post-RSA in nonagenarians is consistent with previous studies on octogenarians (patients aged 80–89 years old).^7,12 Despite progress in managing geriatric patients, there is still room to improve hospital stays for older patients. Moreover, both increased age and decreased functional status have been linked to an increased postoperative LOS.¹³ Furthermore, caregiver stress and nursing home placements are potential targets to mitigate the length of stay of older patients, address the rising trends in healthcare costs, and minimize the fiscal burden on our healthcare system.¹⁴ The longer LOS associated with post-RSA in the nonagenarian group can be associated with a higher COC. Nonagenarians have been reported to have higher in-hospital costs than younger populations in procedures such as total joint arthroplasty and THA.¹⁵ In addition, RSA is widely known as an expensive procedure, as implants and staffing costs have only driven prices up.¹⁶ However, RSA is also known as one of the most cost-effective procedures, specifically for elderly populations over 65 years old.^17,18 Making implants more affordable and finding effective ways to coordinate staffing could be a target source to reduce the cost associated with RSA.

Given that nonagenarians were less likely to undergo elective RSA compared to controls, it is important to note that longer lengths of stay and costs of care post RSA could be associated with worse outcomes reported with RSA post fracture.^19,20 Although the use of RSA in patients suffering from fractures was previously a debated topic, poorer outcomes for post-fracture RSA when compared to elective RSA have been reported.²¹

Nonagenarians were at a significantly greater risk of mortality compared to our control group. Limited data on nonagenarians aligns with previous studies on octogenarians, which found higher inpatient mortality rates in older patients.^4,6,7,12 Despite the potential of a successful surgery, the mortality risk of undergoing RSA should be communicated effectively to older patients and their families.

Although we found elective RSA in nonagenarians to be associated with longer LOS, more frequent blood loss anemia, transfusions, and PNA, we feel those risks may be deemed acceptable given the great therapeutic and functional benefits of RSA. Our analyses also discovered that nonagenarians undergoing elective RSA are subject to more favorable outcomes, lower COC, and fewer complications than RSA in their non-elective counterparts, illustrating potential room for optimization in the management of patients needing emergency RSA. To our knowledge, this is the first study that has evaluated the outcomes of nonagenarian patients who undergo elective RSA using a large database.

Nonagenarians who underwent RSA were more commonly Caucasian women, consistent with previous reports of higher proportions of female patients undergoing shoulder arthroplasty in this age range.^7,22 Unfortunately, both female gender and increased age have been linked to greater risks of intraoperative complications.²³ The complication rates differ slightly from previously reported ones in octogenarians. We found nonagenarians to be at increased risk of pneumonia, ARF, blood loss anemia, and required blood transfusions (no significant difference in pneumonia and acute renal failure rates in octogenarians vs. control in Newman et al.).⁷ Remarkably, our study revealed that nonagenarians undergoing RSA did not experience any significant increase in incidences of periprosthetic mechanical complications, periprosthetic dislocations, periprosthetic infections, or periprosthetic fractures. This finding diverges from previous studies that have examined this population after procedures like THA and total knee arthroplasty.^24,25,26,27

Our results may also highlight the health disparity in elderly patients who are offered primary RSA. Interestingly, the racial and ethnic makeup of all patients in our study undergoing RSA was significantly different from the 2020 Census Bureau population estimates of the racial and ethnic composition of the U.S. (Caucasians 61.9%, Hispanic/Latin 18.7%, African-American 12.4%, Asian 6%, Native American 1.1%, Pacific Islander 0.2%).²⁸ This disparity may be explained by the differences in bone mass density (BMD) among various racial groups; however, we did not stratify the nonagenarian population by race and evaluate the individual characteristics of each due to the small sample size. It is of note that previous studies have reported that black individuals have a higher BMD than white patients, which explains higher rates of fractures within this population.²⁹ While the racial disparity in RSA among our cohorts was not the main topic of discussion, this disparity sheds light on an important topic that requires further research.

The strength of our study lies in the large sample size provided by the NIS database to explore the outcomes of RSA in nonagenarians with greater reliability of statistical analyses, while also simultaneously assessing for multiple medical and orthopedic complications. This allows for the exploration of rare complications across hospitals in the U.S. that may not be available in individual institutions. To our knowledge, complications in nonagenarians following RSA has never been studied before. We hope that this study will aid clinicians make informed decisions regarding patient care and resource utilization surrounding the use of RSA in older patients to deliver the most appropriate treatment option. Additionally, we believe this information will be important for necessary discussions with patients and their families that often revolve around risks, benefits, and costs.

We recognize that our study has several limitations. Unfortunately, only administrative information regarding a patient's initial original hospitalization is kept on file in the NIS database, allowing for varying degrees of dependability. Long-term outcome data following patient disposal is crucial in arthroplasty, but it is not available in the NIS database. Consequently, our study, while providing valuable insight into perioperative outcomes and complications of RSA in nonagenarians, has a limited follow-up window and may underestimate long-term risks associated with RSA in this age group. Further research incorporating longer follow up timeframes is required to attain a better understanding of the long-term outcomes associated with RSA in older patients. Lastly, despite its reasonable precision (specificity >92%), information coded in NIS may also be incomplete, potentially hindering our study's power and making it prone to bias.

5 Conclusions

Nonagenarian patients had significantly higher orthopedic and medical complication rates than younger patients. Additionally, nonagenarians also incurred greater COC and LOS, consistent with lower proportions of elective admissions in nonagenarians. This information is useful to aid providers and patients make informed decisions while taking into account factors that are important for both parties, such as patient care, resource utilization, and costs. More efforts should be made to investigate and combat disparities that continue to persist in the U.S. healthcare system.

Funding/scholarship

This research did not receive any specific grant from funding agencies in the public, commercial or not-for-profit sectors.

Institutional Review Boards

IRB exempt since the study uses commercially available database with deidentified patient information.

Author contributions

Michel Toutoungy: Conceptualization, Writing – Original Draft, Writing – Review & Editing, Visualization; Nikit Venishetty: Writing – Original Draft, Writing – Review & Editing Varatharaj Mounasamy: Conceptualization, Writing – Review & Editing, Supervision Michael Khazzam: Conceptualization, Writing – Review & Editing, Supervision Senthil Sambandam: Conceptualization, Methodology, Software, Validation, Formal analysis, Investigation, Resources, Data Curation, Writing – Review & Editing, Supervision, Project Administration.