Bilateral total shoulder arthroplasty: A systematic review of clinical outcomes

Abstract

Background

The purpose of this study was to describe the clinical outcomes following bilateral total shoulder arthroplasty (TSA).

Methods

A systematic search of the PubMed, Embase, and Cochrane Library databases following PRISMA guidelines was performed. English-language literature published from 2010 to 2018 analyzing bilateral TSA (anatomic and/or reverse) with a minimum one-year follow-up was reviewed by two independent reviewers. Study quality was evaluated with the Modified Coleman Methodology Score and the methodological index for non-randomized studies score.

Results

Eleven studies (1 Level II, 3 Level III, 7 Level IV) with 292 patients were included. Two studies reported on bilateral anatomic TSA (n = 54), six reported on bilateral reverse TSA (RTSA; n = 168), two reported on anatomic TSA with contralateral RTSA (TSA/RTSA; n = 31), and one compared bilateral anatomic TSA (n = 26) and bilateral RTSA (n = 13). Among studies, mean revision rate ranged from 0% to 10.53% and mean complication rate ranged from 4.9% to 31.3%. At final follow-up, patients experienced significant overall improvements in range of motion and patient-reported outcome score measurements. However, bilateral anatomic TSA resulted in greater improvements in external rotation compared to bilateral RTSA. Overall patient satisfaction was 91.0%.

Conclusion

The available data indicate that bilateral TSA allows for functional and pain improvements and result in high patient satisfaction.

Level of evidence

IV.

Introduction

The demand for total shoulder arthroplasty (TSA) in the United States has increased significantly over the past 15 years and is projected to continue to increase over the next 10 years.¹ Anatomic TSA has become a reliable surgical technique to treat patients with severe glenohumeral arthritis. Notably, in the absence of an intact rotator cuff, patients may experience superior migration of the humeral head, resulting in poor functional results, glenoid component loosening, and/or early surgical failures following anatomic TSA.² Reverse TSA (RTSA) has emerged as an alternative treatment option in patients with severe glenohumeral arthritis in the setting of a rotator cuff deficiency, and studies have indicated successful mid-term and long-term results following this procedure.^3–5

Many studies evaluating outcomes following unilateral anatomic TSA and unilateral RTSA have reported improvements in function, pain relief, and patient satisfaction.^3,5–10 With the increasing volume and clinical success of shoulder arthroplasties in the United States, it is common for patients with bilateral disease to require bilateral shoulder arthroplasties. Bilateral TSA involves a patient undergoing a shoulder replacement on one side with a separate and subsequent shoulder replacement on the contralateral side. This differs from simultaneous bilateral TSA where both shoulders are replaced during the same procedure. Although many studies have reported outcomes in patients undergoing bilateral total hip and knee arthroplasties,^11–18 few have reported the clinical outcomes in patients undergoing bilateral TSA.

It has been suggested that bilateral TSA (anatomic and/or reverse) may be inadvisable due to difficulties with activities of daily living (ADLs) in the immediate post-operative period.^4,19,20 In patients undergoing unilateral TSA, they are able to compensate in their ADLs by using their contralateral extremity; however, in those undergoing bilateral TSA, they may not be able to effectively compensate in the same manner while recovering. The purpose of this systematic review was to describe the clinical outcomes following bilateral TSA. The authors hypothesized that patients would experience functional improvements in both shoulders, report high satisfaction rates, and successfully maintain their ADLs following this procedure.

Materials and methods

This systematic review was conducted according to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines using a PRISMA checklist. Two independent reviewers (JS & DAH) searched PubMed, Embase, and the Cochrane Library up to 1 October 2018. The following search phrase was used: “bilateral total shoulder” (arthroplasty OR replacement). A total of 69 studies were reviewed by title and/or abstract to determine eligibility based on the following inclusion criteria: studies reporting clinical outcomes of patients who underwent bilateral TSA, studies with a minimum one-year follow-up from the most recent surgery, studies that were published in English, and retrospective and prospective studies of Level I-IV evidence. Exclusion criteria included studies reporting clinical outcomes of patients who underwent simultaneous bilateral TSA, studies solely reporting clinical outcomes of patients who underwent unilateral TSA, cadaveric or animal studies, non-clinical studies, and case reports. Ten studies and one conference abstract met inclusion and exclusion criteria (Figure 1). Data extraction was performed independently (JS), and when presented with comparative studies, only data from the population that met the aforementioned inclusion and exclusion criteria were extracted. Data from the conference abstract was obtained from the corresponding poster presentation at the American Academy of Orthopaedic Surgeons conference. Funding and third-party involvement were not required to obtain any of the collected data.

Figure 1.

PRISMA 2009 flow diagram. From Moher et al.⁴⁸ For more information, visit www.prisma-statement.org

Reporting outcomes

Outcomes extracted included demographic data (age, follow-up, and interval between arthroplasties (IBA)), revision rate, complication rate, patient satisfaction, range of motion (ROM), and patient-reported outcome scores (PROs). ROM included forward flexion (FF), external rotation (ER) at the side and at 90° abduction, internal rotation (IR) at the side and in abduction, and abduction. PROs included Constant-Murley scores (CMS),²¹ Visual Analog scores for pain (VAS-pain),²² Simple Shoulder Test (SST) scores,²³ The Shoulder and Elbow Surgeons Shoulder (ASES) scores,²⁴ and Subjective Shoulder Value (SSV) scores.²⁵ ROM and PROs were evaluated preoperatively and postoperatively.

Study methodology assessment

The quality of study methodology was evaluated using the Modified Coleman Methodology Score (MCMS) based on a scaled potential score ranging from 0 to 100.²⁶ Scores ranging from 85 to 100 are excellent, 70–84 are good, 55–69 are fair, and less than 55 are poor. Risk of bias was assessed using the methodological index for non-randomizing studies (MINORS)²⁷ score, which incorporates 8 items to assess overall bias and 12 items in comparative studies. The MINORS score has a scaled potential score ranging from 0 to 16 for non-comparative studies and 0 to 24 for comparative studies. Each item is scored 0 (not reported), 1 (reported but inadequately), or 2 (reported adequately).

Statistical analysis

The individual study quality and heterogeneity related to patient population and treatment prevented pooling of outcome data and meta-analyses calculations. Therefore, descriptive statistics are displayed for numerical demographic data (age, follow-up, and IBA), PROs, and ROM. Categorical variables (revision and complication rates, and patient satisfaction) were reported as percentages.

Results

Included studies

Ten studies (1 Level II, 4 Level III, 5 Level IV),^28–37 published between 2010²⁸ and 2018,³⁵ and one conference abstract (Level III)³⁸ met inclusion and exclusion criteria (Figure 1). Of these, two studies^28,29 reported on bilateral anatomic TSA, six studies^30–34,37 reported on bilateral RTSA, and two studies^35,36 reported on bilateral anatomic TSA/RTSA. The poster presented by Cox et al.³⁸ compared outcomes following bilateral anatomic TSA to bilateral RTSA.

Assessment of study quality

Table 1 shows the MCMS from the included studies (mean MCMS, 56.8), two of which (RTSA) achieved good scores, and seven of which (anatomic TSA, 2; RTSA, 4; anatomic TSA/RTSA, 1) achieved fair scores. This indicates that the overall quality of this systematic review was fair. A MCMS for the conference abstract was unable to be determined.

The results of the methodological quality assessment of included studies using the MINORS score are presented in Table 3. The mean MINORS scores for the non-comparative and comparative studies were 10.1 and 17.5, respectively.

Table 3.

Methodological items for non-randomized studies (MINORS) score.

Study	MINORS
Non-comparative studies
Gruson et al.28	9
Fabricant et al.29	8
Stevens et al.30	12
Mellano et al.32	10
Levy et al.33	12
Wirth et al.37	10
Cox et al.35	10
Latif et al.36	8
Comparative studies
Wiater et al.31	17
Morris et al.34	18

Patient demographics

A total of 253 patients were included in this systematic review, including 54 who underwent bilateral anatomic TSA,^28,29 168 who underwent bilateral RTSA,^{30–34, 37} and 31 who underwent bilateral anatomic TSA/RTSA.^35,36 Patient age at the time of the first surgery ranged from 45²⁹ to 92³⁷ years (anatomic TSA, 45²⁹ to 87²⁹ years; RTSA, 51³⁴ to 92³⁷ years). The percent of patients available at final follow-up ranged from 67%³⁷ to 100%.^{28,30,32–36}

The included poster directly compared the outcomes of 26 patients who had undergone bilateral anatomic TSA to 13 patients who had undergone bilateral RTSA. Patients were matched by gender, age, and time from first arthroplasty. Additional patient demographics and study design information can be found in Tables 1 and 2.

Table 1.

Studies included.

Study	PDR	Country	Study design	MCMS	Follow-up, n/total (%)	Males (%)	Age, years	IBA, months	Follow-up, months
TSA
Gruson et al.28	2002–2007	United States	Retrospective review, Level IV	56	13	30.8%	74.2	7.40	30.9
Fabricant et al.29	2002–2012	United States	Retrospective review, Level III	64	41/57 (72%)	43.9%	70.0	16.0	52.0
RTSA
Stevens et al.30	2004–2012	United States	Case series, Level IV	60	15/15 (100%)	33.3%	72.9	21.6	33.4
Wiater et al.31	2004–2010	United States	Prospective cohort, Level II	71	16/20 (76%)	25.0%	72.0	13.7	33.0
Mellano et al.32	2004–2013	United States	Case series, Level IV	63	50/50 (100%)	38.0%	71.8	14.0	61.0
Levy et al.33	2007–2013	United Kingdom	Case series, Level IV	66	19/19 (100%)	21.1%	74.5	18.2	48.4
Morris et al.34	2004–2011	United States	Retrospective review, Level III	66	11/11 (100%)	27.3%	67.1	8.00	36.8
Wirth et al.37	NR	Switzerland	Case series, Level IV	76	57/57 (100%)	29.8%	75.0	32.1	12a
TSA/RTSA
Cox et al.35	2004–2015	United States	Retrospective review, Level III	55	19/19 (100%)	42.1%	70.6	14.0 (median)	TSA: 62.3 RTSA:38.1b
Latif et al.36	1992–2009	France	Retrospective review, Level III	63	12/13 (92%)	16.7%	TSA: 70.8 RTSA: 72.9	33.6	TSA: 79.3 RTSA: 54.0b
Poster
Cox et al.38	2004–2015	United States	Retrospective review, Level III	–	Bilateral TSA: 26 Bilateral RTSA: 13	23.1%	TSA: 71.2 RTSA: 73.4	TSA: 20.4 RTSA: 18.6	Bilateral TSA: 63.4 Bilateral RTSA: 62.7

TSA: total shoulder arthroplasty; RTSA: reverse total shoulder arthroplasty; PDR: procedure date range; IBA: interval between arthroplasties; NR: not reported; MCMS: Modified Coleman Methodology Score.

When available, patient age at first surgery, interval between surgeries, and follow-up time are reported as a mean. Follow-up, n/total (%) represents the number of patients available at follow-up.

^aStudy involved multiple follow-up periods; however, patients were lost at the 24-month follow-up, so all data reported are from the 12-month follow-up.

^bOverall mean follow-up from the final procedure was not recorded, rather, follow-up was reported based on procedure type. The number of patients available for follow-up was not clearly reported in two studies,^13,23 therefore, only the number of patients is reported.

Table 2.

Studies included.

Study	Study summary	Control/Comparison group
TSA
Gruson et al.28	Retrospective review of patients who underwent staged bilateral TSA.	None
Fabricant et al.29	Retrospective cohort design study investigating outcomes of patients who underwent TSA with four different intervals between arthroplasties: <6 months, 6–12 months, 12–24 months, and >24 months.	None
RTSA
Stevens et al.44	Retrospective review of patients who underwent staged bilateral RTSA.	None
Wiater et al.31	Prospective cohort study comparing outcomes of patients with severe cuff deficiency who undergo staged bilateral RTSA and patients who undergo unilateral RTSA.	A matched control of patients who had undergone unilateral RTSA was selected for each shoulder based on age, pre-operative diagnosis, gender, prosthesis, and length of follow-up.
Mellano et al.32	Retrospective review of patients who underwent staged bilateral RTSA.	None
Levy et al.33	Review of the prospectively collected data of patients who underwent RTSA.	None
Morris et al.34	Review of the prospectively collected data of patients who underwent RTSA for rotator cuff arthropathy.	A matched control of patients who underwent unilateral RTSA was selected for each shoulder based on age, gender and diagnosis.
Wirth et al.37	Review of the prospectively collected data of patients who underwent RTSA.	None
TSA/RTSA
Cox et al.35	Retrospective review of patients who underwent bilateral shoulder arthroplasties with a TSA and a contralateral RTSA.	None
Latif et al.36	Retrospective review of patients who underwent bilateral shoulder arthroplasties with a TSA and a contralateral RTSA.	None

TSA: total shoulder arthroplasty; RTSA: reverse total shoulder arthroplasty.

Shoulder diagnoses and pathologies

Total shoulder arthroplasty

Gruson et al.²⁸ indicated that 25 shoulders (96.2%) undergoing anatomic TSA were diagnosed with primary osteoarthritis (OA) and 1 shoulder (3.8%) was diagnosed with post-capsulorraphy arthritis. Additionally, one shoulder underwent a concomitant open acromioplasty.

Fabricant et al.²⁹ did not report the patients' shoulder diagnoses and pathologies.

Reverse total shoulder arthroplasty

Indications for RTSA included rotator cuff tear arthropathy (CTA),^30–34 rheumatoid arthritis,^31,33 OA,^31–33 failed humeral head replacement,³¹ prior failed hemiarthroplasty,³² and failed rotator cuff repair.³³ The number of patients who had previous rotator cuff surgeries were noted by Stevens et al.³⁰ and Mellano et al.³² who reported 11 (36.7%) and 15 (15%) shoulders, respectively. Levy et al.³³ indicated that the RTSA was performed as a revision arthroplasty in seven shoulders (18.4%). It is important to note that the results of revision arthroplasty are less favorable than primary arthroplasty and outcomes in these patients should be viewed as distinct from those in patients undergoing primary RTSA. Revised implants included two stemmed hemiarthroplasties (Nottingham), three resurfacing prostheses (Copeland), one stemmed TSA (Bimodular), and one stemmed RTSA (Delta).

Surgical technique and prosthesis

Information on the surgical techniques and prostheses can be found in Table 4.

Table 4.

Surgical techniques and prostheses.

Study	Surgical technique	Concomitant procedures	Prostheses
Total shoulder arthroplasty
Gruson et al.28	Standard Deltopectoral Approach (n = 26).	Lesser tuberosity osteotomy (n = 25) Subscapularis tendon elevation (n = 1)	Cemented prosthesis (n = 24) Uncemented prosthesis (n = 2)
Fabricant et al.29	NR	NR	Bio-modular choice system (n = 14) Comprehensive total shoulder system (n = 68)
Cox et al.35	Standard Deltopectoral Approach (n = 19)	NR	Global AP system (n = 15) Aequalis (n = 3) Titan System (n = 1)
Reverse total shoulder arthroplasty
Stevens et al.30	Standard Deltopectoral Approach (n = 30)	Glenoid augmentation due to deformity or insufficient bone (n = 9) Latissimus dorsi transfer (n = 1) Platelet-rich plasma (n = 23)	Equinoxe (n = 19) Aequalis (n = 6) Encore (n = 4) Delta (n = 1)
Wiater et al.31	Standard Deltopectoral Approach (n = 32)	NR	TMRS (n = 10) Delta (n = 3) Aequalis (n = 3)
Mellano et al.32	Standard Deltopectoral Approach (n = 100)	Glenoid augmentation due to deformity or insufficient bone (n = 37) Latissimus dorsi transfer (n = 2)	TMRS (n = 70) Encore (n = 15) Aequalis (n = 13) Delta (n = 1)
Levy et al.33	Standard Deltopectoral Approach (n = 1) Neviaser-Mackenzie Anterosuperior Approach (n = 37)	NR	Verso (n = 38)
Morris et al.34	NR	NR	Aequalis (n = 22)
Wirth et al.37a	Standard Deltopectoral Approach (n = 114)	NR	Promos Reverse (n = 34) SMR Reverse Aequalis Univers Reverse
Cox et al.35	Standard Deltopectoral Approach (n = 19)	NR	TMRS (n = 7) Delta (6) Aequalis (n = 3) Titan (n = 2) DJO (n = 1)

NR: not reported; TMRS: Trabecular Metal Reverse Shoulder (Zimmer Biomet, Warsaw, IN); Bio-Modular Choice System (Biomet Inc, Warsaw, IN); Comprehensive Total Shoulder System (Biomet Inc, Warsaw, IN); Global AP System (DePuy Synthes, Warsaw, IN); Aequalis (Wright Medical, Memphis, TN); Aequalis Reverse (Tornier, Bloomington, MN); Titan (Integra, LifeSciences, Plainsboro, NJ); Encore Reverse (DJO Surgical, Vista, CA); Delta (DePuy, Paoli, PA); Promos Reverse (Smith & Nephew Orthopaedics AG, Rotkreuz, Switzerland); SMR Reverse (Lima Corporate S.P.A., Udine, Italy); Univers Revers (Arthrex Inc, Naples, FL); DJO (DJO Surgical, Vista, CA); Verso (Innovative Design Orthopaedics, London, UK; formerly Biomet, Swindon, UK).

^aWirth et al. report the types of prostheses used, but only the number for the Promos Reverse prosthesis which was used in the majority of cases.

Clinical outcomes

Revision rate and complications

Revision rates were reported in nine studies.^{28–34,36,37} Overall, revision rates ranged from 0%^30,33–35 to 10.53%³⁷ (anatomic TSA, 2.43% to 7.69%; RTSA, 0% to 10.53%; anatomic TSA/RTSA, 0% to 8.33%). In patients undergoing bilateral anatomic TSA, three shoulders required revision surgery at a mean follow-up of 13.2 months. Re-operation in these shoulders were due to post-traumatic loosening of the glenoid component (n = 1), periprosthetic join infection (n = 1), and pain and stiffness related to component size (n = 1). In patients undergoing bilateral RTSA, 10 shoulders required revision surgery due to instability (n = 5), periprosthetic joint infection (n = 2), decoupling of the glenosphere component from the base plate (n = 1), suspicion of septic arthritis (n = 1), and stem loosening (n = 1). Only Wirth et al.³⁷ recorded the time to revision surgery in bilateral RTSA patients and indicated that the six re-operations in their study occurred at a mean follow-up of 4.6 months. In patients undergoing anatomic TSA with contralateral RTSA one anatomic TSA required revision two months postoperatively following an anterior dislocation.

Overall, complication rates ranged from 4.88%²⁹ to 31.3%³¹ (anatomic TSA, 4.88% to 15.4%; RTSA, 10.5% to 31.3%; anatomic TSA/RTSA, 12.5% to 26.3%). Common complications included fractures of the acromion, scapular spine, or humeral shaft,^30,32–34 and periprosthetic infection.^29,32 Wirth et al.³⁷ reported 17 complications; however, they only expanded on those that required re-operation (n = 9). More information on the complications exhibited in each study can be seen in Table 5.

Table 5.

Type of complications exhibited.

Study	Type of arthroplasty	Type of complications
Gruson et al.28	Bilateral TSA	Post-traumatic loosening of the glenoid component (n = 1). Fixed anterior subluxation with limitation of motion (n = 1).
Fabricant et al.29	Bilateral TSA	Hematogenous infection (n = 1). Pain/stiffness related to component size (n = 1).
Stevens et al.30	Bilateral RTSA	Bilateral scapular spine insufficiency fractures (n = 1). Periprosthetic fracture (n = 1).
Wiater et al.31	Bilateral RTSA	Scapular spine nonunion (n = 2). Transient brachial plexopathy (n = 1). Instability (n = 1). Transient radial nerve palsy (n = 1).
Mellano et al.32	Bilateral RTSA	Prosthetic instability (n = 3). Acromial fracture (n = 5). Periprosthetic joint infection (n = 1).
Levy et al.33	Bilateral RTSA	Traumatic scapular spine fracture (n = 2).
Morris et al.34	Bilateral RTSA	Humeral shaft fracture (n = 1). Radial nerve traction injury (n = 1).
Wirth et al.37	Bilateral RTSA	Prosthesis dislocation (n = 5). Infected hematoma (n = 1). Progressive decoupling of the glenosphere component from the baseplate (n = 1). Stem loosening (n = 1). Ventral synovial impingement (n = 1). Specific complications not reported (n = 8).
Cox et al.35	Bilateral TSA/RTSA	TSA: Rotator cuff tear (n = 1). Failed subscapularis repair (n = 1). Type B periprosthetic fracture (n = 1). RTSA: Acromial stress fracture (n = 1). Lower extremity deep venous thrombosis (n = 1).
Latif et al.36	Bilateral TSA/RTSA	TSA: Anterior dislocation (n = 1). RTSA: Neuropathy of the axillary and suprascapular nerves (n = 1).

Range of motion

Neither study performing bilateral anatomic TSA reported both pre- and post-operative ROM outcomes.

Five studies^{30–32,34,39} which performed bilateral RTSA reported pre- and post-operative ROM outcomes. All five studies^{30–32,34,39} indicated significant improvements in FF, three studies^32,34,39 reported significant improvements in ER and abduction, and two studies^32,39 reported significant improvements in IR. Notably, Stevens et al.³⁰ indicated a mean loss in ER following bilateral RTSA.

Latif et al.³⁶ reported pre- and post-operative ROM outcomes in patients undergoing bilateral anatomic TSA/RTSA and indicated significant improvements in FF and IR for both anatomic TSA and RTSA shoulders. They also report significant improvements in ER in the anatomic TSA shoulder and a mean loss in ER in the RTSA shoulder.

The poster by Cox et al.³⁸ indicated that patients undergoing bilateral anatomic TSA reported mean improvements of 51°, 18°, and 4° in FF, ER, and IR, respectively, compared to mean improvements of 63° and 2° in FF and IR and a mean loss of 6° in ER in bilateral RTSA patients. All ROM values can be seen in Table 6.

Table 6.

Range of motion (ROM).

Study	Forward flexion		External rotation		Internal rotation		Abduction
	Preop	Postop	Preop	Postop	Preop	Postop	Preop	Postop
TSA
Gruson et al.28	NR	NR	NR	NR	NR	NR	NR	NR
Fabricant et al.29	NR	154°	NR	70°	NR	L3	NR	145°
RTSA
Stevens et al.30	81°	116.5°,*	31.5°	24°	L4	L1	103.5	109.5
Wiater et al.31	54°	123.5°,*	16.5°	18°	LSJ	TCJ	NR	NR
Mellano et al.32	72°	136°,*	26°	45°,*	39° (ABD)	58° (ABD)*	84°	121°,*
Levy et al.33	58°	143°,*	20°	32°,*	9° (ABD)	81° (ABD)*	60°	130°,*
Morris et al.34	41.5°	130°,*	12.5° (ABD)	28° (ABD)*	LSJ	L4	40.5°	129°,*
Wirth et al.37	NR	133.5°	NR	22.5°	NR	NR	NR	NR
TSA/RTSA
Cox et al.35	88°	139°	22°	34°	LSJ	L2	NR	NR
Latif et al.36	TSA: 85.8° RTSA: 85.8°	TSA: 143.3°,* RTSA: 142.1°,*	TSA: 2.9° RTSA: 13.3°	TSA: 42.9°,* RTSA: 12.1°	TSA: Sacrum RTSA: Sacrum	TSA: T8* RTSA: L1*	NR	NR
Poster
Cox et al.38	TSA: 107° RTSA: 75.5°	TSA: 156 RTSA: 135.5	TSA: 26 RTSA: 25.5	TSA: 42.5 RTSA: 24.5	TSA: L5 RTSA: LSJ	TSA: L1 RTSA: L3	NR	NR

LSJ: lumbosacral junction; TCJ: thoracolumbar junction; preop: preoperative; postop: postoperative; ABD: in abduction; NR: not reported.

Data are reported as mean values between both shoulders.

^*Significant improvement.

Patient-reported outcomes

All PRO scores are shown in Table 7. All studies reporting pre- and post-operative outcome scores indicated significant improvements in PROs following bilateral TSA (anatomic and/or reverse).

Table 7.

Patient reported outcomes (PROs).

	CMS		VAS-Pain		SST		ASES		SSV
Study	Preop	Postop	Preop	Postop	Preop	Postop	Preop	Postop	Preop	Postop
TSA
Gruson et al.28	38.1	72.5*	6.9	0.9*	3.0	10.0*	NR	NR	NR	NR
Fabricant et al.29	NR	72.0	NR	NR	NR	9.0	NR	NR	NR	NR
RTSA
Stevens et al.30	33	60.0*	NR	NR	4.0	9.0*	36.0	79.5*	NR	NR
Wiater et al.31	28.9	57.3*	6.8	1.6*	NR	NR	31.3	76.0*	24.1	74.5*
Mellano et al.32	NR	NR	5.5	0.7*	2.5	8.9*	37.5	76.7*	NR	NR
Levy et al.33	18.7	65.1*	NR	NR	NR	NR	NR	NR	21	92.0*
Morris et al.34	25.2	86.1*	NR	NR	NR	NR	25.2	70.8*	NR	NR
Wirth et al.37	32	66.0*	NR	NR	NR	NR	NR	NR	NR	NR
TSA/RTSA
Cox et al.35	NR	NR	NR	TSA: 0.7 RTSA: 1.5	NR	TSA: 8.4 RTSA: 7.4	NR	TSA: 87 RTSA: 81.5	NR	NR
Latif et al.36	TSA: 27.9 RTSA: 35.5	TSA: 76.8* RTSA: 72.6*	NR	NR	NR	NR	NR	TSA: 89.6 RTSA: 82.4	NR	TSA: 85.4 RTSA: 82.5
Poster
Cox et al.38	NR	NR	NR	TSA: 0.4 RTSA: 0.8	NR	NR	NR	TSA: 94.2 RTSA: 84.6**	NR	NR

NR: not reported; CMS: Constant-Murley Scores; VAS: Visual Analog Scores; SST: Simple Shoulder Test; ASES: American Shoulder and Elbow Surgeons Shoulder Score; SSV: Subjective Shoulder Value.

Data are reported as mean values between shoulders.

^*Significant improvement between pre-operative and post-operative scores.

^**Significant difference between TSA and RTSA values.

Seven studies^{28,30,31,33,34,36,37} indicated significant improvements between pre-operative and post-operative CMS scores (p < 0.05). Notably, Latif et al.³⁶ indicated no significant differences in the mean post-operative CMS scores for anatomic TSA and RTSA shoulders (anatomic TSA: 77; RTSA: 73; p = 0.25).

Three studies^28,31,32 reported significant improvements between pre-operative and post-operative VAS-pain scores (p < 0.05). Cox et al.³⁵ indicated no significant differences between mean post-operative VAS-pain scores for anatomic TSA and RTSA shoulders (anatomic TSA, 0.7 ± 1.4; RTSA, 1.5 ± 2.3; p = 0.31). Similarly, the poster by Cox et al.³⁸ indicated an insignificant difference between post-operative VAS-pain scores in bilateral anatomic TSA and bilateral RTSA patients (p > 0.05).

Three studies^28,30,32 indicated significant improvements between pre-operative and post-operative SST scores (p < 0.05), and Cox et al.³⁵ indicated no significant differences in post-operative SST scores between anatomic TSA and RTSA shoulders (TSA, 8.4; RTSA, 7.4; p = 0.35).

Four studies^30–32,34 reported significant improvements between pre-operative and post-operative ASES scores (p < 0.05), and Cox et al.³⁵ indicated no significant differences between anatomic TSA and RTSA shoulders in ASES scores (TSA, 87; RTSA, 81.5; p = 0.38). However, the poster by Cox et al.³⁸ indicated significantly greater ASES scores in patients undergoing bilateral anatomic TSA than those undergoing bilateral RSTA (TSA, 94.2; RTSA, 84.6; p < 0.05).

Two studies^31,33 reported significant improvements between pre-operative and post-operative SSV scores (p < 0.05). Latif et al.³⁶ did not report pre-operative values, but they note an insignificant difference between anatomic TSA and RTSA shoulders in SSV scores (anatomic TSA, 85.4%; RTSA, 82.5%; p = 0.63).

Patient satisfaction

Six studies (anatomic TSA, 2; RTSA, 3; anatomic TSA/RTSA, 2)^{28,29,31,34–36} reported patient satisfaction rates. Overall satisfaction with the procedures ranged from 81%³⁴ to 100%.²⁸ These studies also compared patient satisfaction rates between shoulders.

Of the studies reporting on bilateral anatomic TSA, Gruson et al.²⁸ indicated 100% satisfaction on both the first and second arthroplasty and Fabricant et al.²⁹ reported 83% satisfaction on the first procedure and 90% satisfaction on the second.

In patients undergoing bilateral RTSA, satisfaction rates for the first procedure ranged from 81%³⁴ to 94%,³¹ and 75%³¹ to 81%³⁴ for the second procedure. Additionally, Wirth et al.³⁷ indicated that at the 12-month follow-up 92% of patients who underwent bilateral RTSA indicated that they would do the bilateral procedures again.

Of the studies reporting on anatomic TSA/RTSA, Latif et al.³⁶ indicated that 92% of patients were satisfied with the TSA shoulder, and similarly, 92% were satisfied with the RTSA shoulder. Cox et al.³⁵ used VAS satisfaction scores to describe patient satisfaction following anatomic TSA and RTSA and reported values of 8.7 and 8.0, respectively (p = 0.448). Cox et al.³⁵ also measured arthroplasty preference and indicated that 68.4% of patients preferred the RTSA shoulder, 5.3% preferred the TSA shoulder, and 26.3% reported no preference.

The poster by Cox et al.³⁸ reported greater patient satisfaction following bilateral anatomic TSA when compared to bilateral RTSA. There was an insignificant difference in the dominant arm (TSA, 9.7 ± 0.9; RTSA, 8.5 ± 1.8; p = 0.057); however, in the non-dominant arm the patient satisfaction was significantly greater in bilateral anatomic TSA patients (TSA, 9.7 ± 0.8; RTSA, 8.6 ± 1.1; p = 0.012).

Personal hygiene and ADLs

Six studies^30–34,36 examined the impact on ADLs following bilateral shoulder arthroplasties. Neither Gruson et al.²⁸ nor Fabricant et al.²⁹ reported these outcomes following bilateral anatomic TSA.

Two studies^30,32 indicated that the majority of patients noted no changes to their personal hygiene habits and ADLs following bilateral RTSA. Stevens et al.³⁰ reported that patients were able to perform personal hygiene and toilet needs “normally” in 80% of shoulders. However, all patients were able to perform personal hygiene and toilet needs with at least one shoulder. Mellano et al.³² reported that 94% of patients noted that they used the same hand when using the toilet, 67% noted no change in their hygiene habits, 50% did not require an assistive device in the shower, and 97% did not require an assistive device on the toilet. Furthermore, 66% of patients were able to wash the opposite shoulder with the contralateral hand. Similarly, Morris et al.³⁴ indicated that all patients were able to use at least one arm to comb their hair, reach their back pocket, eat with a utensil, wash the armpit of the opposite arm, and manage toileting.

Wiater et al.³¹ noted no significant correlation between the amount of active ER and the mean ASES ADL scores for the first or second shoulder in those undergoing bilateral RTSA (first shoulder, p = 0.484; second shoulder, p = 0.14). Levy et al.³³ reported ADLs requiring external rotation scores and indicated that all patients retained independence with personal hygiene and were able to resume their leisure and sporting activities without limitation.

Of the studies reporting on bilateral anatomic TSA/RTSA, Latif et al.³⁶ indicated significantly greater ASES ADL scores in the anatomic TSA shoulder in the following categories: “lift 10lb above shoulder”, “throw a ball overhand”, “do usual sport”, and “cumulative ADL score” (p < 0.0001 for all).

The poster by Cox et al.³⁸ reported activities of daily living external and internal rotation scores and indicated significantly greater results in the anatomic TSA shoulder in both dominant (anatomic TSA, 35.3; RTSA, 32.1; p = 0.001) and non-dominant arms (anatomic TSA, 35.5; RSA, 32.5; p = 0.001).

Sub-group analyses performed in the included studies

Fabricant et al.²⁹ measured the effect that the IBA had on PROs and patient satisfaction. They indicated that patients with less than six months between arthroplasties demonstrated significantly better UCLA scores than 6–12 month interval patients (p = 0.04), greater CMS than all other groups (p < 0.01), and greater SST scores compared with the 6–12 month and 12–24 month interval patients (p < 0.01). Additionally, they reported that patients in the <6 months interval group had 100% satisfaction rate, where those in the 6–12 months, 12–24 months, and >24 months interval groups reported satisfaction rates of 67%, 83%, and 78%, respectively.

Discussion

The principle finding of this study is that at a minimum follow-up of 12 months after the contralateral (second) shoulder arthroplasty, patients undergoing bilateral anatomic TSA, RTSA or anatomic TSA/RTSA experience overall improvements in ROM, PROs and reportedly high rates of patient satisfaction. Additionally, the majority of patients were able to successfully perform ADLs and were often able to complete perineal hygiene using one arm. When comparing outcomes of bilateral anatomic TSA to bilateral RTSA, patients who underwent bilateral anatomic TSA had superior functional results and satisfaction; however, patients with bilateral RTSA were still satisfied with their results.

These results indicated consistently low mean revision rates following each type of bilateral shoulder arthroplasty; however, complication rates were varying, with patients undergoing bilateral anatomic TSA experiencing lower complication rates than patients undergoing bilateral RTSA or anatomic TSA/RTSA. It is important to note, however, that only two studies examined bilateral anatomic TSA and bilateral anatomic TSA/RTSA compared to six studies that examined bilateral RTSA. Also, the low sample size of the included studies may indicate indefinite conclusions primarily related to overall mean revision and complication rates. Furthermore, the overall heterogeneity of the studies makes the interpretation of the literature difficult, thus making this review primarily descriptive in nature.

Due to the potential for functional limitations in both shoulders following these procedures, concerns still remain regarding the impact on the patient's ADLs. Moreover, it has been indicated that ER is expected to be limited following RTSA,⁴ and because adequate rotational movements are necessary for performing ADLs, surgeons may be even more tentative to perform bilateral RTSA or anatomic TSA/RTSA.^4,19,31 Overall mean decrease in ER values following bilateral RTSA was reported by both Stevens et al.³⁰ and Cox et al.³⁸ Although overall improvements in ER are indicated in the remaining studies reporting this outcome, the results of this systematic review indicate relatively lower post-operative ER values following RTSA when compared to anatomic TSA. The reduced improvements in ER may correspond with greater pre-operative rotator cuff damage. Mellano et al.³² reported greater ER improvements in patients undergoing bilateral RTSA when the patient is presenting with OA rather than rotator CTA, or with rotator CTA rather than a major irreparable rotator cuff tear. In cases with severe rotator cuff damage, latissimus dorsi transfers combined with RTSA have been successful for restoring active ER.^40–46

However, discrepancies in post-operative ER may not significantly influence usage rates. A recent study⁴⁷ indicated no significant differences in shoulder motion or frequency between patients undergoing unilateral anatomic TSA and unilateral RTSA. They also indicated that shoulder motion after anatomic TSA or RTSA was not significantly different from the contralateral asymptomatic side. This suggests that while maximum ER may be limited following RTSA when compared to anatomic TSA, this limitation may not affect daily motion.

Furthermore, Wiater et al.³¹ indicated no significant correlation between the amount of active ER and the mean ASES ADL scores in either shoulder following bilateral RTSA. This suggests that although ER plays a role in accomplishing ADLs, patients may still successfully perform perineal hygiene and other ADLs with some limitations in ER. All studies reporting on ADLs indicated that the majority of patients experienced little to no change in hygiene habits and ADLs. These results suggest that bilateral anatomic TSA and/or RTSA may not impact the patient's ability to perform ADLs to the degree that it was once believed.

The two studies^35,36 reporting clinical outcome scores following anatomic TSA/RTSA indicated an insignificant difference between anatomic TSA and RTSA shoulders in all included PROs. In contrast, the poster by Cox et al.³⁸ reported significantly greater post-operative ASES scores and patient satisfaction following bilateral anatomic TSA when compared to bilateral RTSA. This suggests that more favorable outcomes may be expected following anatomic TSA; however, patients may still experience similar results following RTSA. Interestingly, Cox et al.³⁵ indicated that the majority of patients undergoing bilateral anatomic TSA/RTSA preferred the RTSA shoulder rather than the anatomic TSA shoulder. This result may be due to the pre-operative differences in pain and limitations between shoulders. Anatomic TSA and RTSA are performed for different reasons (generally osteoarthritis versus rotator cuff arthropathy) which may influence outcomes.

Overall patient satisfaction rates following bilateral TSA (anatomic and/or reverse) were consistently high across all included studies. Fabricant et al.²⁹ reported increased patient satisfaction after the second anatomic TSA when compared to the first and indicated that patients more frequently reported the second shoulder recovery as being easier than the first. This may be the result of having a better contralateral support system in which to rely and provide a less painful assist during the recovery period for the second replacement. Also, patient expectations may impact the outcomes, namely, knowing what to expect from the first arthroplasty may influence the recovery of the second. However, Wiater et al.³¹ indicated decreased patient satisfaction following the second RTSA when compared to the first. These conflicting results suggest that more qualitative research is needed to determine if one procedure is easier from the patient's perspective.

Fabricant et al.²⁹ indicated that the time between arthroplasties may have an effect on PROs and on patient satisfaction, with shorter intervals corresponding to improved scores. The same study suggests that a shorter proximity between the two surgical procedures may permit a shared recovery time where the rehabilitation of both shoulders could have occurred simultaneously. This may result in an overall decreased length of recovery and the patient's perceived time away from normal functioning. Additionally, the patient's satisfaction with the first procedure may influence the time between arthroplasties. Namely, if the patient is not satisfied with the initial surgery, the IBA may be extended as the patient becomes more reluctant to move forward with the second procedure. When preparing for bilateral anatomic TSA, it may be recommended to minimize the IBA, but more research is necessary to determine if the IBA has a definitive role on outcomes following bilateral anatomic TSA.

Limitations

It is important to note the limitations of this study. The quality of the included studies is limited based on the low MINORS scores. The majority of studies included are Level III or IV and retrospective in nature, therefore, selection bias is a notable limitation. Also, among the included studies there is variance relating to surgical techniques, reported outcomes, concomitant injuries, and patient demographics. The sample sizes for the studies were small, and due to the heterogeneity and variance in reported outcomes, the overall sample size for each outcome was limited and the presence of detection bias was increased. Additionally, the length of follow-up was limited in the Wirth et al.³⁷ study as post-operative outcomes are examined at 12 months, rather than 24 months, due to a loss of follow-up.

Conclusions

Although literature on clinical outcomes following bilateral TSA is limited, the available data indicate that bilateral anatomic TSA, RTSA, and anatomic TSA/RTSA allow for functional and pain improvements and results in high patient satisfaction. Future research should examine the roll that the IBA has on clinical outcomes and whether there is a trend relating to the outcomes of the procedures.

ORCID iDs

John-Rudolph H Smith https://orcid.org/0000-0001-6206-1966

Darby A Houck https://orcid.org/0000-0003-0367-8987