Day case versus inpatient total shoulder arthroplasty: A systematic review and meta-analysis

Abstract

Background

A number of papers have been published comparing the safety and efficacy of day case and inpatient anatomic or reverse total shoulder arthroplasty. However, no systematic review of the literature has been published to date. The aim of this review was to determine if day case total shoulder arthroplasty (length of stay <24 h) leads to similar outcomes as standard-stay inpatients (length of stay ≥24 h).

Methods

The US National Library of Medicine (PubMed/MEDLINE), EMBASE, and the Cochrane Database of Systematic Reviewers were queried for publications utilizing keywords that were pertinent to total shoulder arthroplasty, day case, outpatient and inpatient, clinical or functional outcomes, and complications. In order to determine the quantitative impact of day case total shoulder arthroplasty on readmission and revision rate, a meta-analysis was performed on articles that observed 30- or 90-day readmission or revision.

Results

Eight articles were found to be suitable for inclusion in the present study which included 6103 day case total shoulder arthroplasty and 147,463 inpatient total shoulder arthroplasty. Following meta-analysis, there was no significant difference among patients who underwent day case total shoulder arthroplasty compared to inpatient total shoulder arthroplasty regarding revision rates (OR: 1.001; 95% CI: 0.721–1.389; p = 0.995) and 30-day readmission rates (OR: 0.940; 95% CI: 0.723–1.223; p = 0.646). In contrast, patients who underwent day case total shoulder arthroplasty were less likely to have a readmission within 90 days compared to their inpatient counterparts (OR: 0.839; 95% CI: 0.704–0.999; p = 0.049). Two out of eight studies reported comparable baseline clinical characteristics among groups, while five studies reported significant differences and one study did not provide information regarding clinical characteristics, such as medical comorbidities or American Society of Anaesthesiologists'(ASA) score. No significant difference among groups was found in all or almost all studies regarding mortality rates, and rates of cardiac complications, cerebrovascular events, thromboembolic events, pulmonary complications, cardiac complications, and nerve complications. Finally, results were rather conflicting regarding the correlation of day case total shoulder arthroplasty to the rate of surgical site infections.

Conclusions

This study showed that day case total shoulder arthroplasty might lead to similar rates of mortality, complications, revisions, and readmissions compared to inpatient total shoulder arthroplasty when used in a selected population of younger, healthier, and more male patients. In contrast, there was no consensus regarding the impact of day case total shoulder arthroplasty on the rate of surgical site infections. Finally, further research of higher quality is required to establish patient demographic criteria, ASA score, or comorbidity index cut off that might be used to define day case-treated patients who seem to have equivalent outcomes compared to inpatient-treated patients.

Level of evidence: Systematic review of level III studies (lowest level included).

Introduction

The shift toward performing orthopedic surgical procedures in the day case setting and in ambulatory surgical centers is a recognized trend of the last 20 years, while a multitude of patient, surgeon, and facility factors have led to the explosive growth of the day case model.^1–4 Policymakers and hospitals are frequently looking at length of stay (LOS) after surgery as an area of focus for improvement, with recent interest in day case total joint arthroplasty. ⁵ Overall, improved surgical techniques, pain management, and perioperative patient management have led to declining LOS. ⁵

With an aging population, it has been projected that shoulder arthroplasty will increase by 755% in patients older than 55 years by 2030, ⁶ which represents an increasing burden on our medical system in terms of costs and resources.

Day case total shoulder arthroplasty (TSA) has been touted as an alternative to inpatient arthroplasty owing to the potential for decreased overall operating costs, enhanced profitability, and patient–provider incentivization without sacrificing care quality.^7,8 However, although some studies have shown promise in using day case TSA, the results may not be generally applied. ⁹ Furthermore, while reports have emerged demonstrating that TSA can be performed safely in the day case setting, selecting patients who are suited for outpatient TSA remains challenging. ¹⁰ Therefore, understanding the outcomes of day case TSA is imperative to surgical predicate decision making. ⁹

A number of papers have been published comparing the safety and efficacy of day case and inpatient TSA. However, no systematic review of the literature has been published to date. The aim of this review was to determine if day case TSA leads to similar outcomes as standard-stay inpatients. Specifically, we aimed to answer the following questions: (1) Are the baseline characteristics statistically comparable in the day case and the inpatient groups? (2) Does day case TSA lead to increased early mortality rate compared to inpatient TSA? (3) Is day case TSA a significant risk factor for early and short-term complications compared to inpatient TSA? (4) Does day case TSA result in increased readmission rate compared to inpatient TSA? (5) Does day case TSA correlate with increased revision rate compared to inpatient TSA? (6) Is day case TSA associated with decreased blood transfusion rate compared to inpatient TSA? Our hypothesis is that day case TSA in selected patients leads to similar outcomes compared to inpatient TSA.

Materials and methods

We performed a systematic review of the literature according to the PRISMA guidelines in order to identify articles reporting the rate of outcomes and complications following anatomic or reverse TSA among patients in the day case (LOS <24 h) and inpatient setting (LOS ≥24 h). The US National Library of Medicine (PubMed/MEDLINE), EMBASE, and the Cochrane Database of Systematic Reviewers were queried for publications utilizing keywords that were pertinent to TSA, day case and inpatient, clinical or functional outcomes, and complications. Only abstracts that evaluated both inpatient and day case TSA were included in this review. The specific search terms are further shown in Table 1.

Table 1.

Search strategy.

Database	PubMed, Cochrane, EMBASE
Date	August 2019
Strategy	#1 AND #2 AND #3
Limit	Human AND English
#1	“Arthroplasty, Replacement, Shoulder”[Mesh] OR total shoulder arthroplasty[tw] OR tsa[tw] OR Total shoulder replacement[tw] OR total shoulder replacements[tw] OR TSR[tw] OR open shoulder surgery[tw]
#2	Outpatient [tw] OR inpatient[tw] OR “Ambulatory Surgery Center”[tw] OR “ASC”[tw] OR “Same Day” [tw] OR “Rapid Recovery”[tw]
#3	ROM[tw] OR range of motion[tw] OR Treatment outcome[tw] OR Clinical outcome[tw] OR “Range of Motion, Articular”[Mesh] OR “Treatment Outcome”[Mesh] OR “Patient Reported Outcome Measures”[Mesh] OR “Patient Outcome Assessment”[Mesh] OR “Outcome Assessment (Health Care)”[Mesh] OR “Postoperative Complications”[Mesh]

Inclusion and exclusion criteria

Inclusion criteria were original articles investigating outcomes and complications of patients who underwent primary TSA treated in a day case setting in comparison with patient treated in an inpatient setting. The inclusion criteria were (1) studies describing human subjects of any age and sex, (2) studies that include a population of at least 10 patients who underwent anatomic or reverse TSA, (3) studies that present data on complication rates and/or other outcome measures for patients who underwent TSA stratified based upon day case or inpatient setting, and (4) studies that described a functional outcome score among patients who received TSA in day case or inpatient setting.

Exclusion criteria consisted of (1) case reports or series with <10 patients; (2) review articles; (3) expert opinions; (4) letters to editors; (5) instructional course lectures; (6) studies involving animals or cadavers, or in vitro investigations; (7) book chapters; (8) abstracts from scientific meetings; (9) unpublished reports; (10) studies in which no subjects underwent anatomic or reverse TSA; (11) studies analyzing different surgeries in conjunction with TSA; (12) studies that did not include location of TSA; and (13) non-English language publications. A cross-reference search of the selected articles was also performed to obtain other relevant articles.

Data collection

Two independent reviewers separately conducted the search by title and abstract. If the title and abstract of each study contained insufficient information to determine its appropriateness for inclusion, the full manuscript was reviewed. The full texts of the chosen papers were obtained to decide whether they were suitable for the purpose of the study. If there was disagreement between the two reviewers, discrepancies between the two authors were resolved by discussion.

The level of evidence in the included studies was determined using the Oxford Centre for Evidence-Based Medicine—Levels of Evidence. ¹¹ Selective reporting bias was assessed for all included articles in this analysis. The methodological quality of each study and the different types of detected bias were assessed independently by each reviewer with the use of modified Coleman methodology score. ¹²

The two reviewers independently extracted the data including the title, year of publication, authors, study design, number of patients, number of joints, functional outcome score, PROMs, ROM, postoperative complication, revision rate, and readmission rate.

Meta-analysis

In order to determine the quantitative impact of day case TSA on readmission and revision rate, a meta-analysis was performed on articles that observed 30- or 90-day readmission or revision. Comprehensive Meta-Analysis software (Biostat Inc., Englewood, NJ) was used to conduct the meta-analysis. Due to the heterogeneity present among the articles, a random-effects model meta-analysis was chosen. Outcomes are reported as odds ratio (OR), with 95% CI and p-value.

Results

Search results

Using the outlined search criteria, 438 studies were identified (Figure 1). Among those, 258 duplicate articles were identified and removed from the search. Following the removal of duplicate articles, the 180 remaining articles were subject to application of the predetermined inclusion and exclusion criteria. Following application of these criteria, 14 articles were subject to a full text screening process, with six articles excluded from final analysis (Figure 1). Eventually, eight articles met the inclusion criteria and were included in this analysis.^8,13–19

Figure 1.

PRISMA 2009 flow diagram.

All studies included in this review were retrospective case–control studies of level of evidence III.^8,13–19 The total mean modified Coleman score of the review was 42, ranging from 30 to 54, indicating a low methodological quality (Table 2).

Table 2.

Type of study, level of evidence, and modified Coleman methodology score (0–100).

Study	Type	Level of evidence	Coleman
Gallay et al. 18	Retrospective case–control (pilot study)	III	50
Leroux et al. 8	Retrospective case–control	III	30
Brolin et al. 16	Retrospective case–control	III	54
Cancienne et al. 17	Retrospective case–control	III	43
Basques et al. 14	Retrospective case–control	III	36
Arshi et al. 13	Retrospective case–control	III	41
Nwankwo et al. 19	Retrospective case–control	III	42
Bean et al. 15	Retrospective case–control	III	39

Meta-analysis

In total, three studies reported on 30-day readmission^2,11,18 and were determined to be comparable for the meta-analysis.^8,14,17 Following meta-analysis, there was no significant difference seen among patients who underwent day case TSA compared to inpatient (OR: 0.940; 95% CI: 0.723–1.223; p = 0.646; Figure 2).

Figure 2.

Meta-analysis of 30-day readmission rates showing no significant differences among groups.

Among those studies which reported on 90-day readmission,^14–17,19 four studies were determined to be comparable for the meta-analysis,^14,15,17,19 whereas one study did not showcase any 90-day readmissions and was unable to be included for analysis. ¹⁶ Following meta-analysis, patients who underwent day case TSA were less likely to have a readmission within 90 days compared to their inpatient counterparts (OR: 0.839; 95% CI: 0.704–0.999; p = 0.049; Figure 3).

Figure 3.

Meta-analysis of 90-day readmission rates favours day case TSA.

Among studies that report on revision rate,^{8,13,16,17,19} only four were determined to be comparable for meta-analysis.^8,13,17,19 Following meta-analysis, no difference was observed in revision rate among patients who received day case or inpatient TSA (OR: 1.001; 95% CI: 0.721–1.389; p = 0.995; Figure 4).

Figure 4.

Meta-analysis of revision rates showing no significant differences among groups.

Systematic review

Demographics

In total, 153,566 patients were included in this review. Six thousand one hundred three patients underwent day case TSA (50.1% females) while 147,463 patients underwent inpatient TSA (52.3% females). The mean age ranged from 18 years ⁸ to 79 years ¹⁷ in the day case group and from 54.2 years ¹⁶ to 79 years ¹⁷ in the inpatient group. Finally, the mean follow-up varied between 1 and 12 months (Table 3).

Table 3.

Baseline characteristics per group and follow-up per study.

Study	Number of patients	Outpatient: Gender (% male)	Outpatient: Age	Inpatient: Gender (% male)	Inpatient: Age	Follow-up
Gallay et al. 18	16	NR%	53	NR%	61	9–32 months
Leroux et al. 8	7197	58.1%	18–64 (53.5%)	43.2%	65–74 (38.8%)	30 days
Brolin et al. 16	60	76.67%	52.6 (33–68)	46.67%	54.2 (28–63)	90 days
Cancienne et al. 17	5165	45.3%	70–79 (45.5%)	45.3%	70–79 (45.6%)	30 days (readmission, cost) 90 days (medical complications) 1 year (shoulder specific complications)
Basques et al. 14	123,347	52.4%	65–69 (42.8%)	37.9%	70–74 (23.4%)	30 days (readmission) 90 days (complications)
Arshi et al. 13	17,542	7.3%	70–74 (8.7%)	70.2%	70–74 (100.3%)	14 days (pneumonia, ARF), 30 days (cerebrovascular accident, MI), 60 days (DVT, PE), 1 year (revisions)
Nwankwo et al. 19	182	53.39%	68.1 (31–90)	42.19%	72.4 (41–88)	90 days
Bean et al. 15	57	50%	59.8	48.7%	59.9	90 days

ARF: Acute Respiratory Failure; DVT: Deep Vein Thrombosis; MI: Myocardial Infarction; NR: Not Referred; PE: Pulmonary Embolism.

Comparison of the baseline demographic and clinical characteristics

All studies of this analysis compared baseline demographic and/or clinical characteristics between inpatient and day case TSA groups.^8,13–18 Two out of eight studies (25%) reported comparable baseline clinical characteristics among groups,^16,17 while five studies^8,13–15,18 reported significant differences and one study did not provide information regarding clinical characteristics, such as medical comorbidities or ASA score. ¹⁹

Brolin et al. ¹⁶ and Cancienne et al. ¹⁷ documented similar comorbidities index among groups. In addition, Brolin et al. ¹⁶ reported no significant differences in ASA scores. On the contrary, both Arshi et al. ¹³ and Basques et al. ¹⁴ found that the day case group had significantly decreased incidence of medical comorbidities compared to the inpatient group, while Bean et al. ¹⁵ and Leroux et al. ⁸ reported that the day case group had significantly lower ASA score.

Most patients in both groups were found with ASA grade 2 or 3. However, Bean et al. ¹⁵ documented that most day case ASA scores were ASA grade 2 and most inpatient ASA scores were grade 3, whereas Gallay et al. ¹⁸ noted that the day case-treated patients’ ASA scores were grade 1 and 2 and the inpatients’ ASA scores were grade 2 and 3.

While two studies^16,17 found similar demographic characteristics among groups, there were six studies that documented significant differences in the demographic characteristics of the two groups.^{8,13–15,18,19} These studies reported that the day case group included younger and more male patients compared to the inpatient group (Table 4).

Table 4.

Comparability in the initial demographic/clinical characteristics among groups as well as readmission rates, transfusion rates, and revision rates per group.

Study	Comparable initial groups	Inpatient’s mean length of hospital stay	Readmission rate	Revision Rate	Blood transfusion rate
Gallay et al. 18	No (patients in Group I lower ASA)	1 day	9–32 mo: 0 (0.00%) 0 (0.00%)	NR	NR
Leroux et al. 8	No (outpatients significantly younger, more males, lower ASA and BMI scores, and lower rates of pulmonary disease and hypertension)	2.1 ± 1.8 days	30 d: 2 (1.74%) 173 (2.93%)	30 d: 0 (0.00%) 75 (1.07%)	30 d: 1 (0.58%) 315 (4.48%)
Brolin et al. 16	Yes	2 days	90 d: 0 (0.00%) 0 (0.00%)	90 d: 0 (0.00%) 0 (0.00%)	90 d: 0 (0.00%) 1 (3.3%)
Cancienne et al. 17	Yes	NR	30 d: 40 (5.7%) (outpat) 225 (5.0%) (inpat) 90 d: 66 (9.3%) (outpat) 401 (9%) (inpat)	1 y: 18 (2.5%) (outpat) 107 (2.4%) (inpat)	90 d: 25 (3.5%) 209 (4.7%)
Basques et al. 14	No (outpatients younger, more males, and higher rates of obesity but lower rates of diabetes, CAD, CHF, and CKD)	≥24 h	30 d: 22 (0.60%) 990 (0.83%) 90 d: 75 (2.04%) 3442 (2.87%)	NR	90 d: 18 (0.49%) 1102 (0.92%)
Arshi et al. 13	No (outpatients higher incidence in Midwest and South)	>24 h	1 y: NR	6 mo: 16 (1.03%) 172 (1.08%) 1 y: 22 (1.41%) 223 (1.39%)	NR
Nwankwo et al. 19	No (outpatients younger and lower ASA)	≥24 h	90 d: 3 (2.5%) 6 (9.4%)	90 d: 0 (0.00%) 1 (1.6%)	NR
Bean et al. 15	No (outpatients lower ASA score)	24 h	90 d: 0 (0.00%) 1 (2.5%)	NR	NR

ASA: American Society of Anaesthesiologists; BMI: Body Mass Index; CAD: coronary artery disease; CHF: Congestive Heart Failure; CKD: Chronic Kidney Disease; NR: Not Referred.

Mortality

The overall mortality rate was 0.3% (1 out of 349 patients) for patients who underwent day case TSA and 0.2% (12 out of 7163 patients) for those treated with inpatient TSA. Four studies examined postoperative mortality rates.^8,15,18,19 Two of them reported on 90-day mortality rate,^15,19 while one study looked at 30-day mortality rate ⁸ and another one reported on mortality rate with a range from 9 to 32 months after surgery. ¹⁸ No significant difference among groups was found regarding mortality rates in any of these four studies^8,15,18,19 (Table 5).

Table 5.

Complications and mortality rates after outpatient and inpatient TSA.

Study	Infection rate (n, %)	Wound issues rate (n, %)	Nerve issues rate (n, %)	Thromboembolic event rate (n, %)	Cerebrovascular accident rate (n, %)	Myocardial infarction rate (n, %)	Acute renal failure rate (n, %)	Pneumonia rate (n, %)	Mortality rate (n, %)
Gallay et al. 18	9–32 mo: Outp: 0 (0.00%) Inp: 0 (0.00%)	9–32 mo (1 d): Outp: 3 (37.5%) Inp: 0 (0.00%)	9–32 mo: Outp: 0 (0.00%) Inp: 0 (0.00%)	9–32 mo: Outp: 0 (0.00%) Inp: 0 (0.00%)	9–32 mo: Outp: 0 (0.00%) Inp: 0 (0.00%)	9–32 mo: Outp: 0 (0.00%) Inp: 0 (0.00%)	9–32 mo: Outp: 0 (0.00%) Inp: 0 (0.00%)	9–32 mo: Outp: 0 (0.00%) Inp: 0 (0.00%)	9–32 mo: Outp: 0 (0.00%) Inp: 0 (0.00%)
Leroux et al. 8	30 d: Outp: 0 (0.00%) Inp: 48 (0.68%)	30 d: Outp: 0 (0.00%) Inp: 6 (0.09%)	30 d: Outp: 0 (0.00%) Inp: 9 (0.13%)	30 d: Outp: 0 (0.00%) Inp: 48 (0.68%)	30 d: Outp: 0 (0.00%) Inp: 9 (0.13%)	30 d: Outp: 1 (0.58%) Inp: 12 (0.17%)	30 d: Outp: 0 (0.00%) Inp: 3 (0.04%)	NR	30 d: Outp: 0 (0.00%) Inp: 12 (0.17%)
Brolin et al. 16	90 d: Outp: 0 (0.00%) Inp: 0 (0.00%)	NR	NR	90 d: Outp: 0 (0.00%) Inp: 1 (3.3%)	NR	90 d: Outp: 0 (0.00%) Inp: 0 (0.00%)	NR	90 d: Outp: 0 (0.00%) Inp: 0 (0.00%)	90 d: Outp: 0 (0.00%) Inp: 0 (0.00%)
Cancienne et al. 17	90 d: Outp: 21 (3%) Inp: 119 (2.7%)	NR	NR	90 d: Outp: 21 (3%) Inp: 133 (3%)	90 d: Outp: 21 (3%) Inp: 125 (2.8%)	90 d: Outp: 7 (1%0 Inp: 40 (0.9%)	90 d: Outp: 32 (4.5%) Inp: 193 (4.3%)	90 d: Outp: 6 (0.8%) Inp: 35 (0.8%)	NR
Basques et al. 14	90 d: Outp: 12 (0.33%) Inp: 621 (0.52%)	90 d (hematoma): Outp: 4 (0.11%) Inp: 389 (0.32%)	90 d: Outp: 1 (0.03%) Inp: 35 (0.03%)	90 d: Outp: 11 (0.30%) Inp: 1052 (0.88%)	NR	NR	90 d: Outp: 11 (0.30%) 1010 Inp: (0.84%)	90 d: Outp: 11 (0.30%) Inp: 1149 (0.96%)	NR
Arshi et al. 13	6 mo: Outp: 13 (0.84%) Inp: 91 (0.57%) 1 y: Outp: 14 (0.90%) Inp: 104 (0.65%)	NR	NR	60 d: Outp: 11 (0.71%) Inp: 291 (1.82%)	30 d: Outp: 0 (0.00%) Inp: 124 (0.78%)	30 d: Outp: 0 (0.00%) Inp: 110 (0.69%)	14 d: Outp: 35 (2.25%) Inp: 452 (2.83%)	14 d: Outp: 0 (0.00%) Inp: 13 (0.08%)	NR
Nwankwo et al. 19	90 days: Outp: 0 (0.00%) Inp: 0 (0,00%)	90 d: Outp: 4 (3.4%) Inp: 2 (3.1%)	90 d: Outp: 1 (0.8%) Inp: 1 (1.6%)	NR	90 d: Outp: 0 (0.00%) Inp: 2 (3.1%)	90 d: Outp: 0 (0.00%) Inp: 0 (0.00%)	NR	NR	90 d: Outp: 1 (0.8%) (28 d) Inp: 0 (0.00%)
Bean et al. 15	NR	NR	90 d: Outp: 1 (0.05%) Inp: 2 (0.05%)	90 d: Outp: 0 (0.00%) Inp: 1 (0.03%)	NR	90 d: Outp: 0 (0.00%) Inp: 1 (0.03%)	NR	NR	90 d: Outp: 0 (0.00%) Inp: 0 (0.00%)

NR: Not Referred.

Cardiac complications

The overall rate of cardiac complications was 0.3% (8 out of 2610 patients) for patients who underwent day case TSA and 0.6% (163 out of 27,609 patients) for those treated with inpatient TSA. Seven out of the eight studies compared cardiac complications among groups.^8,13,16–19 All these seven studies reported that there were not any significant differences between day case- and inpatient-treated groups in the rates of cardiac complications (Table 5).

Pulmonary complications

The overall rate of pulmonary complications was 0.3% (17 out of 5792 patients) for patients who underwent day case TSA and 0.9% (1197 out of 140,338 patients) for those treated with inpatient TSA. Five studies looked at pneumonia rates.^{13,14,16–18} All five studies showed no significant differences at pneumonia rates between inpatient- and day case-treated groups^{13,14,16–18} (Table 5).

Acute renal failure

The overall rate of acute renal failure was 1.3% (78 out of 5935 patients) for patients who underwent day case TSA and 1.1% (1658 out of 147,332 patients) for those treated with inpatient TSA. Five studies stratified the acute renal failure rates between day case-TSA-treated and inpatient-TSA-treated cases.^{8,13,14,17,18} Four of them depicted that there was not any significant difference between the two groups.^8,13,17,18 In contrast, one study noted a higher risk (p value < 0.001) of acute renal failure for the outpatient group compared to the day case group¹⁴ (Table 5).

Thromboembolic events (VTE)

The overall VTE rate was 0.7% (43 out of 5985 patients) for patients who underwent day case TSA and 0.01% (1526 out of 147,399 patients) for those treated with inpatient TSA. Overall, seven studies provided information regarding the rate of VTE.^8,13–18 Six studies showed no significant differences between the two groups.^8,13,15–18 In contrast, one study reported a significantly higher (p value <0.001, OR = 2.6) VTE rate in the inpatient group compared to the day case group¹⁴ (Table 5).

Cerebrovascular events

The overall rate of cerebrovascular accidents for the day case cohort was 0.4% (21 out of 5935 patients), while the overall rate of cerebrovascular accidents for the inpatient cohort was 0.9% (260 out of 27,542 patients). Five studies stratified the rate of cerebrovascular accidents per group.^8,13,17–19 All five studies showed no significant differences among groups regarding the rate of cerebrovascular accidents (Table 5).

Surgical site infection (SSI) rate

The overall infection rate was 2.3% (137 out of 6083 patients) for patients who underwent day case TSA and 0.6% (892 out of 147,463 patients) for those treated with inpatient TSA. Seven studies investigated the early postoperative SSI rate requiring irrigation and debridement and/or arthrotomy.^{8,13,14,16–19} Five of them^8,16–19 illustrated no significant differences among groups. Among those, there were two studies^16,17 with comparable demographic and clinical characteristics, while the other three^8,18,19 reported that the day case group included younger, healthier, and more male patients compared to the inpatient group. In contrast, two studies^13,14 noted significantly different rates of early SSI among groups. Both these studies^13,14 also included younger, healthier, and more male patients in the day case group. Specifically, Basques et al. ¹⁴ found increased rate of early postoperative infections in the inpatient TSA group compared to the day case group (p value = 0.002). On the contrary, Arshi et al. ¹³ reported increased infection rates (p value < 0.001 at one year) in the day case TSA group compared to the inpatient TSA group.

Soft-tissue complications

Four studies compared the rates of soft-tissue complications among groups.^8,14,18,19 The superficial wound complications rate was 0.4% (17 out of 3792 patients) in the day case group and 0.3% (397 out of 126,950) in the inpatient group. Three out of these four studies did not find any significant difference among groups,^8,18,19 while the fourth one showed that the inpatient group was associated with increased rate (p value = 0.006) of subcutaneous hematomas compared to the day case group¹⁴ (Table 5).

Nerve complications

The overall nerve complications rate was 0.08% (3 out of 3812 patients) for patients treated with day case TSA and 0.04% (47 out of 126,987 patients) for those who underwent inpatient TSA. Five studies compared all-cause nerve issues rates between day case and inpatient TSA.^{8,14,16,18,19} All of these studies noted that there were not any significant differences among two groups (Table 5).

All-cause revision rate

The overall revision rate at 12-month follow-up was 1.8% (40 out of 2261 patients) for the day case-treated patients and 1.6% (330 out of 20,446 patients) for the inpatients treated with TSA.^13,17 Five studies compared all-cause revision rates between day case and inpatient TSA to show that there were not any significant differences among groups at the early to short follow-up^{8,13,16,17,19} (Table 4).

Readmission rate

The all-cause readmission rate at 30-day follow-up was 1.5% (64 out of 4372 patients) in the day case group and 1% (1388 out of 131,337 patients) in the inpatient group treated with TSA,^8,14,17 whereas the overall readmission rate at 90-day follow-up was 3.3% (144 out of 4367 patients) in the inpatients and 3% (3850 out of 124,444 patients) in the day case-treated patients.^14,16,17 No readmission rate was reported in the study by Arshi et al. ¹³ at one-year follow-up, while Gallay et al. ¹⁸ reported no readmissions at all during their study period from 9 to 32 months. Readmission rate was significantly higher at 90-day follow-up in the inpatient cohort in the study by Nwankwo et al. ¹⁹ (p value = 0.049, OR = 0.252) and at 30-day follow-up in the inpatients older than 85 years and with COPD in the study by Basques et al. ¹⁴ (p value = 0.016). All- cause readmission rates were similar in the studies by Leroux et al., ⁸ Bean et al., ¹⁵ Brolin et al., ¹⁶ and Cancienne et al. ¹⁷

Clinical and functional outcome variables

In total, two out of the eight studies reported on clinical and functional subjective outcome variables.^15,18 Gallay et al. found similar mean visual analogue scale (VAS) pain scores between the two groups, while Bean et al. ¹⁵ reported significantly lower mean VAS pain score two weeks postoperatively in the day case group compared to the inpatient group.

Perioperative blood transfusion

Four studies assessed the rate of blood transfusion in the different groups of patients. The overall transfusion rate in the day case cohort was 1% (44 out of 4402 patients), while the overall transfusion rate in the inpatient cohort was 1.2% (1627 out of 131,367 patients). Two studies showed significantly increased transfusion rates in the inpatient group compared to the day case group.^14,17 Another two studies reported no significant differences in the transfusion rates between the two groups, although they noted a non-significant trend toward increased blood transfusions in the inpatient group^8,16 (Table 4).

Discussion

Our meta-analysis showed that patients treated with day case TSA did not have any significant differences in 30-day readmission rates, while day case TSA was associated with decreased 90-day readmission rates compared to the inpatient TSA. In addition, based on our meta-analysis, there was not any difference in revision rates among groups in the short term (12 months). According to our systematic analysis, all or almost all studies reported no differences among groups in mortality rates as well as rates of cardiac complications, respiratory complications, nerve complications, thromboembolic events, and cerebrovascular events. Finally, most studies did not find any differences among groups with regards to wound complications and acute renal failures.

In contrast, six out of the eight studies of this review reported significant differences in the demographic and baseline clinical characteristics of the two groups.^{8,13–15,18,19} The typical pattern observed was younger, healthier and male patients chosen for the day case TSA procedure. However, there was no evidence to establish patient demographic criteria, ASA score, or comorbidity index cut off that might be used to define day case-treated patients who have equivalent outcomes compared to inpatient-treated patients.

However, there were two studies^16,17 which involved matched-paired groups with statistically comparable baseline clinical (ASA grade, comorbidity index) and demographic characteristics and, mainly, patients suffering from mild systemic diseases (ASA grade 2). In both studies with comparable baseline clinical and demographic characteristics, it was demonstrated that there were no differences among groups in the all-cause revision and readmission rates as well as complication rates, such as cardiac complications, pulmonary complications, thromboembolic events, acute renal failures, and SSIs. Based on these findings, one must conclude day case TSA is a safe procedure in patients with mild (ASA 2) systemic diseases. This does not mean that all patients are ideal candidates for day case TSA. We still feel that patients who are at increased risk for cardiac events or they are suffering from a severe systemic disease that is a constant threat to life should be rather recommended to select an inpatient procedure for their treatment.

Furthermore, although the overall SSI rate of the day case patient cohort (2.3%) of the studies included in this analysis was higher than the overall SSI rate of the inpatient cohort (0.6%), we could not pool together and compare these rates in a meaningful statistical manner, since the follow-up documentation of this outcome varied widely among studies. Furthermore, five^8,16–19 studies reported similar rates of SSI among groups. Two of them^16,17 were the only studies included in this review with comparable baseline demographic and clinical (ASA/comorbidity index) characteristics between the two groups. On the contrary, one large database study ¹⁴ reported significantly higher rates of SSI in the inpatient group compared to the day case group. However, the inpatient TSA group of this study ¹⁴ consisted of significantly older and more female patients with increased comorbidities that might be related to the increased SSI rates observed in this group. In contrast, another large database study ¹³ reported an increased rate of SSI in the day case group, despite the fact that they were younger and healthier compared to the inpatient group. Overall, based on the current literature, there was no consensus on the association between day case TSA and rate of SSI, and further studies of higher quality are required to lead to definite conclusions.

It is notable that all studies assessing all-cause revision rate depicted that there were not any differences between day case and inpatient TSA. In contrast, there was some evidence to show increased rates of blood transfusions in the inpatient cohort. Particularly, two studies found significantly increased transfusion rates in the inpatient group compared to the day case group and another two studies noted a trend toward a non-significant increase in the transfusion rate of the inpatient cohort.^16,17 We assume that the inpatient TSA population might be associated with increased needs of blood transfusion as a result of the higher incidence of co-existing medical comorbidities which was reported in them compared to the day case TSA cohort.

The biggest limitation of this review was that most studies assessed heterogeneous populations with significantly different baseline characteristics among groups that make it difficult to draw conclusions about safety. Specifically, since day case TSA cohort in most studies included younger and healthier patients compared to the inpatient TSA patient cohort, it is possible that inpatient TSA patients might have done better if their demographics and comorbidities would have been comparable to those of the day case TSA patients. Other limitations included a lack of prospective studies, randomization, and blinding. The quality of the studies was low and potential bias in the retrospective design might influence the results. Moreover, the studies varied widely in regards to follow-up and number of patients, while there were not any studies to report mid- to long-term outcomes. On the other hand, the total number of patients was adequately high, while the rationale was to compare outpatient and inpatient TSA as for the early postoperative morbidity and mortality rates.

Conclusions

This study showed that day case TSA might lead to similar rates of mortality, complications, revisions, and readmissions compared to inpatient TSA when used in a selected population of younger, healthier, and more male patients. In contrast, there was no consensus regarding the impact of day case TSA on the rate of SSIs. Finally, further research of higher quality is required to establish patient demographic criteria, ASA score, or comorbidity index cut off that might be used to define day case-treated patients who seem to have equivalent outcomes compared to inpatient-treated patients.

ORCID iDs

Michael-Alexander Malahias https://orcid.org/0000-0001-8463-5975

Dimitris Karanikas https://orcid.org/0000-0002-9570-0653