Safety and Cost Effectiveness of Outpatient TSA

Results

Study Inclusion

The database search returned 230 articles for assessment, and two additional articles were added after reference review of included articles. A total of 129 articles existed after removing duplicates, of which 37 made it through the title and abstract screening process and underwent a full article text review. Ultimately, there were no disagreements between reviewers on which should be included. After full-text review, 21 outpatient TSA studies met the inclusion criteria for the systematic review (Figure 1). All studies except for one study^[8] (which was performed in Canada) were performed in the United States and were published from 2005 to 2020. One (5%) level II prospective cohort study was noted,^[7] and no level I randomized trials. Seventeen (81%) were level III, retrospective outpatient versus inpatient comparative^{[8–10,20–24,27,29–36]} and three (14%) were level IV, retrospective outpatient-only case series.^[25,26,28] Ten studies (48%) used state or national registries for data collection,^{[10,20–23,27,32,34–36]} and 11 (52%) involved single-center or multicenter patient data.^{[7–9,24–26,28–31,33]} These 21 studies accounted for 25,808 outpatient TSAs and 231,408 inpatient TSAs. Five studies included only anatomic TSA,^{[8,9,22,29,31]} 1 study included only reverse TSA,^[33] 10 studies included anatomic and reverse TSA,^{[10,20,21,23,24,26–28,32,36]} 3 studies included anatomic, reverse, and hemiarthroplasty,^[25,30,35] and 2 did not specifically state which type of TSA was included.^[7,34] Details on all included studies are compiled in Supplemental Table I (https://links.lww.com/JAAOS/A712).

(Enlarge Image)

Figure 1.

Chart showing the PRISMA flow diagram. PRISMA = Preferred Reporting Items for Systematic Reviews and Meta-Analyses, TSA = total shoulder arthroplasty

Study Quality Assessment

Mean Methodological Index for Non-Randomized Studies scores were 18.4 ± 1.7 for comparative studies (n = 18) and 12.7 ± 0.6 for case series (n = 3). The two independent reviewers scored each study identically. Individual scores can be found in the Appendix (https://links.lww.com/JAAOS/A715).

Patient Demographics

Of the 15 comparative cohort studies that recorded patient demographic differences among inpatient and outpatient cohorts,^{[8–10,20,22–24,27,29,30,32–36]} inpatients were recorded as older in 11 studies,^{[8,10,20,23,27,29,30,32,34–36]} had higher BMI in 6 studies,^{[10,20,27,29,33,36]} and higher American Society of Anesthesiologists (ASA) Physical status classification in 10 studies.^{[10,20,23,24,27,29,30,34–36]} Two of the comparative studies were matched cohorts with no statistical differences in any patient demographics.^[9,22]

Readmissions, Emergency Department Visits, and Conversion to Overnight Stay

Fourteen studies, including 6,133 outpatient TSAs, described readmission rates, with six also describing postoperative ED visits (Supplemental Table II, https://links.lww.com/JAAOS/A713). The pooled outpatient readmission rate was 3.9% (range, 0% to 9.3%), and ED visit rate was 11.1% (range, 0% to 37.5%). Only one of 10 studies comparing readmission rates between outpatient and inpatient TSA found significantly higher readmissions after inpatient TSA (2.9% versus 2.0%, P < 0.001).^[23] Although not statistically significant, one study found a trend toward increased readmissions after outpatient TSA (7.8% versus 6.3%, P = 0.053),^[20] whereas another trended toward more readmissions after inpatient surgery (9.4% versus 2.5%, P = 0.068).^[30] Three articles^[22,25,30] discussed reasons for outpatient readmissions; medical complications (n = 57, 67.0%), implant-related complications (n = 17, 20.0%), and uncontrolled pain (n = 11, 13.0%) were the most common. Five studies found no statistical difference in ED visits between outpatient and inpatient TSA. Ten studies, with 446 outpatient TSAs, evaluated the rate of SDD, with only six patients (1.3%) requiring an unplanned overnight stay.^{[7–9,24–26,28–31]} Reasons for failed SDD were surgery delays (n = 1, 16.7%),^[7] patient convenience (n = 1, 16.7%),^[28] hypoxia (n = 1, 16.7%),^[28] or not meeting postanesthesia care unit (PACU) discharge criteria (n = 3, 50.0%).^[30]

Complications

Fourteen studies, including 6,937 outpatient TSAs, evaluated perioperative complications (Supplemental Table III, https://links.lww.com/JAAOS/A714). The total, shoulder-related, and medical-related complication rates were 7.6% (range, 0% to 24.2%), 3.5% (range, 0 to 14%), and 4.2% (range, 0 to 19.1%), respectively. Revision surgery and blood transfusion rates were 2.2% (range, 0% to 6%) and 0.9% (range, 0 to 3.5%), respectively. Finally, infection, venous thromboembolism, and mortality rates were 0.7% (range, 0 to 3%), 0.7% (range, 0% to 3%), and 0.6% (range, 0 to 1.2%), respectively.

Eleven studies compared complication rates between outpatient and inpatient TSA.^{[8–10,22–24,27,29,30,32,33]} Only one study found outpatient TSA with higher complication rates, specifically, in the rates of 1-year irrigation and débridement/arthrotomy (0.9% versus 0.65%, P < 0.001).^[32] Five studies found outpatient TSA were at a lower risk of overall complications,^[10,33] medical complications,^[22,23] and shoulder-related complications (eg, dislocation, manipulation under anesthesia, surgical site infection, capsulitis, and hematoma).^[23,32] Six studies found no difference in complications between cohorts.^{[8,9,24,27,29,30]}

Costs

Four studies, including 3,298 outpatient TSAs, evaluated charges and reimbursements for the episode of care.^[20–22,31] Ode et al^[20] found a significant difference in outpatient versus inpatient TSA charges ($37,395 versus $62,905, P < 0.001), which was similar to what Gregory et al^[21] found ($22,907 versus $76,109, P < 0.001). Ode et al^[20] additionally identified lower outpatient TSA charges in an ambulatory surgery center (ASC) versus the hospital setting ($31,790 versus $55,990, P < 0.001). Cancienne et al^[22] looked at actual reimbursement differences paid by insurance companies, which were significantly less for outpatient TSA ($14,722 versus $18,336, P < 0.001). Walters et al^[31] also evaluated outpatient TSA in the ASC and found that a bundled payment program reduced total charges of care, mainly through reduced implant charges ($42,410 versus $44,530, P = 0.024).

Patient Selection

Patient selection was discussed in 10 studies,^{[8,9,24–26,28–31,33]} with all studies highlighting the need for preoperative clearance by a medical or anesthesia specialists before outpatient candidacy. Fournier et al^[26] sought to validate a stepwise, patient selection algorithm for outpatient TSA in a freestanding ASC. They held that patients younger than 70 years were candidates; they must have no evidence of preoperative anemia or previous venous thromboembolism and required evaluation of pulmonary (no more than one risk factor and no chronic obstructive pulmonary disease [COPD] with the need for home oxygen) and cardiac comorbidities (no anticoagulation use and two or fewer cardiac stents). After meeting the criteria of this algorithm, all patients would meet with ASC anesthesia staff for optimization and final risk stratification. Using this algorithm, all 61 patients were able to successfully undergo planned SDD, with no 90-day readmissions. Two additional studies used this algorithm, with all patients having successful SDD and no hospital readmissions.^[29,31] Five studies highlighted the need to assess for an adequate social support before outpatient consideration,^{[8,24,25,30,33]} and two discussed proximity of residence to the hospital or ASC.^[8,24] All but one study included ASA Physical Status scores of 3 or less for outpatient candidacy, with one study requiring 2 or less.^[8] Three studies listed preoperative opioid dependence as a contraindication.^[9,24,25] One study had additional qualifications, including a BMI less than 30, age younger than 65, and no ambulatory aid dependence.^[24] The study by Leroux et al^[28] highlighted that sleep apnea and previous coronary or cerebrovascular events were not an absolute contraindication if they were treated and stable.

Three studies used the American College of Surgeons National Surgical Quality Improvement Program to perform regression analyses to identify TSA patients who would be less likely have hospital LOS less than 24 hours after TSA.^[34–36] Overall, increasing age, hypertension, COPD, diabetes, bleeding disorder, chronic heart failure (CHF), dialysis, cancer, poor functional status, and ASA class 3 or 4 were all risk factors for LOS of greater than 24 hours.

Patient Satisfaction and Clinical Outcome Scores

Four studies evaluated patient satisfaction after outpatient TSA.^[7,8,28,29] Gallay et al^[8] surveyed eight patients who underwent outpatient TSA, with all responding that they were "very satisfied" and would undergo outpatient TSA again. Ilfeld et al^[7] found that all six of the patients who underwent outpatient TSA reported 10 of 10 satisfactions. Leroux et al^[28] surveyed 35 outpatient TSAs and found that 94.3% of patients felt ready for discharge, had controlled pain on night 1 after surgery, and would have outpatient TSA again. Of the 35 patients, 97% rated their experience as good or excellent.^[28] Nelson et al^[29] compared satisfaction between outpatient TSA at an ASC with inpatient TSA and found no differences in satisfaction with the location (4.65 versus 4.37, P = 0.18) or overall experience (4.4 versus 4.6, P = 0.30); however, more patients preferred to undergo outpatient versus inpatient TSA (94.2% versus 71.7%, P = 0.035) if given the option.

Four studies evaluated clinical outcomes after outpatient TSA.^[8,24,25,33] Bean et al^[24] found decreased visual analog scores (VAS) for pain after outpatient TSA at 2 weeks postoperatively (2.0 versus 3.0, P = 0.044), but no difference at further follow-up. Charles et al^[25] did not find any differences in Single Assessment Numeric Evaluation, American Shoulder and Elbow Surgeons, VAS, or ROM when comparing ASC with hospital-based outpatient TSA. Erickson et al^[33] did not find any differences in these scores between outpatient and inpatient reverse TSA up to 2 years postoperatively. Gallay et al^[8] found equal VAS pain scores and ROM between outpatient and inpatient TSA.

Risk Factors for Outpatient Readmission and Complications

Two studies evaluated the risk factors for readmission after outpatient TSA among two large insurance data bases.^[22,23] Basques et al^[23] used the Medicare Standard Analytical File database and found that patients older than 85 years of age (odds ratio [OR] 34.1, P < 0.001) with COPD (OR 2.7, P = 0.036) had higher risk of readmission. Cancienne et al^[22] used the PearlDiver database and identified diabetes mellitus (OR 1.5, P = 0.025), peripheral vascular disease (OR 1.7, P = 0.025), CHF (OR 2.0, P = 0.014), chronic lung disease (OR 1.8, P = 0.021), depression (OR 2.4, P < 0.0001), and chronic anemia (OR 2.2, P = 0.001) to increase the risk of readmission.

Perioperative Protocols and Same Day Discharge Criteria

Eleven studies described their outpatient perioperative protocols including anesthesia techniques, multimodal pain regimens, utilization of tranexamic acid (TXA), rehabilitation, and follow-up protocols.^{[7–9,24–26,28–31,33]} Five studies used general anesthesia with local anesthetic injection before closure,^{[9,24,26,29,31]} and the six others used general anesthesia with interscalene,^[7,25,28,33] continuous brachial plexus,^[8] or supraclavicular blocks.^[30] Common preoperative and PACU oral medications included oxycodone, acetaminophen, gabapentin, celecoxib, ondansetron/promethazine, famotidine, and aspirin, with additional hydromorphone or morphine if needed for breakthrough pain. Intravenous antibiotic protocols were described in six studies and included single antibiotic use depending on allergy profile of cefazolin, clindamycin, or vancomycin,^{[8,24,25,30,33]} or a combination of cefazolin and vancomycin.^[9] Four studies used a single preoperative dose,^[9,25,30,33] one study used a redose in PACU,^[24] and another gave additional oral antibiotics until the removal of a peripheral nerve catheter.^[8] Two studies described using TXA to help prevent blood loss.^[24,28]

Seven studies reported their rehabilitative protocols,^{[7,9,25,26,29–31]} which were all very similar and included a restricted amount of passive ROM with Codman exercises up to 6 weeks postoperatively, gentle active ROM at 6 to 12 weeks, and then full lifting of restrictions at 12 weeks. Postoperative day 1 phone calls by nurses or ASC staff were commonly performed to assess for pain, numbness, nausea, urinary retention, surgical dressing care, and any other postsurgical-related issues.^{[7,9,24,26,29,31]} Seven studies also described the SDD criteria.^{[7,9,24,26,29–31]} Important discharge milestones included the ability to safely ambulate, to urinate with control, to tolerate a diet without nausea, and a VAS for pain of four or less, controlled with oral medications. Standard PACU criteria including stable respiratory status, oxygen saturation, hemodynamic stability, and the level of consciousness were needed before discharge.