To evaluate the safety and feasibility of outpatient total hip and knee arthroplasties by comparing the results of 90-day follow-up after the operation between propensity matched cohorts that undergo inpatient vs outpatient total hip and knee arthroplasties.

From December 2019 to December 2021, a total of 198 hip and knee arthroplasties were performed in Ambulatory Surgical Center of Shanghai Changzheng Hospital. The inclusion criteria: age ≤70 years old; plasma albumin ≥35 g/L; hemoglobin ≥110 g/L; American Society of Anesthesiologists (ASA) grade Ⅰ and Ⅱ; body mass index (BMI) ≤30 kg/m². Exclusion criteria: severe joint deformity; patients with history of cardiovascular and cerebrovascular diseases, diabetes mellitus, arrhythmia, and severe liver or renal insufficiency and so on. Among them, 185 patients as the outpatient group were successfully propensity matched to 185 inpatients undergoing total hip and knee arthroplasty as the inpatient group, based on gender, age, ASA grade, and BMI. The matched cohort included 227 total hip arthroplasties(THAs) and 143 total knee arthroplasties(TKAs). Reoperation and readmission incidences, major and minor complications, emergency department visits, and unplanned outpatient visits within 90 days after surgery were retrospectively analyzed in the two groups. Wilcoxon rank sum test and Chi-square test were used for statistical analysis.

The demographics of patients undergoing outpatient surgery (outpatient group) and inpatient surgery (inpatient group) were similar, which meant a successful match. The rates of reoperations in both groups were 1.1% (χ²=0, P>0.05), and readmissions rates were 2.2%、2.7% respectively in two groups(χ²=0, P>0.05). No statistically significant difference was found in total complications, major complications, or minor complications(χ²=0.178, 0, 0.069, all P>0.05). Incidence of dislocation after THA were 0.8% and 0.9% respectively(χ²=0, P>0.05); the rates of arthrofibrosis requiring manipulation under anesthesia after TKA were 3.0% and 3.9% respectively(χ²=0, P>0.05), rates of emergency department visits were 1.6% and 2.2% respectively(χ²=0, P>0.05); all these data showed no statistically significant difference between two groups. In terms of unplanned outpatient visits, the incidences in the outpatient and inpatient groups(13.0% vs 5.9%, χ²=5.333, P=0.021), in the first 93 and last 92 of the outpatient group(18.3% vs 7.6%, χ²=4.665, P=0.031), and in the first 93 of the outpatient group and the inpatient group (5.9% vs 18.3%, χ²=10.394, P =0.001) showed statistically significant differences; while there was no statistically significant difference in incidence between the inpatient group and the last 92 of the outpatient group(5.9% vs 7.6%, χ²=0.280, P>0.05.

The results indicate that the model of outpatient hip and knee arthroplasties has similar safety and efficacy to the traditional surgical model among appropriately selected patients.