Background The safety of outpatient versus inpatient total hip arthroplasty (THA) in the higher BMI population has not been described. The current study compared 90-day complications and 5-year revisions in patients following primary THA in the outpatient versus inpatient settings with BMI ≥ 35 kg/m2, a “high-risk” group as defined by prior studies.

Methods In this case-control study, adult patients with BMI ≥ 35 kg/m2 undergoing primary THA from 2010 to October 2022 were identified in the PearlDiver Mariner165 orthopedics database. Patients diagnosed with infection, trauma, or neoplasm of the hip within 90 days prior to surgery were excluded. Patients were categorized by service location of THA: inpatient or outpatient. 90-day postoperative adverse events were compared between groups with multivariable logistic regression controlling for age, sex, and Elixhauser Comorbidity Index (ECI). With Bonferroni correction, significance was established at P < 0.004. 5-year revision rates were compared with logrank test.

Results Of high-BMI patients undergoing THA, 39,396 were inpatient and 11,505 were outpatient. These groups were exact matched 1:1 on age, sex, and ECI, yielding 11,455 patients in each matched group. Multivariate analysis demonstrated that outpatient THA was associated with lower 90-day odds ratios (ORs) for: deep vein thrombosis (OR: 0.47), pulmonary embolism (OR: 0.63), surgical site infection (OR: 0.69), and readmissions (OR: 0.40), in addition to multiple minor adverse events (P < 0.001 for each). Outpatient THA in the study population was not associated with differences in 90-day hip-related adverse events (dislocation, fracture) or 5-year all-cause revisions (P = 0.5).

Conclusion Compared to inpatient THA, outpatient THA for those with BMI ≥ 35 kg/m2 was associated with lower odds of various severe and minor 90-day adverse events, with similar short-term hip-related outcome measures and 5-year revision rates. These findings are encouraging when considering performing outpatient THA in obese patients.