To our knowledge, no study has compared the difference between these two NA techniques. Early postoperative adverse events like uncontrolled pain, orthostatic hypotension, urinary retention, and prolonged motor block are linked to late patient mobilization, prolong hospitalization and failure to discharge in outpatient setting. The type of anesthesia used may have an important impact. Therefore, this study has the potential to improve the already established ERAS program and improve patients care perioperative and postoperative. Showing that SED-EA and SA are equivalent will allow for a more efficient and reliable technique for THA/TKA ERAS program that can be further translated into other lower limb surgeries.

Primary objective is to compare the overall complication rate within 72 hours after surgery, categorized according to the Clavien-Dindo classification (15), between both techniques following THA and TKA surgery.

Secondary objectives are to compare the following perioperative and postoperative events between both groups:

1. Perioperative

   1. Preoperative pain levels and opioid/analgesics consumption
   2. Time needed to perform the technique (from the first handling the needle to sterile drapes removal from the back of the patient)
   3. Time needed for the SA or SED-EA to achieve adequate sensory block (from LA injection to the absence of cold feeling at T8 allowing surgical incision
   4. Intraoperative blood loss
   5. Need for dose adjustment intraoperatively
   6. Hemodynamic instability defined by hypotension (-20% from basal values prior to entering the OR, at the time of the consent).

2. Post-operative

   1. Time to motor and sensory function return
   2. Time to mobilization
   3. Pain evaluated with Visual analog scale immediately after surgery and up to 72 hours after surgery
   4. Opioid consumption up to 48 hours
   5. Hospital LOS and incidence of failed discharge at planned time
   6. Complications related to the technique performed (Post-dural puncture headache, local infection, hematoma etc.)

HYPOTHESIS We hypothesize that the incidence of the overall complication rate within 3 days after surgery, categorized according to the Clavien-Dindo classification will be equivalent between both groups; SED-EA and SA.

Inclusion Criteria:

• Patient has symptomatic hip or knee OA requiring primary joint replacement (not associated with unusual treatments such as bone graft, concomitant osteotomy, revision implant, etc.).
• Patient who is candidate for our ERAS program
• Patient understands the study condition
• Patient capable of giving informed consent.
• Someone to accompany the patient to the Pre-admission Clinic and hospital the day of the surgery and to be available in the first postoperative week during home recovery.

Exclusion Criteria:

• Patients with contraindication to NA (spinal anatomical abnormalities, coagulation disorders, infection at the puncture site)
• Allergy to LAs used in the study
• Unable to communicate with the investigators, unable to read the questionnaire, unable to keep track and notes of the medication taken at home
• Lack of home services offered by the local community service centre in the area.
• BMI > 40.
• Psychiatric disease limiting participation or interfering with the ability to provide consent or assessment
• Need for long-term urinary Foley catheter post-op.
• Allergies to sulfonamides or other medications specified in the protocol.
• Cognitive impairment or communication problem
• Pulmonary embolism or deep vein thrombosis in the past year.
• Need for long-term anticoagulation therapy.
• Current corticotherapy or systemic corticotherapy in the past year (unless confirmation of a cortrosyn test prior to surgery).
• Systemic disease involvement (diabetes, heart, kidney, blood, etc.) necessitating special perioperative care (intensive care, multiple transfusions, dialysis, etc.).
• Coagulation disorder increasing the risk of intraoperative and postoperative bleeding including thrombocytopenia (platelet count lower than 80), hemophilia, prolonged INR (1,4 and over) and any order coagulation disorder deemed a contra-indication to neuraxial anesthesia..
• Locomotor problem, other than the joint to be replaced, imposing functional limitations that prevent movement without technical or physical assistance.
• Neurological or balance disorder.
• Living space incompatible with home care.
• Clcr < 30 ml/min (Cockcroft-Gault formula).
• Pregnancy

The group assignment will be given to the intraoperative anesthesia team by writing, by a research team member who will not participate in postoperative assessments. The intraoperative anesthesia team will not be blinded to the group assignment, but surgeons, patients, and assessors will be blinded. For the SA group, a sham epidural catheter will be placed on the patient's back by the anesthesia team in order to keep the surgeons, patients and assessors blind to the randomization group.