This study was a randomized, double-blind, sham procedure-controlled evaluation of tidal lavage as a treatment for knee osteoarthritis. We achieved blinding by raising a drape during the study procedure and mimicking the sensory aspects of tidal lavage, but not flushing the knee, in the sham lavage group. We gave all study participants subcutaneous anesthesia with lidocaine and then bupivacaine and then attempted aspiration of the knee, removing up to 5 ml of synovial fluid for examination and then injecting the knee with 20 ml of bupivicaine.

In people who received the sham treatment, we placed a 16-gauge catheter in the lateral suprapatellar position just to the knee joint capsule and infused small volumes (less than 5 ml per "exchange" x 20) of saline into the subcutaneous tissue while manipulating the anterior knee to mimic efforts to shift fluid within the knee during the "aspiration" phase of each exchange. Patients could see the supply bag of sterile saline but could not see the waste bag of this closed drainage system.

For tidal lavage patients, we made the catheter puncture in the lateral suprapatellar pouch and repeatedly distended the knee with 30-50 ml of sterile saline, then aspirated this volume (about 20 exchanges for a total of 1 liter of flush solution).

At the end of the procedure, we told the patients that their knee might be swollen due to retained saline, and to expect this additional swelling to resolve over 24-48 hours, during which time they were to minimize activity. The person who did the procedure (the principal investigator) then left the room. The study nurse, who was not present for the procedure and was blinded to the procedure's identity, asked the patients which treatment they thought they received (tidal lavage or sham lavage). We scheduled patient followups with this study nurse for 3, 6, and 12 months, and patients completed questionnaires at each visit. Questionnaires were both arthritis-specific (WOMAC) and global (Quality of Well-Being).