PURPOSE: To investigate whether patient-controlled oral analgesia (PCOA) used by individuals receiving a total knee replacement could reduce pain, increase patient satisfaction, reduce opioid use and/or reduce opioid side effects when compared with traditional nurse (RN)-administered oral analgesia.METHODS: Patients who underwent an elective total knee replacement at a quaternary care centre (Toronto Western Hospital, Toronto, Ontario) were randomly assigned to either PCOA or RN-administered short-acting oral opioids on postoperative day 2. Subjects in the RN group called the RN to receive their prescribed short-acting opioid. Subjects in the PCOA group kept a single dose of their prescribed oral opioid at their bedside and took this dose when they felt they needed it, to a maximum of one dose every 2 h. Study outcomes, collected on postoperative day 2, included pain (measured by the Brief Pain Inventory – Short Form), patient satisfaction (measured by the Pain Outcome Questionnaire Satisfaction sub-scale – component II), opioid use (oral morphine equivalents), opioid side effects (nausea, pruritis and/or constipation) and knee measures (maximum passive knee flexion and pain at maximum passive knee flexion, performed on the operative knee).RESULTS: Study outcomes were analyzed twice. First, for a subset of 73 subjects who remained in their randomly assigned group (PCOA group, n=36; RN group, n=37), randomized analyses were performed. Second, for the larger sample of 88 subjects who were categorized by their actual method of receiving oral opioids (PCOA group, n=41; RN group, n=47), as-treated analyses were performed. There were no differences in study outcomes between the PCOA and RN groups in either analysis.CONCLUSION: PCOA was not superior to RN administration on study outcomes. However, PCOA did not increase opioid use or pain. PCOA remains an important element in the patient-centred care facility.