Manipulation under anaesthesia versus lysis of adhesions for arthrofibrosis of the knee: a 6-month randomized, multicentre, non-inferiority comparative effectiveness protocol.

Abstract

BACKGROUND\nRecent studies show similar outcomes between lysis of adhesions (LOA) and manipulation under anaesthesia (MUA) in patients with arthrofibrosis within 3 to 6 months after arthroscopic knee procedures. As MUA offers positive efficacy with less expense and more convenience, the first consideration in clinical practice shifting to MUA may save much medical cost when MUA is non-inferior to surgery. There is a lack of evidence comparing the clinical outcomes of these two surgical techniques. The purpose of our study is to determine whether MUA is non-inferior to LOA for improving clinical outcomes in patients with loss of flexion within 3 and 6 months after arthroscopic knee surgery.\n\n\nMETHODS\nThis multicentre randomized controlled trial (RCT) will recruit 360 patients within 3 to 6 months after arthroscopic knee surgery with a magnetic resonance imaging (MRI)-confirmed arthrofibrosis to prove non-inferiority of MUA. Patients are randomly assigned to the LOA group (n=180) or the MUA group (n=180) in a 1:1 ratio with random sequences using the SPSS Statistics V.22. The standard surgical procedures will be performed within 2 weeks after randomization, and each patient will receive unified perioperative physical therapy (PT). The primary outcome measure is the range of knee flexion. Secondary outcome measures include physical function, activity level, general health, knee pain, and the occurrence of adverse events. Safety measures include physiological parameters (i.e., blood pressure, pulse, respiratory rate) and complications (i.e., component damage, patellar tendon tear and fracture during surgical procedures, neurological or internal medicine conditions, venous thromboembolism, bleeding and infection postoperatively). The noninferiority margin is defined as a difference of 10 degrees on the range of knee flexion between groups and was assessed with a one-sided α of .025. To investigate relative effectiveness between the groups, linear mixed-effects model will be used for continuous data, and generalised estimating equation for dichotomous data. Analyses will follow the intention-to-treat principle.\n\n\nDISCUSSION\nFindings of our study will help clinicians determine the risk-benefit balance of MUA and LOA, two frequently used surgical procedures for patients with knee arthrofibrosis.\n\n\nTRIAL REGISTRATION\nChinese Clinical Trial Registry (www.chictr.org.cn); ChiCTR2000033467.