Is Manipulation Under Anesthesia Effective in Improving Patient Reported Outcomes After Total Knee Arthroplasty? A Matched Cohort Analysis.

Abstract

Introduction: Manipulation under anesthesia (MUA) after total knee arthroplasty (TKA) is considered effective for postoperative stiffness, but strong scientific justification is lacking. This study compared outcomes in two matched cohorts: patients who met criteria and underwent MUA and patients who met criteria but did not undergo MUA. Methods: MUA (experimental) cases had ≤ 90° flexion 4-weeks postoperatively and underwent MUA surgery within 12 weeks of the index TKA. Control cases had ≤ 90° flexion 4-weeks postoperatively and did not undergo MUA. The latter group was alternatively treated with aggressive flexion exercises, frequent follow-up, and pain control modalities per surgeon discretion. 42 MUAs performed by three surgeons between 2011 and 2017 at the same center using the same standardized clinical and rehabilitation protocols were retrospectively reviewed. Six MUAs were excluded for potential confounds. The remaining 36 MUA cases were matched one-to one on sex (p= 1.00), age (p=0.893), race (p=0.938), BMI (p=0.069), and implant manufacturer (p= 1.00) to 36 control cases. Outcome variables included amount of improvement in flexion from preoperative baseline to latest follow-up and standardized PROMS. Covariates potentially affecting outcomes were taken into account. Results: Overall MUA incidence during the time period was 1.9%. Experimental and control groups did not differ on preoperative fibromyalgia, depression, and narcotic use; or intraoperative analgesia (p≥0.084). Four control patients and no experimental patients had inflammatory disease (p=0.054), and six of the former compared to none of the latter had lumbar spine pain or disease (p=0.025). Flexion data are provided in Table 1. Mean pre-primary TKA flexion was significantly greater in experimental patients (112.4 vs. 98.6°, p=0.002). On average, between pre-primary surgery and latest follow-up, experimental patients lost 10.7° of flexion compared to a gain of 12.7° by control patients (p<0.001). Pre-primary to latest follow-up improvement in pain walking on level ground (-3.3 and -4.7 points, p=0.190) and climbing stairs (-3.7 vs. -5.1, p=0.192) did not significantly differ between experimental and control patients respectively. As shown in Figure 1, prior to primary surgery experimental and control patients had similar activity levels (p=0.624). At latest follow-up, however, control patients were significantly more active than experimental patients (Figure 1, p=0.009). Figure 2 shows the proportions of patients in each group who reported their knee never feels normal prior to primary surgery (p=0.580) and at latest follow-up (p=0.0004). Surgery significantly improved this metric for control but not experimental patients. At latest follow-up 88.6% of control patients and 50% of experimental patients were satisfied or very satisfied with their knee surgery (p=0.001). Conclusion: Patients with ≤ 90° flexion 4-weeks after TKA who underwent MUA had significantly worse flexion and PROM scores than matched control patients who did not undergo MUA. These findings question the effectiveness of MUA as a legitimate treatment for postoperative TKA stiffness.