Optimal Handling of the Patella in Tourniquet‐Free Total Knee Arthroplasty: Eversion or Lateral Retraction?

OBJECTIVE: To explore the optimal handling of the patella during total knee arthroplasty (TKA) without the intraoperative application of a tourniquet. METHODS: A total of 104 patients undergoing primary unilateral TKA without the intraoperative use of tourniquets from December 2018 to March 2019 in our center were included in this prospective randomized double‐blinded study, including 42 men and 62 women with a mean age of 66.3 ± 7.8 years and a minimum follow‐up of 1 year. Patients were randomly divided into an eversion group (n = 52) and a retraction group (n = 52) based on the intraoperative handling of the patella (eversion or lateral retraction). Primary outcome measures, including the visual analog scale at rest (rVAS) and the visual analog scale in motion (mVAS) for both anterior knee pain and thigh pain, opioid consumption, active range of motion (aROM), passive range of motion (pROM), the time needed for return of the straight‐leg raise (SLR), and 90° knee flexion, were recorded by an independent observer, who also noted secondary outcome measures, including operation time, length of stay (LOS), patella‐related (patellar tilt and baja) and other complications, knee swelling, Hospital for Special Surgery (HSS) scores, and the 12‐item Short Form Health Survey (SF‐12) scores. RESULTS: There were no significant differences between the two groups in terms of baseline parameters. At 24, 48, and 72 h postoperatively (PO), patients in the eversion group experienced more severe thigh pain than those in the retraction group (24 h: 2.6 ± 0.8 vs 2.2 ± 0.5 [P = 0.003]; 48 h: 2.0 ± 0.5 vs 1.8 ± 0.4 [P = 0.026]; 72 h: 1.1 ± 0.4 vs 0.9 ± 0.5 [P = 0.012], respectively) and consumed more opioids (24 h: 22.3 ± 7.7 vs 15.1 ± 8.9 mg [P < 0.001]; 48 h: 27.3 ± 9.3 vs 21.4 ± 10.5 mg [P = 0.003]; 72 h: 23.1 ± 8.2 vs 19.8 ± 7.6 mg [P = 0.036], respectively), but no significant difference was found in anterior knee pain (both rVAS and mVAS) preoperatively or at 24, 48 or 72 h, 3 weeks, 2 or 6 months, or 1 year PO (P > 0.05). Throughout the 1‐year follow‐up, patients in the retraction group showed significantly better function, including greater aROM and pROM at all time points (P < 0.05) and a shorter period of time needed for return to SLR (1.9 ± 0.7 vs 2.2 ± 0.8 days [P = 0.044]) and 90° knee flexion (1.2 ± 0.7 vs 1.9 ± 0.8 days [P < 0.001]). In addition, no between‐group difference was found in operation time, complication rate, swelling of the knee, or SF‐12 score during the follow‐up period (P > 0.05). The eversion group had a significantly longer LOS (5.61 ± 1.92 vs 4.93 ± 1.45 days [P = 0.044]) and worse HSS score within 2 months PO (3 weeks PO: 77.4 ± 7.6 vs 81.7 ± 7.2 [P = 0.004]; 2 months PO: 85.1 ± 7.2 vs 88.5 ± 6.1 [P = 0.011]), but at the third follow‐up (6 months PO), the HSS score in the two groups became comparable (P > 0.05). No patellar tilt occurred in either group and there was only one case of patellar baja in the eversion group, with no significant between‐group difference (P > 0.05). CONCLUSION: We recommend lateral retraction rather than eversion for optimal handling of the patella during TKA because the postoperative functional recovery is better and thigh pain is relatively less severe.