Key points

Introduction

Introduction

Methods

Institutional board approval was obtained for this study. The authors retrospectively reviewed the office charts, operating room notes, and medical records of all patients undergoing simultaneous bilateral TKAs by a single surgeon at 2 institutions from April 2011 through February 2012. Twenty-nine patients (58 TKAs) formed the study cohort. The authors included all patients undergoing simultaneous bilateral TKA during this time frame and no patients were excluded.

The senior surgeon, Dr. Cushner started using PSI for TKA in April 2011. The decision to offer PSI versus conventional instrumentation was based entirely on the surgeon’s discretion. Based on whether PSI had been used or not, 2 patient groups were formed ( Table 1 ). The non-PSI group included 14 patients (28 knees). This group comprised 8 women and 6 men, with a mean age of 59 years (Mean [SD] 6.5) and mean body mass index (BMI) of 29 (Mean [SD] 4.5). The PSI group consisted of 15 patients (30 knees) with 9 women and 6 men, having a mean age of 57 years (Mean [SD] 6.6) and mean BMI of 30 (Mean [SD] 5.1). Preoperative diagnosis was osteoarthritis in all knees.

All patients received a similar design of posterior-stabilized cemented TKA by the same manufacturer. All surgeries were performed using either the Legion or Journey II posterior-stabilized TKA standard instrumentation with OrthAlign for tibial cut in non-PSI group system (Smith & Nephew, Memphis, Tennessee) and the Visionaire (Smith & Nephew) PSI was used in PSI group. The latter is an MRI-based PSI system, which, in addition to the MRI, uses long-leg radiographs for preoperative planning.

Patients donated 2 units of autologous blood and were given 3 erythropoetin injections (40,000 IU each) (Procrit, Centocor Ortho Biotech, Horsham, Pennsylvania), 3 weeks, 2 weeks, and 1 week prior to surgery. All patients received oral sustained-release oxycodone (20 mg) (OxyContin, Purdue Pharma, Stamford, Connecticut) and celecoxib (200 mg) (Celebrex, Pfizer, New York, New York) preemptively.

Spinal anesthesia was given to all patients followed by bilateral femoral nerve blocks. A pneumatic tourniquet set to 300 mm Hg to 350 mm Hg (based on body habitus) was used in all cases and was inflated prior to skin incision and deflated after skin closure and application of compression dressing.

A medial parapatellar arthrotomy was used for exposure. In the non-PSI group, a handheld accelerometer-based surgical navigation system (KneeAlign, OrthAlign, Aliso Viejo, California) was used to make the tibial resection first, followed by distal femoral resection using a conventional intramedullary alignment rod. The handheld navigation was not used for femoral resection. The remaining femoral cuts were performed after setting the femoral rotation based on the transepicondylar axis and Whiteside line. The femoral canal was capped in all non-PSI cases with a bone plug made from resected bone. In the PSI group, all femoral cuts were performed first as recommended by the manufacturer and tibial resection was performed last using the fabricated guides. Osteophyte removal and soft tissue releases were performed to balance flexion-extension gaps. Patella resurfacing was performed in all knees. Closure was performed over a deep reinfusion drain, which was clamped for the first 4 hours and removed on postoperative day (POD) 1. Blood was reinfused if the output was more than 100 mL in the first 6 hours after unclamping of drains. Closure was performed using knotless barbed sutures (Quill Self-Retaining System, Angiotech Pharmaceuticals, Vancouver, British Columbia, Canada), using #2 for closure of the arthrotomy, #0 for the subcutaneous layer, and #2-0 for subcuticular closure.

A sterile compression dressing was then applied and compression maintained for 24 hours. Cryotherapy was initiated. Postoperatively, patients received short- and long-acting oral oxycodone, acetaminophen, and celecoxib, with intravenous morphine as needed. Patients were encouraged to get out of bed on POD1 and active and passive knee range-of-motion exercises were initiated. Subcutaneous enoxaparin injections were used for thromboembolism prophylaxis starting 18 to 24 hours after surgery and continued for 2 weeks postoperatively. The criteria for blood transfusion were standardized and included either presence of signs or symptoms of anemia or hypovolemia unresponsive to fluid resuscitation or Hgb concentration less than 7.5 g/dL. Complete blood cell count was performed daily for the first 3 days of inpatient stay and then as needed. Physical therapy was provided daily, starting POD1. Patients were allowed full weight bearing with a walker frame. According to the prespecified home discharge criteria set forth by the 2 hospitals, patients were allowed to go home if they were able to walk 150 feet (46 m), independently transfer from a bed and chair, and go up and down a flight of stairs. Patients were discharged to a skilled nursing facility if they failed to meet the discharge to home criteria. Social work was consulted on POD1 to facilitate eventual discharge.

Tourniquet time was evaluated as a measure of the length of the surgical procedure. Tourniquet time was added for the 2 sides to generate the total tourniquet time. The reduction in Hgb concentration was used as a measure of blood loss and was calculated based on serial postoperative Hgb levels using preoperative Hgb (obtained just prior to surgery) as baseline. Blood transfusion rates were analyzed and compared between groups. Inpatient LOS was assessed in days with the day after surgery counted as the first day and the day of discharge counted as the last day. All thromboembolic episodes were noted up to 3 months postoperatively. Similarly, all early reoperations and complications were noted.

Data were analyzed to determine if there was a significant difference between the 2 groups in demographics, preoperative and postoperative Hgb values, maximum Hgb reduction, average number of allogeneic units transfused in the 2 groups, tourniquet time, LOS, and disposition. Independent t tests were used for comparing continuous data and chi-square test (with Fisher exact test) for comparing categorical data. The level of statistical significance was set at P = .05. Statistical analyses were performed using SPSS, Version 16, software (Chicago, Illinois).