Double level osteotomy (DLO) for severe genu varum is not a common technique. We performed our first computer-assisted double level osteotomy (CADLO) in March 2001 and we published our preliminary results in 2005 and 2007. The rationale to perform this procedure is to avoid oblique joint line in order to have less difficulty in case of revision to a total knee arthroplasty (TKA). The goal of this paper is to present the results of 37 cases operated on between August 2001 and January 2010.

The series was composed of 35 patients (two bilateral), nine females and 26 males, aged from 39 to 64 years old (mean age: 50.5 +/− 7.5). We operated on 20 right knees and 17 left ones. The mean BMI was 29.3 +/− 4.3 for a mean height of 1.71 m and a mean weight of 85.8 kg. The functional status was evaluated according to the LYSHÖLM and TEGNER score. The mean score was of 42.4 +/− 8.9 points (22–69). According to modified AHLBÄCK criteria we operated on seven stage 2, 22 stage 3, five stage 4 and two stage 5. We measured HKA (Hip-Knee-Ankle) angle using RAMADIER's protocol and we also measured the femoral mechanical axis (FMA) and the tibial mechanical axis (TMA) to pose the right indication. These measures were respectively: 168° +/− 3.4° (159°–172°), 87.5° +/− 2.1 (83°–91°) for the FMA and 83.7° +/− 2.6° (78°–88°) for the TMA.

The inclusion criteria were a patient younger than 65 years old with a severe varus deformity (more than 8° − HKA angle ≤ to 172°) and a FMA at 91° or less. All the osteotomies were navigated using the ORTHOPILOT® device (B-BRAUN-AESCULAP, TUTTLINGEN, GERMANY). The procedure was performed as follows: after inserting the rigid-bodies and calibrating the lower leg, we did first the femoral closing wedge osteotomy (from 4 to 7 mm) which was fixed by a an AO T-Plate, and secondly, after checking the residual varus, the high tibial opening wedge osteotomy using a BIOSORB® wedge (Tricalcium phosphate) and a plate (AO T-plate or C-plate). The goals of the osteotomy were to achieve an HKA angle of 182° +/− 2° and a TMA angle of 90° +/− 2°.

The functional results were evaluated using the LYSHÖLM-TEGNER score and the KOOS score. The patients answered the questionnaire at revision or by phone, and the radiological results were assessed by plain radiographs and standing long leg X-Rays between three and six months postoperatively.

We had no complication in this series but one case of recurrence of the deformity related to an impaction of the femoral osteotomy on the medial side. Two patients were lost to follow-up after removing of the plates (24 months) but were included in the results because the file was complete at that date. All the patients were assessed at a mean follow-up of 43 +/− 27 months (12–108). The mean LYSHÖLM-TEGNER score was 78.7 +/− 7.5 points (59–91) and the mean KOOS score was 94.9 +/− 3.3 points (89–100). Thirty-five patients were satisfied (18) or very satisfied (17) of the result. Only two were poorly satisfied. Regarding the radiological results, if we exclude the patient who had a loss of correction, the goals were reached in 32 cases (89%) for the HKA angle and in 31 cases (86%) for the TMA with only one case at 93°. The mean angles were: 181.97° +/− 1,89° (177°–185°) for HKA, 89.86° +/− 1,85° (85°–93°) for TMA and 93.05° +/− 2.3° (89°–99°) for FMA. At that mid-term follow-up no patient had revision to a total knee arthroplasty.

DLO is a very demanding technique. Navigation can improve the accuracy of the correction compared to non computer-assisted osteotomies. The functional results are satisfying and the satisfaction of the patients is very high. Despite the difficulty of the procedure, complications are, in our hands, very rare. We recommend DLO for severe genu varum deformity in young patients to avoid oblique joint line, which will be difficult to revise to TKA.

Purpose of the study: The risk of homologous blood transfusion for primary total knee arthroplasty (TKA) varies between centres from 4 to 72%. Measurement of mean blood loss for patients undergoing primary total hip or knee arthroplasty led us to adopt a global blood-sparing protocol.

Material and methods: This was a prospective observational study involving 90 TKA (TC-SB) implanted by one surgeon in 56 women and 33 men, mean age 74 years, ASA 3 for 29%, mean BMI=31. Programmed autologous transfusion and blood recovery were not applied. Alpha epoeitine (EPO was administered in the preoperative period in accordance with the predicted serum haemoglobin at discharge. Tranexamic acid (TA) was injected intravenously during the operation.

Results: The mean operative time was 63 minutes, mean tourniquet time 51 minutes. During the perioperatively period, the mean 7-day blood loss was 566±254 ml red cells and 1560±643 total blood. Mean serum haemoglobin at discharge was 1.0 g/dl. Fourteen percent of patients had a postoperative serum haemoglobin less than 10 g/dl (lowest level 8.4 g/dl). One patient had a transfusion; this was a man in his nineties who was operated on without a tourniquet and without TA because of a poor arterial system. Nearly 86% of patients received TA with a significant (p< 0.05) decrease in bleeding compared with those who did not receive TA (365 ml total blood). EPO was delivered for 22% of patients (2.8 injections on average) to reached 14.2 g/dl the day before the operation.

Discussion: This overall strategy for blood sparing enabled us to reduce the risk of transfusion and limit costs. In this series, the prescription of EPO was not targeted to reach a set level preoperatively but was based on the predicted serum haemoglobin level at discharge. Thus seven patients with a preoperative haematocrit below 39 did not receive EPO. None of these patients were transfused. Their discharge haemoglobin was between 10.7 and 12.4g/dl. Use of TA should be included in the calculation of the predicted discharge haemoglobin level in order to better determine candidates for EPO injections.