The aim of this prospective study was to compare the analgesic efficacy of three techniques for knee ligamentoplasty using an integration score, the SIA score. Hypothesis. The use of a statistical approach to integrate data from pain scores and morphine consumption will identify four patient subgroups that may hamper the accuracy of the comparative study Material. - We included 165 patients undergoing primary reconstruction of the anterior cruciate ligament of the knee. All patients underwent spinal anaesthesia. The first group, systemic analgesia (SA), received balanced systemic analgesia postoperatively for a minimum of five days, based on Paracetamol, Diclofenac and a morphine PCA. In addition to the systemic analgesia already described, the second group, femoral analgesia (FA), will benefit from a femoral peri-nervous catheter in the crural position. 20 ml of bupivacaine at 0.125% concentration is injected; maintenance is set up immediately with a continuous flow of 8 ml/h for 36 h. The third intra-articular analgesia (IAA) group received, in addition to the same systemic analgesia, an infusion through an epidural catheter of 20 ml of 0.125% bupivacaine, followed by maintenance with 8 ml/h of the same local anesthetic via an electric syringe pump for 36 h. We evaluated cumulative morphine consumption, pain estimation at rest using a numerical scale (EN) from 1 to 10, and an integration of these two parameters. Results. - The mean age was identical for all three groups, at around 30 years. Analgesia in the femoral peri-nerve group was more effective, with a reduction in mean EN compared with the other two groups. Morphine consumption was lower in the AF group. The lowest mean rank was in the AF group, followed by the AIA group, and the difference was statistically significant with an H = 6.89 and a p= 0.032. The difference was significant between the AS group and the AF group (p = 0.09); the other inter-group differences were not significant. 23.2% of patients had an AIS score between [-100, -200], and were in little pain and consumed little morphine (effective treatment); 14.2% of patients had an AIS score between [100, 200] and were in great pain and consumed a lot of morphine (morphine-resistant or very sensitive to pain). For the difference (ranked pain score - ranked morphine), 4.5% of cases had values between [- 100, - 200] and were therefore not very painful but consumed a lot of morphine (propensity), while 3.9% had values between [100, 200] and were therefore very painful but consumed little morphine (intolerant). Discussion. - This study suggests that the use of a statistical approach to integrating data from pain scores and morphine consumption, such as the Silverman Integrating Approach (SIA), will yield four patient subgroups that may hamper the accuracy of the comparative study, including the morphine-resistant subgroup and the propensity subgroup. Comparison of the means of the therapeutic efficacy group will eventually be more precise. Continuous femoral block really does provide added analgesia and morphine sparing, and should therefore be preferred wherever possible to intra-articular or systemic therapy.