Treatment of osteoarthritis of the knee remains a challenging problem since the evolution of the disease may be different in each compartment of the knee, as well as the state of the ligaments. Total knee arthroplasty may provide a reliable long-lasting option but do not preserve the bone stock. In another hand, compartmental arthroplasty is a bone and ligament sparing solution to manage limited osteoarthritis of the knee affecting the medial, lateral or the patello-femoral compartment.1, 2, 3

Patient's selection and surgical indication are based on the physical examination and on the radiological analysis including full-length x-rays and stress x-rays. Clinical experience has shown the need for high flexion in patients who have both high flexibility and a desire to perform deep flexion.

Additionally the shape differences related to anatomy or the patient expectations after the surgery may also affect the surgeon decision. 4

The limited incision into the extensor mechanism allows a quicker recovery which represents a functional improvement for the patient additionally to the cosmetic result. A dedicated physiotherapy starting on the following day allowing weight bearing exercises protected by crutches and focusing on early mobilization and range of motion combined to a multimodal pain management approach is critical despite the type of individualized solution chosen for the patient knee. 5

Since bony landmarks may be different form a patient to another one as well as anatomical shapes, several tools have been developed in order to provide the surgeons an assisted tool during the surgery adapted to each knee, this include navigation, patient specific instrumentation and robotic surgery.

Improving the adaptation between the implant and the patient bone during total hip arthroplasty (THA) may improve the survival of the implant. This requires a perfect understanding of the tridimensional characteristics of the patient hip. The perfect evaluation of the tridimensional anatomy of the patient hip can be done pre-operatively using X-rays and CT-scan. All patients underwent a standard x-rays evaluation in the same center according to the same protocol. Pre-operatively, the frontal analysis of the hip geometry was performed and the optimal center of rotation, CCD angle, neck length and lever arm was analyzed to choose the optimal solution for proper balance of the hip in order to obtain adequate range of motion, appropriate leg length, and correct tension of the abductors muscles. Standard or lateralized monoblock stems can be valid or modular neck shape can be choosen among 9 available shape. These 9 frontal shapes are available in standard, anteverted or retroverted shapes, leading to 27 potential neck combinations. In case of important hip deformation, a custom implant can be used in order to balance the extra-medullar geometry without compromising the intra-medullary adaptation of the stem.

We prospectively included 209 hips treated in our institution with total hip arthroplasty performed using a supine Watson-Jones approach and the same anatomic stem. The mean patient age was 68 years and the mean BMI 26 Kg/m². Intra-operatively the sagittal anatomy of the hip was analyzed and standard, ante or retro modular necks were tested for the frontal shape defined pre-operatively.

According to the pre-operative frontal planning, non-standard necks were required in 24 % of the cases to restore the anatomy of the hip. Intra-operatively, a sagittal correction using anteverted neck was required in 5% of the cases and retroverted necks in 18% of the cases. Harris hip score improved from 56 to 95 points at min. 5 year follow-up. No leg length discrepancy greater than 1 cm was observed. Restoration of the lever arm (mean 39.3 mm, range 30 to 49 mm) and of the neck length (55.2, range 43 to 68 mm) was adapted for 95% compared to the non operate opposite side. Disturbed anatomy like in DDH or post-traumatic cases may require additional solutions to balance the hip such combined osteotomy or customized stem and neck.

Introduction:

Recent studies have concluded that gender influences hip morphology at the time of surgery as well as dysplastic development of the hip. This may lead to a particular choice of implant including stem design and/or neck modularity. In this study we hypothesized that not only gender but also morphotype and etiology (primary osteoarthritis versus aseptic osteonecrosis) may be a significant factor to predict the anatomy of the hip at the time of total hip arthroplasty (THA).

Treatment of limited osteoarthritis of the knee remains a challenging problem. Total knee arthroplasty may provide a reliable long-lasting option but do not preserve the bone stock. In another hand, compartmental arthroplasty with or without osteotomy is a bone and ligament sparing solution to manage limited osteoarthritis of the knee. Considering the renewed interest for combined compartmental implants we aimed to evaluate the average 12-year clinical and radiological outcome of a consecutive series of patients treated with compartmental knee arthroplasty combined or not with osteotomy.

We retrospectively reviewed all 255 patients (274 knees) treated in our institution with a compartmental arthroplasty combined or not with an osteotomy for a diagnosis of either bi or tricompartmental osteoarthritis of the knee between April 1972 and December 2000. The series included: 100 cases of combined lateral and medial UKA, 77 combined medial UKA and patello-femoral arthroplasty (PFA), 19 cases of combined Bi-UKA and PFA, 14 cases of UKA and high tibial osteotomy (HTO), 7 cases of combined lateral-UKA and PFA and HTO, 16 cases of combined lateral-UKA and PFA and 13 cases of combined bi-UKA and HTO.

Patient’s selection and surgical indication was based on the physical exam and on the radiological analysis including full-length x-rays and stress x-rays. Clinical and radiological evaluations were performed at a minimum follow-up of 5 years (mean, 12 years; range, 5–23 years) by an independent observer.

The Knee Society knee and function scores improved respectively from 43 to 89 and from 47 to 90 at last-follow-up. The mean active knee flexion improved from 116° ± 6° (range, 100°–145°) pre-operatively to 129° ± 5° (range, 117°–149°) at final follow-up. The restoration of the mechanical axis of the knee was achieved in all the cases. Dramatic failures were observed for patient with uncemented PFA. Considering revision for any reason as the endpoint, the 17-years survivorship was 0.68 (95% confidence interval: 0.62 to 0.75).

Our results suggested that combined compartmental arthroplasty with or without osteotomy can restore function and alignment of the knee in compartmental arthritis. This combined surgery represents a bone and ligament sparing alternative to TKA which can be considerate as a true minimally invasive solution.

Purpose of the study: It has been demonstrated that the anatomy of the distal femur differs by gender. The ratio of the mediolateral/anteroposterior distance, the shape of the distal femur, and the orientation of the trochlea differ between males and females. To adapt to these differences, prostheses specifically designed for female patients (TKAgender) were developed. The purpose of our study was to compare the first objective and subjective outcomes with these prostheses.

Material and methods: Thirty women who underwent bilateral surgery within a six month interval for degenerative joint disease of the knee between March 2006 and March 2008 were included in a comparative prospective study. The side receiving the gender implant was determined at random. Operative and postoperative protocols were the same excepting the femoral implant. Patients were not informed of which knee had received the gender prosthesis. Mean age in this series was 67 years and mean BMI 26. All implants were cemented. At minimum one year follow-up, objective and subjective analysis included specific questions concerning preference and quality of life presented by an independent observer.

Results: The Knee Society clinical scores were comparable in the two groups, as were the results for the different items of the KOOS score. For preferences: the patients preferred the gender knee in 75% of the cases (p< 0.01), they reported less noise or cracking sounds in the anterior part of the knee for 68% (p=0.03) and had the impression that the knee recovered faster in 64% (p=0.04).

Discussion: To our knowledge, this is the first report of results concerning implants specifically designed for female patients. At short-term, the only difference in the patients’ preference was a subjective feeling of less impairment for the patellar track. It will be interesting to follow these patients to assess the long-term impact on the patellofemoral articulation.

Purpose of the study: Besides enabling a precise calculation of the needs for blood (in ml) for each type of orthopaedic surgery, an adequate estimation of average total blood loss (TBL) as a function of total blood volume (TBV) enables initiation of blood sparing techniques. Thus, when the balance is negative, erythropoietin could be advisable. The purpose of this work was to analyse the impact of prescribing erythropoietin as a function of the expected blood balance on the rate of homologous blood transfusions. We also wanted to determine the financial impact of this prescription strategy.

Material and methods: This prospective study included 229 patients (153 THA and 76 TKA) who underwent surgery from January 2005 to December 2008 with a standard anaesthesia, analgesia and surgical protocol. TBL was considered to be 10% of TBV for THA and TKA. Mean gain in serum haemoglobin (Hb) per injection of erythropoietin (1 ampoule Eprex) was 0.8 g/dl. Using the patient’s Hb level at the preop exam (Hbpreop), the anaesthetist determined the volume available (VA) for each patient and for each type of operation. When VA was negative, the Hb needed to reach the desired level (HbA) so that VA=0 (zero risk of transfusion) was calculated. The number of Eprex ampoules prescribed (amp) was determined from the formula: (HbA-Hbpreop)/0.8. Hb levels on day 0, day 1, day 3(discharge) and day 28 were noted, as were the number of homologous transfusions and the total number of ampoules of Eprex delivered.

Results: Sixty-six patient received one amp (33.2%), 96 had two amp (42%), 40 three amp (17.5%) and 17 four amp (7.4%). For 220 patients (92%), Eprex delivered was within the approved dose (92%). The mean Hb levels were: Hbpreop=12.4, HbA=13.5, Hbd0=13.9, Hbdischarge=11.2, and Hbd28=11.9. Total consumption was 456 amp instead of 916 (229x4), i.e. a savings of 460 amp corresponding to 173 880 euros (460 x 378 euros). There were six homologous transfusions (2.6%).

Discussion: In orthopaedic surgery, adapted prescription of erythropoietin as a function of the TBL enabled, in this series, a simplification of the prescription procedure and avoided the risks related to high Hb levels > 15g with a financial savings of 760 euros per operation (173880 euros for 229 patients) without increasing the rate of homologous transfusion. This strategy is currently being evaluated for prosthesis revision surgery.

Improving the adaptation between the implant and the patient bone during total hip arthroplasty (THA) may improve the survival of the implant. This requires a perfect understanding of the tridimensional characteristics of the patient hip. The perfect evaluation of the tridimensional anatomy of the patient hip can be done pre-operatively using CT-scan and in case of important hip deformation, a custom implant can be used. When this solution is not available, modular necks may be a reliable alternative using standard x-rays and intraoperative adaptation. We aimed to evaluate:

The usefulness of modular neck to restore the anatomy of the hip and

We prospectively included 209 hips treated in our institution with a modular neck total hip arthroplasty between January 2006 and December 2007. All patients underwent a standard xrays evaluation in the same center according to the same protocol. Pre-operatively, the frontal analysis of the hip geometry was performed and the optimal center of rotation, CCD angle, neck length and lever arm was analyzed to choose the optimal modular neck shape among 9 available shape. These 9 frontal shapes are available in standard, anteverted or retroverted shapes, leading to 27 potential neck combinations. The mean patient age was 68 years and the mean BMI 26 Kg/m² All the procedures were performed supine using a Watson-Jones approach and the same anatomic stem. Intra-operatively the sagittal anatomy of the hip was analyzed and a standard, ante or retro modular necks were tested for the frontal shape defined pre-operatively.

According to the pre-operative frontal planning, nonstandard necks were required in 24 % of the cases to restore the anatomy of the hip. Intra-operatively, a sagittal correction using anteverted neck was required in 5% of the cases and retroverted necks in 18% of the cases.

Harris hip score improved from 56 to 95 points at last follow-up. No leg length discrepancy greater than 1 cm was observed. Restoration of the lever arm (mean 39.3 mm, range 30 to 49 mm) and of the neck length (55.2, range 43 to 68 mm) was adapted for 95% compared to the non operate opposite side. No loosening was observed.

According to our results modular neck combined are useful and reliable to restore optimal hip geometry and in this series 25% of the patient would have had imperfect extra-medullary hip geometry with a standard prosthesis. The good clinical and radiological short-term results should be confirmed at longer follow-up.

Unicompartmental knee arthroplasty (UKA) is a logic procedure when osteoarthritis or avascular necrosis is limitad to one femorotibial compartment. The indications for the procedure includes osteoarthrosis or osteonecrosis with full-thickness loss of articular cartilage limited to one of the tibiofemoral knee compartments. Physical examination should ensure full range of knee motion. Frontal and sagittal knee stability has to be tested. A particular attention should be given to the state of the anterior cruciate ligament. The status of the patellofemoral joint should be analysed by physical examination and patellofemoral view at 30, 60 and 90° of flexion. Preoperative anteroposterior varus and valgus stress radiographs should be done to confirm the complete loss of articular cartilage in the involved compartment, the full thickness cartilage in the opposite compartment and the possibility of full correction of the deformity to neutral.

The so-called minimally invasive surgery (MIS) procedure using a specific instrumentation is able to provide quicker recovery since the extensor mechanism disruption is eliminated. More importantly the radiological evaluation has shown that precise implantation of the components is possible with an MIS approach which is important for the long term results of the arthroplasty. The clinical results at ten years of follow-up of cemented metal-backed UKA performed through a conventional approach have shown results comparable to those obtained with total knee arthroplasty. The in vivo kinematic evaluation of patients implanted with UKA has shown that kinematics similar to the normal knee can be obtained, enhancing the importance of a functional anterior cruciate ligament.

Recent design improvements have increased the femorotibial area of contact to accommodate high flexion angles. Additionally our experience has demonstrated that modern UKA is a valid alternative for young and active patients with unicompartmental tibiofemoral noninflammatory disease, including both osteoarthritis and avascular necrosis. Compared to medial UKA lateral UKA represents in our experience only 5% of all UKA implantations.

However the long term results of lateral UKA compares at least equally with those reported for medial UKA.

Introduction: Women have gender specific shape of the distal femur. To fit these gender characteristics, gender specific femoral implants were developed for total knee arthroplasty (TKA). We aimed to compare

objective and subjective functional improvement;

Materials and Methods: 30 women (60 knees) operated on successively (6 months in between) for a bilateral TKA between March 2006 and March 2008 by the same surgeon were included in this prospective study. The same surgical protocol and the same post-operative management protocol were applied for both sides. Mean age was 67±3 and mean BMI 26±4 Kg/m². At a minimum follow-up of one year, evaluation objective and subjective functional improvement, patient satisfaction and preference and cost-utility analysis were performed double blind.

Results: Knee Society knee score and Knee Osteoarthritis Outcome Score (KOOS) improvements were comparable in both groups. However, 75% of the women preferred their gender TKA (p< 0.001). 68% of the women described less crepitus or anterior knee bothering after gender TKA (p=0.003) and 64% had faster recovery with the gender implant (p< 0.001). The cost-utility analysis was favorable for the gender knee.

Discussion: No objective or subjective superiority in terms of functional improvement was shown with gender specific implants at this short-term follow-up. However significant differences in terms of patient satisfaction and preference and a favorable cost-utility analysis were observed. These results should now be confirmed at longer-follow-up. Despite comparative functional improvement, patient satisfaction and preference were higher for the side implanted with a gender specific TKA in this prospective comparative study.

Introduction and Method: For 16 years, now, we have been using custom made femoral stems (titanium stem, HA coated) based on preoperative patient CT scans in young patients presenting with symptomatic osteoarthritis of the hip. The aim was to provide optimum initial fit-and-fill of the stem in the femoral medullary canal, conferring the best chance of secondary osteoin-tegration. This, with a goal of long term survivorship. The custom stem also enables dialling-in correction of the (often abnormal) femoral neck version (in young patients presenting with arthritis) to a more normalised 15 degrees of anteversion.

Results: We present the long term results (5–16 year, mean of 10 years), clinical, survivorship and radiological, of 312 primary total hip arthroplasties in 280 patients, all of whom were under the age of 50 years of age (mean age 40).

At 10 years we have a survivorship, if femoral aseptic loosening is used as an end point, of 97.6%. There was a deep infection rate of 1.2%, and a dislocation rate of 1.9%. There were no cases of thigh pain, and no intra-operative femoral neck/shaft fractures.

Discussion: The under 50’s with hip arthritis requiring total hip arthroplasty are a highly challenging group, they are young, active and tend to have distorted anatomy. Our results are superior to those previously published using either cemented or standard uncemented stems. Thus, justifying the increased initial financial outlay on the custom-made stem.

Introduction: Due to the diversity of the prosthetic implants for hip arthroplasty, a better matching between the implant and the intra and extra-medullary characteristic of the patient anatomy is now possible. This adaptation however requires a perfect understanding of the tridimensional characteristics of the patient hip anatomy. Little data are available in the literature. We aimed to analyze the muscular and bony anatomy of the hip at the time of arthroplasty.

Material and Methods: data acquisition was performed according a standardized CT-scan in the same center for all patients within the standard workup before arthroplasty. Standardized measurements were obtained after automatic tridimensional bone reconstructions using dedicated software. 549 femurs in 469 patients including primitive coxarthrosis (COX), dysplasic development of the hip (DDH) and aseptic osteonecrosis (ONA) were analyzed. Mean age was 58 and 70% of the patients were women. Tridimensional reconstruction of the muscular anatomy of the hip was performed for 30 patients using manual contouring on dedicated software. Characteristics of the bony and muscular anatomy were then analyzed according to the etiology and correlations between bony and muscular anatomy were evaluated.

Results: Concerning the bone analysis, mean offset was 23.2 ±1.5mm in the DDH group, 40.5±1.2 mm in the COX group and 29.6± 0.9mm in the ONA group(p< 0.001). Neck-shaft angle was 132±25º in the DDH group, 130±0.5º in the COX group and 134±1º in the ONA group (p< 0.001). Mean anteversion was 33±3.5º in the DDH group, 25±3.8º in the COX group and 16±3.2º in the ONA group (p< 0.001). Concerning the muscular analysis, gluteus medius and minimus volumes were correlated with the body mass index and with the gender, but not with patient age (p: NS). Location of the muscular insertion of the gluteus medius and minimus on the greater trochanter were correlated with the femoral anteversion.

Discussion: The results of our study demonstrated that bony and muscular anatomical characteristics were correlated with the etiology of the degenerative joint disease, with the patient body mass index and gender. Surgeons should be aware of these characteristics to improve the patient anatomy reconstruction during the arthroplasty.

Introduction: The data reporting clinical and radiological outcomes after modern unicompartmental knee arthroplasty (UKA) for spontaneous avascular osteonecrosis are limited. We hypothesized that UKA for spontaneous osteonecrosis may be as reliable and durable as it is for osteoarthritis.

Materials and Methods: We retrospectively reviewed 40 cemented UKA operated for spontaneous osteonecrosis of the knee in two different centers between 1989 and 2004. Twenty-six patients were women and 14 men, mean patient age was 67 years (range, 45 to 84) and mean body mass index was 27.4 Kg/m2 (range, 18 to 44). Clinical and radiological evaluations were performed by an independent observer in each center according to the same protocol at a minimum follow-up of 3 years (mean 8 years; range, 3–17.5 years).

Results: The mean Knee Society Knee score improved from 60 preoperatively to 95 post-operatively. The mean Knee Society Function score improved from 50 preoperatively to 89 post-operatively. Restoration of an appropriate lower-limb mechanical axis was achieved for 36 knees (90%). Two knees were revised for aseptic loosening. The Kaplan–Meier survivorship was 95% at 12 years.

Discusssion and conclusion: Our data suggest UKA is a reasonable solution for restoring clinical function and radiological lower-limb alignment for spontaneous osteonecrosis of the knee, with a durable survivorship.

A clinical and radiographic study was conducted on 97 total hip arthroplasties (79 patients) performed for congenital hip dislocation using three-dimensional custom cementless stem. The mean age was 48 years (17 to 72). The mean follow up was 123 months (83 to 182).

According to Crowe, there were 37 class 1, 28 class 2, 13 class 3 and 19 class 4. The average lengthening was 25 mm (5 to 58 mm), the mean femoral anteversion 38.6° (2° to 86°) and the correction in the prosthetic neck −23.6° (71° to 13°). The average Harris hip score improved from 58 to 93 points. Six hips (6.2%) required a revision. The survival rate was 97.7% ± 0.3% at 13 years.

Custom cementless stem allows anatomical reconstruction and good functional results in a young and active population with disturbed anatomy, while avoiding a femoral osteotomy.

We retrospectively reviewed 35 cemented unicompartmental knee replacements performed for medial unicompartmental osteoarthritis of the knee in 31 patients ≤50 years old (mean 46, 31 to 49). Patients were assessed clinically and radiologically using the Knee Society scores at a mean follow-up of 9.7 years (5 to 16) and survival at 12 years was calculated. The mean Knee Society Function Score improved from 54 points (25 to 64) pre-operatively to 89 (80 to 100) post-operatively (p < 0.0001). Six knees required revision, four for polyethylene wear treated with an isolated exchange of the tibial insert, one for aseptic loosening and one for progression of osteoarthritis.

The 12-year survival according to Kaplan-Meier was 80.6% with revision for any reason as the endpoint. Despite encouraging clinical results, polyethylene wear remains a major concern affecting the survival of unicompartmental knee replacement in patients younger than 50.

Purpose of the study: Malposition of the acetabular implant of a total hip arthroplasty can provoke dislocation, limited joint movement, and early wear. The purpose of this prospective randomized study was to assess the efficacy of a image-free navigation system to achieve correct acetabular position for total hip arthroplasty.

Material and methods: The navigation software for the acetabular component used intraoperative anatomic acquisition. The prospective randomized study included two groups of 20 patients. In the first group, the acetabular implant was inserted using the computer-assisted system and in the second using the conventional method. The same operator performed all procedures via an anterolateral approach and using the same non-cemented hemispheric cup. The postoperative assessment was performed by an independent investigator who noted the cup inclination on the standard ap view and cup anteversion on the computed tomography; then using a dedicated system for 3D reconstruction, the same values were determined one month after surgery.

Results: Each group included ten men and ten women, mean age 63 years and mean body mass index 24. For the computed-assisted group, mean additional operative time for navigation was 13 minutes (range 8–20 min). The intraopeartive concordance with the surgeons subjective impression was excellent for 12 patients and good for 8. Mean intraoperative values were for the computer-assisted group were: for operative mode: inclination 30° (25–46°), anteversion 14° (0–25°), radiographic mode: inclination 35° (25–47°), anteversion 13° (0–26°), anatomic mode: inclination 36°, anteversion 19° (0–27°). There was no significant difference between the intraoperative and postoperative values for the computer-assisted group. There was no significant difference between the average values between the computer-assisted and conventional groups but the standard deviation was smaller in the computer-assisted group.

Discussion and conclusion: The image-free navigation system enables reliable positioning of the prosthetic cup for total hip arthroplasty and increases the precision of acetabular implantation without increasing significantly operative time. This first step must be integrated into the computerized preoperative planning for total hip arthroplasty. The next step will be to use the navigation system for implanting the femoral component.

Purpose of the study: Acetabular dysplasia is a recognized cause of premature hip degeneration. With increasing use of arthroplasty, the role of conservative treatment can be debated. The purpose of this work was to describe technical advances achieved with Ganz triple periacetabular osteotomy and evaluate long-term results.

Material and methods: This study included 32 dysplastic hips in 28 patients treated by Ganz triple osteotomy and assessed a mean 12 years follow-up (range 2 – 20 years). Mean age was 32 years (range 18–47). There were 24 women and four men. Hip joint measurements were made on preoperaive standard x-rays with complementary recentered views if needed as well as computed tomography (CT) to better distinguish progressive degeneration. For early patients, the iniail osteotomy involved three cuts (ilioischial, iliopubic, ilial) starting close to the acetabulum and performed via three approaches: sub coxofemoral, intrapelvic, extrapelvic. The first technical change involved osteotomy of the anterosuperior iliac spine and an oblique iliac cut farther from the acetabulum.

Results: Mean preoperative angles were: 135° (121 to 150°) for CC’D, 23.2° (3 to 40°) for HTE, 8.4° (−14 to 22°) VCE, 11.3° (−26 to 32°) for VCA. The postoperative values were: 134.5° (121 to 150°) for CC’D, 9.5° (−9 to 20°) for HTE, 31.7° (14 to 60°) for VCE, 31.7° (10 to 48°) for VCA. Six patients required total hip arthroplasty on average four years later (range 2 – 9 years), including one patient with aseptic necrosis of the acetabulum.

Discussion and Conclusion: This study confirms the usefulness of triple periacetabular osteotomy for conservative treatment of acetabular dysplasia. In light of our results, the following changes have been instituted:

all three cuts are performed via a single intra-pelvic approach;

Purpose of the study: Changes in prosthetic design to adapt to knee flexion greater than 120 degrees can modify the bone-prosthesis fixation and also displace the femorotibial contact. The purpose of this study was to analyze mid-term results in a consecutive series of 186 arthroplasties and to examine the femorotibial kinematics in vivo.

Material and methods: A posterior stabilized cemented prosthesis with a plateau with motion limited to rotation was used. Design changes concerned: lengthening of the posterior femoral condyle, scooping out the poly-ethylene anteriorly with reorientation and change in the height of the posterior stabilization stem. The same technique was used for all patients who followed the same rehabilitation protocol. Mean age was 69 years (range 22–87). All patients were evaluated clinically with the IKS score and radiologically on the anterioposterior and lateral images. An in vivo analysis of the femorotibial kinematics in the weight bearing condition was also performed in 20 patients under fluoroscopic control with automatic 3D modelization.

Results: Mean follow-up was 40 months (range 2–5 years). Mean IKS function score improved from 34 preoperatively to 96 at last follow-up. The knee score improved from 53 on average to 91 at last follow-up. The mean flexion was 115° (range 45–135°) preop-eratively and 120° (115–145°) at last follow-up. One implant was removed for infection and arthrolysis was performed for one case of stiff joint. Radiographically: the mean postoperative femorotibial alignment was 179° (178–181°), the mean tibial slope 3.8° (0–10°°, the mean patellar height (0.8° (0.56–1°), and the mean elevation of the joint space (4.5 mm. There were two cases of progressive lucent lines in the tibial zone which were stable at last follow-up. All patients analyzed showed a mean posterior displacement of the femorotibial point of contact of 9.7 mm at flexion.

Discussion and conclusion: Changes in prosthesis design to adapt to greater range of flexion do not appear to have a negative effect at mid-term on implant fixation. The clinical flexion ranges obtained were encourageing and the correlation with kinematic results show that the degree of preoperative flexion remains a determining factor for the postoperative outcome. Posterior displacement of the femoro-tibial point of contact, observed in all patients examined fluoroscopically, certainly contributed to the good postoperative flexion.

Purpose of the study: Compter-assisted surgery enables improved precision of prosthetic implantations, but the basis of data acquisition remains variable. The purpose of this study was to assess the radiological quality of a total knee arthroplasty (TKA) implanted with a computer-assisted surgical technique with or without pre-operative imaging.

Material and methods: This was a case-control study of a group of 40 patients who underwent TKA implanted with a navigation systm (N+) which was compared with a control group of patients who underwent the same procedure with a conventional technique (N-). The two groups were comparable for: age, gender, BMI, preoperative HKA. The same surgeon operated all patients using the same cemented posterior stabilized TKA. Outcome was analyzed by an independent operator. The same navigation system was used for all knees, with, for the first 20 knees, acquisition based on preoperative computed tomography and for the next 20 knee, intra-operative acquisition. Postoperatively, six radiographic parameters were studied for each knee on the ap and lateral views. An optimal interval was determined for each parameter and the number of optimal criteria was noted for each knee.

Results: The mean HKA was 177.5° in the N- group and 179.2° in the N+ group. The angle of implantation of the femoral piece was 90.3° in the N- group and 90° in the n+ group. The mean posterior tibial slope was 3.5° in the N+ group and 3.1° in the N- group. There was a significant difference for the tibial prosthetic angle in favor of the N+ group, i.e. 89° compared with 87.3° for the N- group. The overall quality of the implantation was considered optimal for 54.5% of knees in the N+ group and for 29.8% in the N- group. There was no significant difference between computed tomographic acquisition and intraoperative acquisition.

Discussion and conclusion: This study demonstrates that the results exhibit a distribution closer to the ideal values for the navigation group but that the difference is solely significant for the tibial implantation. This improvement requires a longer operative time of 18 minutes. The lack of any difference between the computed tomographic acquisition and the intraoperative acquisition suggests that intraoperative acquisition should be favored for reasons of cost and simplicity. Computed tomography imaging can still be useful for a precision of the biepicondylar line in certain complex situations such as revision arthroplasty.

A clinical and radiological study was conducted on 97 total hip replacements performed for congenital hip dislocation in 79 patients between 1989 and 1998 using a three-dimensional custom-made cementless stem. The mean age at operation was 48 years (17 to 72) and the mean follow-up was for 123 months (83 to 182).

According to the Crowe classification, there were 37 class I, 28 class II, 13 class III and 19 class IV hips. The mean leg lengthening was 25 mm (5 to 58), the mean pre-operative femoral anteversion was 38.6° (2° to 86°) and the mean correction in the prosthetic neck was −23.6° (−71° to 13°). The mean Harris hip score improved from 58 (15 to 84) to 93 (40 to 100) points. A revision was required in six hips (6.2%). The overall survival rate was 89.5% (95% confidence interval 89.2 to 89.8) at 13 years when two hips were at risk.

This custom-made cementless femoral component, which can be accommodated in the abnormal proximal femur and will correct the anteversion and frontal offset, provided good results without recourse to proximal femoral corrective osteotomy.

Introduction: Deep flexion may affect both femorotibial contact pattern and patellofemoral interface. The objective of this study was to conduct the first in vivo kinematic analysis that determines the 3D motions of the femorotibial and patellofemoral joints, simultaneously from full extension into deep flexion.

Methods: Three-dimensional femorotibial and patello-femoral kinematics were evaluated during a deep knee bend using fluoroscopy for five subjects having a normal knee, five having an ACL-deficient knee and 20 subjects having a TKA designed for deep flexion.

Results: The average weight-bearing range-of-motion was 125 degrees, significantly higher than in previous studies. On average, subjects experienced 4.9o of normal axial rotation and only three subjects experienced an opposite rotation pattern. On average, subjects experienced −9.7 mm of posterior femoral rollback (PFR) and all subjects experienced at least −4.4 mm of PFR. These subjects experienced less patellofemoral translation than the normal knee, but the average motion was similar in pattern to the normal knee. On average, the subjects having a TKA experienced patella tilt angles that were similar to the normal knee.

Discussion: It is assumed that femorotibial kinematics can play a major role in patellofemoral kinematics. Altering the patella motion and/or the patellar ligament rotation could lead to much higher forces at the patel-lofemoral interface. In this study, these subjects experienced kinematic patterns that were very similar to the normal knee and it can be deducted that forces acting on the patella were not significantly increased for TKA subjects compared with the normal subjects.

Using radiography and computer tomography (CT) we studied the morphology of 83 hips in 69 Caucasian adults with osteoarthritis secondary to developmental dysplasia of the hip (DDH). A previously published series of 310 hips with primary osteoarthritis was used as a control group. According to the Crowe classification, 33 of the dysplastic hips were graded as class I, 27 as class II and 23 as class III or class IV.

The intramedullary femoral canal had reduced mediolateral and anteroposterior dimensions in all groups compared with the control group. Only in Crowe class II hips was the femoral neck-shaft angle increased. The proximal femur had more anteversion in all the developmental dysplasia of the hip groups, ranging from 2° to 80°. Templated measurement of acetabular dimensions for plain radiography closely matched measurements taken by CT.

The results of our study confirm the observations previously confined to the Japanese population.

Purpose: Most revisions of total knee arthroplasty (TKA) occur before the fifth year. The leading cause is prosthesis malalignment. Computer-assisted surgery is presented as a technique which improves implantation precision. The purpose of this study was to evaluate the radiographic quality of computer-assisted TKA implantation in comparison with conventional implantation.

Material and methods: A prospective randomised study was conducted with a total of 82 patients assigned randomly to conventional instrumentation (group 1) or computer-assisted implantation (group 2). The two groups were comparable for age, gender, body mass index, side, and preoperative femorotibial angle (HKA). The same surgeon operated all patients using the same cemented posterostabilised prosthesis. Radiographically, the HKA, the theta angle between the mechanical and anatomic femoral axis, the angle of the femoral and tibial implantations, and the posterior tibial slope were noted. All measures were taken by an independent operator who was unaware of the operative technique.

Results: There was no statistical difference in the HKA 177.5° in group 1 and 179.2° in group 2 (p=0.13); the theta angle was 6° in group 1 and 5.9° in group 2 (p=0.78); the femoral implantation angle was 90.3° in group 1 and 90° in group 2 (p=0.74); the posterior tibial slope was 3.5° in group 1 and 3.15° in group 2 (p=0.65). There was a statistical difference in the tibial angle, 87.3° in group 1 and 89° in group 2 (p=0.012).

Discussion: This study demonstrates a significant improvement in the navigation group for the position of the tibial implant and no difference for the femoral implant. This might be related to the cut guide which allowed more intra-operative adjustment for the tibia. Blood loss was equivalent in the two groups (503 mg) and operative time was longer in the navigation group (18 minutes).

Conclusion: It appears important to have specifically adapted cut guides for computer-assisted surgery and to reduce operative time.

Purpose: Implantation of a unicompartmental knee pros-thesis (UKP) via a short incisiosn was proposed to improve rapid postoperative recovery. But it is known that long-term survival is highly dependent on implantation quality. The purpose of this study was to evaluate the influence of implantation quality assessed radiographically and that of rapid recovery after unicompartmental arthroplasty using the conventional or minimally invasive approach.

Material and methods: A continuous series of the first 25 cases of UKP implanted using the minimally invasive approach (group 1) was compared with a homogeneous group of matched knees in a series of 145 UKP implanted with the same approach as used for total knee arthroplasty (group 2). The groups were comparable for age, gender, weight, and height. The same medial cemented UKP with a metallic tibial base plate was implanted in both groups. Postoperative radiographic measures included the overall mechanical axis, the femoral and tibial implantation orientations, and the posterior tibial slope.

Results: There was no significantly different between groups 1 and 2 for mechanical axis (3.6° and 4.7° respectively), tibial implantation angle (87° and 88°), or posterior tibial slope (2.6° and 2°). The femoral implantation angle was also comparable (2.3° and 2.9°) with the exception of one knee (9°) in group 1. Hospital stay, use of crutches and resumed activity were significantly shorter in group 1. Blood loss (222 and 244 ml respectively), postoperative pain (visual analogue scale), and full flexion at one year were comparable between the groups.

Discussion: The basic difference between the two approaches is the absence of patellar eversion and rupture of the extensor system with the minimally invasive method. This probably explains the more rapid recovery of function and activity in this group. The one case of significant femoral deviation noted in this group points out the need for greater care in positioning the specific guides for the short incision approach. To evaluate the long-term results of UKP implanted with the minimally invasive approach, the same criteria for patient selection and prosthetic implantation should be used.

Purpose: Impacted piecemeal allografts for nonce-mented hemispheric cups raises a problem of primary stability in the case of extensive bone defects. The high centre of rotation of the oversized cup further increases bone loss, requiring an extralong neck. The purpose of this study was to describe the use of impacted piecemeal grafts associated with a pressfit supporting ring with reposition of the centre of rotation.

Material and methods: The piecemeal grafts were impacted into the acetabulum to fill the defect. The hydroxyapatite coated ring was pressfit for primary stability then stabilised with axial screws in the upper paste. A distal hook on the obturator foramen repositioned the centre of rotation. The study group included 103 cases of acetabular reconstruction, including 34 for aseptic loosening and type 2 and 3 acetabular substance loss. Clinical and radiographic assessment was performed at 5 and 12 years.

Results: Mean patient age was 58 years, mean weight was 64kg. The Harris score improved from 53 points preoperatively to 88 points at last follow-up. Radiographically, there were no cases of cup migration according to the Massin classification, and the centre of rotation (Pierchon) was anatomic in 66% horizontally and in 44% vertically. There were two lucent lines in zone 2 and mean polyethylene wear was 0.015 mm per year. Graft integration (Conn) was identical to the host in 84% with disappearance of the interface in 67%. There were three dislocations treated without changing the implant and two revisions for infection.

Discussion: Several theoretical and clinical studies have shown that the high centre of rotation increases stress on the implants and decreased abductor force. The results obtained in this study with a maximum 12 year follow-up show that indications for this pressfit technique associating reposition of the centre of rotation, fixation for stability, and restoration of bone stock can be widened. Limitations are bone destructions with rupture of the pelvic girdle.

Introduction: Recent studies showed that the position of the center of rotation and the prosthetic neck may infl uence implant fixation in hip arthroplasty. The purpose of this study is to evaluate the use of modular necks and their limits to restore hip geometry after the arthroplasty.

Methods: The study radiographically evaluates 117 cases of unilateral hip arthroplasty using a modular neck / head concept (Wright Medical). The analysis evaluated on a weight-bearing frontal pelvis view: center of rotation, horizontal abductor ratio, and vertical abductor index, comparatively to the controlateral hip using the student’t test. On the computerized templates of the association effectively used in each case was measured: neck length, lever arm and neck anteversion.

Results: In craniopodal the center of rotation averaged 0.19 in women and 0.23 in men. In mediolateral it averaged 0.26 in women and 0.32 in men. The mean horizontal abductor ratio was 0.65 in women and 0.70 in men. The mean vertical abductor index was 6:..4° in women and 6.5° in men. The mean neck length was 55.2 mm, the mean lever arm 39.3 mm, and the 15° ante or retroverted neck was noted in 10 %.

Discussion and conclusion: Restoration of the center of rotation was more accurate in mediolateral than in craniopodal, with a higher location as previously noted in the litterature. Abductor function was correctly restored excepted for the vertical index in women, probably due to the large variations of pelvis width. The limits of modular necks are large hip dysmorphy where neck length averages 60 mm, lever arm 45 mm and neck anteversion requires 30° of correction.

Introduction: Minimally invasive surgery (MIS) has recently been proposed for unicompartmental arthroplasty to allow quick function recovery. The purpose of this study is to evaluate retrospectively the indications in which this technique would have been possible and successful.

Methods: The requirements for the minimal invasive technique are: a pre-operative flexion of 100° and a lesion limited to one compartment of the knee. The preoperative status, operative findings and clinical outcome of 160 cases of unicompartmental knee arthroplasty (MG, Zimmer) were analyzed to determine whether the criteria for MIS would have been possible.

Results: Preoperatively 12 knees had a flexion less than 100°. Postoperatively four of them had a limited flexion ranging from 90° to 100°. In 53 knees (33 %), peripheral osteophytes were removed on the opposite tibiofemoral joint. None of them were revised for progression of osteoarthritis. In 33 knees (21 %) a patelloplasty was associated to the procedure for peripheral osteophytes. Two of them were revised at 12 and 20 months for osteoarthritis progression.

Discussion: Removal of peripheral osteophytes can be successfully associated to the unicompartmental procedure when using a conventional surgical approach. In 96 of the 160 knees (60 %) MIS was not recommended either for limited preoperative flexion or for peripheral osteophytes. In 2 knees unicompartmental arthroplasty itself was not the correct indication and total knee arthroplasty would have been the right solution. Finally, in 62 of the 160 knees (39 %) unicompartmental arthroplasty using MIS was indicated. In conclusion unicompartmental knee arthroplasty may be either performed by conventional or minimal approach based on the preoperative clinical and radiological evaluation.

Purpose: Hip prosthesis for sequelar developmental dysplasia of the hip is a therapeutic challenge because of the anatomic deformity and the young age of the patients. The purpose of this work was to report results obtained using a non-cemented femoral stem with an intramedullar design and a prosthetic neck custom-made to match individual anatomy observed on preoperative computerised tomographic.

Material and methods: This study included 257 hips with a mean follow-up of 5.6 years. Mean age at implantation was 55 years (range 17–78). The computed tomography study assessed: dislocation according to Crowe, leg length discrepancy, and acetabular anteversion and diameter. The cup was not cemented and was inserted with an anchor hook in the obturator foramen for implantation in the paleoacetabulum. The medullary canal was prepared using a blunt reamer shaped like the definitive prosthesis. The prosthetic neck was designed individually to match the lever arm and anteverion.

Results: There were 174 cases of dysplasia and 83 dislocations (39% grade 1, 30% grade 2, 14% grade 3 and 17% grade 4). Mean lengthening was 39 mm. The mean ante-verion was 28±16° and the mean anteroposterior diameter of the acetabulum was 51 mm. The Harris clinical score improved from 58 points preoperatively to 93 points at last follow-up. The follow-up x-rays showed osteointegration in 88% of the cases with osteolysis in 5% and one stem impaction. The prosthesis had to be changed for six hips: two for infection, one for dislocation and two for nonfixation. The 11-year survival rate was 97%.

Discussion and conclusion: This study illustrates the anatomic sequelae observed in patients with developmental dysplasia of the hip and demonstrates a surgical solution for these problems. There is no correlation between dislocation and the degree of anteversion so it is difficult to assess the difficulty of inserting a non-cemented stem without preoperative computed tomography. The good 11-year survival is encouraging for this young and active population.

Purpose: The current concern to achieve 120° postoperative flexion in certain patients raises the problems of polyethylene wear, prosthesis stability, and surgical technique. The purpose of this clinical and radiographic comparative study was to analyse pertinent factors

Material and methods: We studied two groups of posterior stabilised total knee prostheses with mobile plateaux. In group A (23 knees), the plateau had rotatory and translation mobility. In group B (36 knees) the plateaux only had rotatory mobility and the posterior condyle was long. We studied clinical flexion before surgery and postoperatively at one year. The radiographic analysis included: mechanical axis, height of the articular space, patellar height, anteroposterior prosthetic surface. In both groups the patients had undergone medial parapatellar arthrotomy and had participated in the same rehabilitation programme.

Results: There was no statistically significant difference between the two groups for age, gender, weight, height, pre-operative diagnosis, or preoperative flexion (mean 120.8°, range 90–130° in group A and mean 120.7°, range 90–140° in group B). Mean postoperative flexion was 114.8° (50–140°) in group A and 130.4° (90–150°) in group B. The difference was statistically significant. There were two episodes of instability in group A and none in group B. There was no difference concerning the pre- or postoperative axis or patellar height. There was a difference in articular space height and anteroposterior prosthetic surface.

Discussion: The gain in flexion achieved in group B is related to the smaller anteroposterior prosthetic surface and the more systematic use of distractors to clear the posterior part of the knee. The design of the posterior condyle could also be involved, allowing better congruence beyond 120°. The difference in stability between the two groups could be related to a difference in the distance between the apex of the tibial attachment and the posterior stabilisation cam. This study demonstrated that increased flexion is possible after total knee arthroplasty in certain patients. It is related to surgical technique but also has direct implications for prosthesis design.

Purpose: The Oxford unicompartmental prosthesis has congruent joint surfaces due to use of a mobile meniscus whose upper part matches the sphere constituted by the femoral component. The purpose of this work was to report results in a series of 75 prostheses studied with mean follow-up of 12.4 years (range 10–14 years).

Material and methods: Mean age of patients was 65 years (38–84), with 67% women and mean body mass index of 28.7 (21.2–40.6). Patients gave their subjective assessment of the arthroplasty and the IKS scores were determined. Radiographically, the mechanical alignment of the lower limb was evaluated pre- and postoperatively on the full stance images. The Kennedy classification (4 grades) was used to assess passage of the mechanical axis through the knee.

Results: At last follow-up, three patients were lost and eight had died. According to their subjective assessment, 84% of the patients were enthusiastic or satisfied. The overall IKS score showed that 52% of the patients were above 180 points, 25% above 150 points. Mean varus was 11° (2–24°) preoperatively and 5.7° (5° valgus to 14° varus) at last follow-up. The mechanical axis passed in the sufficient correction zone (c or 2) in 77% of the knees, undercorrection zone in 17% and in the overcorrection zone in 6%.Seven prostheses were revised for mechanical failure: four for loosening, two for extensive degeneration, and one for pain giving a survival curve of 90% at 12 years. It must be added that two meniscal dislocations occurred, one early due to wound dehiscence and infection and one late, both after a fall.

Discussion: Most loosenings occurred early and at the tibial level. Defective alignment could be incriminated in one case. The other knees exhibited a reactive line under the tibial plateau. The rate of dislocation should probably decline with the use of progressive reamers, not available at the time of this series. Finally none of the prostheses were revised for wear, confirming the proper prosthesis design.

This comparative clinical and radiographic study evaluates the role of implant design and positioning on maximum postoperative flexion after TKR.

Two groups of cemented posterostabilised mobile prosthesis were studied. In group A (23 cases) bearing mobility was possible in rotation and translation. In group B (36 cases) solely rotation was possible and the radius of the posterior femoral condyle was larger. In both groups the same medial transquadricipital approach was used and the postoperative rehabilitation was identical.

There were no significant difference between the two groups for patient age, sex, weight and etiology. The average preoperative flexion was 120.8° in group A and 120.7° in group B. The average postoperative flexion at one year was 114.8° in group A and 130.4° in group B. Instability was noted twice in group A and none in group B. Radiographically there was no difference in mechanical axis or patellar height. There was a difference in joint line elevation and anteroposterior joint size.

Higher flexion angle may be related to smaller antero-posterior joint size, better posterior osteophytes removal and larger posterior femoral condyle radius. Greater stability correlates with greater jumping distance over the tibial post. This study showed that deep flexion can be achieved in some patients after TKR, with implications on surgical technique and implant design.

Purpose: Implantation of a prosthesis on a remodelled femur can cause technical difficulties affecting the outcome of the arthroplasty. We performed a tridimensional study of the femoral anatomy before prosthesis implantation for sequelar congenital hip dislocation.

Material and method: The series included 312 hips in 262 patients. The same radiography and computed tomography work-up was perfomred in all patients. There were 288 women and 84 men, men age 56 years. Mean weight was 66 kg and mean height was 163 cm. The crowe classification was 195 dysplasia, 123 dislocations (41% class I, 27% class II, 13% class III, 19% class IV). Telemetric measurements were: femoral isthma, the centre of the lesser trochanter, limb length discrepancy, the cephalo-cervico-diaphyseal angle. Computed tomographic measurements were: anterio-posterior and mediolateral dimensions and femur funneling, helitorsion between the bichondylar plane and the upper femur, anteroposterior diameter of the acetabulum.

Results: The mean mediolateral and anteroposterior diameters of the femoral canal at the isthma were 9.8 and 13.1 cm respectively in dysplasia and 9.3 and 12.6 cm, 9.4 and 12.7, and 9.7 and 13.6 cm in I, II, and III–IV congenital dislocations respectively. The femoral funneling index varied from 1.9 to 7.6 in dysplasia and from 2.6 to 7.9, 2.1 to 8.4 and 2.1 to 8.7 in I, II, and III–IV congenital dislocations respectively. The mean cephalo-cervico-diaphyseal angle was 129.3°, 131.9°, 136.8°, and 127.4° respectively. Maximal leg length discrepancy was 45, 57, 71, and 82 cm respectively. Mean helitorsion was 22.9° (1°–52°), 36.4° (8°–86°), 43.2° (2°–82°- and 38.4° (6°–68°) respectively. The mean anteroposterior diameter of the acetabulum was 52, 51.2, 53.1; and 49.6 cm respectively.

Discussion and conclusion: The dysplastic or dislocated femur is narrower than the normal femur with wide variations in funneling and cephalo-cervico-diphyseal angle. The mean difference in leg length increases gradually with helitorsion but with wide individual variability, irrespective of the grade. These tridimensional anatomic data can be useful for predicting difficulties in prosthetic treatment of these patients.

Purpose: Cinematic studies after total knee arthroplasty without an anterior cruciate ligament demonstrate abnormal behaviour compared with the normal knee. The purpose of this cinematic analysis was to examine the knee behaviour after implantation of single-compartment prostheses with an intact anterior cruciate ligament.

Material and methods: The femorotibial contact points were analysed by videofluoroscopy in 20 patients executing a complete weight-bearing extension to flexion movement. These patients had medial (n=16) or lateral (n=4) single-compartment implants. The clinical result in all patients was considered to be very good with a mean HSS score of 97.9 points at a mean 56 months postoperatively. The femorotibial contact points were determined using an automatic computerised adaptation-modelling system. An anterior contact on the medial tibial line in the sagittal plane was positive and a posterior contact was negative. The rotation axis in the craniopodal direction was measured between the anteroposterior longitudinal axis of the femoral component and the fixed axis of the tibial component.

Results: The mean position of the contact point for medial single-compartment prostheses was −90.8 mm in complete extension, −1.4 mm at 30° flexion, −2.4 mm at 60°, and −1.7 mm at 90°. Mean position of the contact point for lateral single-compartment prostheses was −4.0 mm at complete extension, −7.9 mm at 30° flexion, −5.7 mm at 60° and −5/7 mm at 90°. Seven patients with a medial implant and two patients with a lateral implant exhibited paradoxical anterior translation of the femur during flexion. On the average, patients with a medial implant had normal 3.3° axial rotation at 90°; axial rotation was 11.2° for patients with a lateral implant.

Discussion and conclusion: Cinematic analysis of the normal knee has demonstrated anterior femorotibial contact in extension and 14.2 mm posterior rolling of the femoral component during flexion. After total knee arthroplasty without preservation of the anterior cruciate ligament, the rolling movement is limited or absent and a paradoxical anterior translation can be observed. In the present study, the first reported on single-compartment implants, demonstrates that movement is similar to that in the normal knee but with major interindividual variability. A posterior contact at extension and a paradoxical anterior translation can also be observed. This suggests progressive development of anterior cruciate ligament laxity over time, which can at least in part explain the premature polyethylene wear observed after implantation of single-compartment knee implants.

Biological fixation of cementless femoral implants requires primary stability by optimal fit in the proximal femur. The anatomy of the bone must then be known precisely. We analysed in vitro the accuracy of bone measurements of 32 femurs and compared the dimensions obtained from radiographs and CT scans with the true anatomical dimensions. Standard radiographs gave only a rough approximation of femoral geometry (mean difference: 2.4 +/- 1.4 mm) insufficiently accurate to allow selection of the best fitting prosthesis from a range of sizes and altogether inadequate to design a custom-made prosthesis. CT scans give greater accuracy (mean difference: 0.8 +/- 0.7 mm) in our experimental conditions, but in clinical practice additional sources of error exist.