Purpose of the study: The objectives of this work were to assess the stability of outcome achieved after Duquennoy procedure to tighten the lateral capsule-ligament structures of the ankle joint. To do this, we reviewed the same series of patients at 3 and 25 years follow-up to analyse the very long-term stability, laxity, and stiffness of the ankle joint and any osteoarthritic degradation.

Material and methods: From 1975 to 1986, we performed 75 Duquennoy procedures on the lateral capsule and ligament structures of the ankle joint. At mean 25 years follow-up (2007) we were able to review 28 ankles (27 patients) with the same method as applied in 1980 (3 years follow-up). The function outcome was assessed with the Good, Karlsson, Duquennoy and Tegner scores. Clinically, other than joint range of motion, were noted varus laxity or anterior drawer. Static x-rays were obtained to search for signs of osteoarthritis and stress images to measure talocrural laxity, in comparison with the preoperative figures and the 1980 data.

Results: There were two early technique failures (one persistent talocrural instability, one subtalar instability). Good and Karlsson scores were good or excellent in 92% of cases. Patients were very satisfied with the operation and the VAS was 8.9/10. Pain and occasional instability were reported by 27% of patients. There was no talocrural no subtalar stiffness at last follow-up. Radiographic laxity declined significantly between the preoperative value and the last review. Only one patient presented a narrow talocrural space (less than 50% narrowing). Eight patients had osteophytes whose presence was not correlated with poor functional outcome. These osteophytes were also present on the controlateral side. Importantly, there was no significant difference in the mean function scores between 3 and 25 years follow-up (Duquennoy score). Similarly joint range of motion and radiographic laxity remained stable over time.

Discussion: The Duquennoy procedure to tighten the lateral capsule and ligament structures is a simple technique with rare complications which provides excellent results both in terms of ligament stability and in terms of preservation from degenerative osteoarthritis. These results remain stable over time. Conversely, the technique has no action on the subtalar joint and should not be proposed as treatment for associated subtalar laxity, if diagnosed.

Purpose of the study: The incidence of revision hip arthroplasty is increasing. In order to overcome certain problems related to loss of femoral bone stock, an original technique was developed combining fragmented allograft material and a metallic mesh with a non filling stem cemented distally in a healthy zone to ensure stability.

Materials and Methods: The clinical data (Postel-Merle-d’Aubigné, PMA score) and radiographic findings (implant migration, loss of bone stock using the SOFCOT and PAPROSKY classification, quality of cementing, filling, and graft aspect, graft lysis, periprosthetic lucency, final aspect of the graft) were collected retrospectively. The operation and the technical difficulties and intraoperative complications were noted.

Results: We report a series of 32 hips treated with this technique and having a mean follow-up of 12.5 years (range 8–20 years). The population studied had particularly significant bone loss (78.2% SOFCOT stage 3 and4). These hips underwent revision for aseptic loosening. The implantation technique required a femoral window in 39.1%. Preventive cerclage was often used (39.1%) but did not prevent fracture or missinsertion in 30.4%. The PMA score improved significantly from 10.6 (7–18) preoperatively to 17 (12–18) at last follow-up. Radiologically, femoral bone regeneration at last follow-up had an aspect of corticalisation in 63.6% of hips, and of cancellous trabeculation in 36.4%. Femoral implant survival was 100% at eight years, and 92.8±6.88% at mean follow-up of 12.5 years. There was only one revision at eleven years for secondary osteolysis related to polyethylene wear in a very active subject implanted before the age of 50 years.

Discussion: The clinical and radiographic results are very satisfactory for this series of femoral revisions using an impacted fragmented graft material and with the longest follow-up reported in the literature. Distal fixation limited migration observed when the stem is entirely cemented in the graft, but did not affect reconstruction which demonstrated long-term stability. This technique, initiated in 1986 without a specific instrument set, had now demonstrated its long-term reliability. The only problem is the length of the operation and the complications related to femoral preparation.

Purpose of the study: Theoretically, long-term functional and radiographic degradation is predictable after ankle fusion, but sound evidence from consecutive analysis of the same cohort is lacking. The purpose of this study was to evaluate the same cohort of patients who underwent ankle fusion.

Material and Methods: The cohort included 52 ankle fusions which had been analyzed in 1984 at seven years (2–22 years) of follow-up then again in 2000 using the same evaluation criteria. Among the 52 patients, six were lost to follow-up, 20 had died and one had undergone leg amputation. The second analysis thus included 25 patients (18 men and 7 women), mean age 62±12.6 years (40–94) at the 2000 assessment performed 23±4.5 years (19–36 years) after the fusion. Functional outcome was assessed with the 100-point Duquennoy scale. Osteoarthritis of the subtalar and mediotarsal joints were assessed preoperatively and at follow-up using the same scale.

Results: The functional outcome did not deteriorate significantly between 1984 and 2000. The mean score was 65.8±22.6 (19–92) in 1984 and 64.7±18.3 (34–90) in 2000 (p=0.67). Fifteen patients (60%) had a good or very good outcome at seven years, and 14 (56%) at 23 years. Between 1984 and 2000, ten patients improved their score (on average 10.4 points, range 1–21 points), two had an unchanged score and thirteen a lower score (on average 10 points, range −1 to −24). Ten of these thirteen patients developed severe intercurrent conditions explaining the degradation. At last follow-up, sixteen ankles were pain free or nearly pain free. Twelve patients considered their ankle as a forgotten problem and had no regrets concerning the operation. The evolution of the subtalar joint in 16 cases (nine fusions including five at the same time as the ankle fusion and four performed within four years) showed that all developed early osteoarthrtic degradation with aggravation between 1984 and 2000, leading in the majority of cases to severe degenerative disease. This osteoarthritis was painful in less than one-third of the cases (including the four secondary subtalar fusions and the four sub-talar fusions which were painful at mobilization). The mediotarsal joints degradation was later and less severe than for the subtalar joint with a majority of moderate osteoarthritis. Ten ankles exhibited compensatory hypermobility of the forefoot measured at more than 15° without pain.

Discussion: This long-term follow-up with two successive assessments using the same evaluation criteria did not demonstrated late degradation of function expected after ankle fusion. It did show however the presence of radiographic degradation of the subtalar joint but with little clinical expression at a minimal follow-up of 19 years. There was no need for complementary fusion between 4 and 23 years follow-up.

Thirty-nine cementless hip replacements using metal-on-metal articulation were consecutively implanted in thirty patients less than fifty years of age and compared with a matched control group of cementless replacements using ceramic-on-polyethylene articulation. The Harris hip score at follow-up (minimum five years) for the metal-on-metal was 94.9 (range, 74–100). After the same follow-up, the results of the ceramic-on-polyethylene were significantly worse: nine osteolyses and seven surgical revisions related to wear. Five-year survival rates were 97% +/− 2% for the ceramic-on-polyethylene and 100% for the metal-on-metal. The metal-on-metal may be recommended to prevent wear problems in younger and more active patients.

The aim of the current study was to assess the results of metal-on-metal articulating components inserted as a primary hip replacement in patients under the age of fifty, comparated with a matched control group using ceramic-on-polyethylene.

Patients and methods

The inclusion of patients was: under fifty years of age and a diagnosis of arthrosis or necrosis of the femoral head.

Femoral stem and cup migration was detected. A variation over five millimeters between the follow-up radiographs was considered as migration.

At the follow-up in 2003, the cobalt concentrations in the whole blood were assessed in the metal-on-metal cohort. The detection limit of cobalt in the whole blood was 0.06 μg/L.

None of the components had migration.

At a mean follow-up of sixty-nine months, the median concentration of cobalt in the whole blood was 0.62μg/L. Only eight patients had cobalt levels greater than 1 μg/L.

Considering a reoperation with the exchange of one of the components as end point, the five year survival rates were 100% for the metal-on-metal group and 97% + 2 for the ceramic-on-polyethylene group. Our study suggests that the metal-on-metal articulation gives a significant improvement in terms of resistance to wear when compared with these conventional bearing components. Our results suggest the metal-on-metal articulation with cementless components can be recommended in the young and active patient to prevent the occurrence of wear and osteolysis. A careful assessment of patients with high levels of whole-blood cobalt should be performed.

Funding: Aucun

Purpose: In young active subjects, total hip arthroplasty (THA) raises the risk of early polyethylene wear eventually warranting the use of alternative bearings. The purpose of this study was to analyze outcome in subjects aged less than 50 years implanted with a primary metal-on-metal prosthesis.

Material and methods: Between 1995 and 1998, thirty-nine THA with a metal-on-metal bearing were implanted without cement in 30 consecutive patients aged less than 50 years, mean age 23–50). There were nine bilateral implantations. The cementless implants had surface treatment without hydroxyapatite and 28 mm heads. The indication for THA was necrosis in twenty cases, osteoarthritis in nineteen, most in patients with hip dysplasia. The Devane classification placed 84% of the patients in levels 4 or 5, indicating heavy work and/or sports activities. Inclusion criteria were: 1) age less than 50 years, 2) significant activity (occupation and/or sports), 3) osteoarthritis or necrosis. Patients were included and followed prospectively. Serum cobalt level was measured at last follow-up. None of the patients were lost to follow-up.

Results: The Postel-Merle-d’Aubigné score (PMA) improved from 12.8±2.2 (7–15) before surgery to 17.2±1 (14–18) at 5.1 years (5–6.3). None of the patients complained of hip pain. Two patients has moderate inguinal pain related to a cupiliopsoas conflict. All patients achieved complete weight bearing on the fourth postoperative day excepting three who resumed weight bearing at six weeks due to an acetabular augmentation graft. There were no cases of implant migration and all implants exhibited signs of osteointegration on the radiograms. There were no postoperative dislocations and no cases of osteolysis could be identified, particularly in the eight patients who had increased serum cobalt (four bilateral implants) and who had no other statistically significant favouring factor.

Conclusion: The results at five years suggest that the second-generation metal-on-metal bearing is a reliable alternative when THA is proposed for young active subjects. Long-term follow-up of this cohort of subjects at high risk of wear is necessary to confirm these encouraging results.

Aims: The goal was to determine if young patients respect the instructions for medical supervision control instructions after a total hip arthroplasty (THA). Methods:We performed 50 THA (Harris cups and ABG I stems) in 39 patients, mean age 38 years ± 11 (15–58). There were thirty-eight 28mm femoral heads (30 zirconia, 8 chromium-cobalt) and 12 chromium-cobalt heads in 22mm. Each patient was told how necessary a regular medical supervision was with the following schedule: controls at 2 months, 1 year then every two years. In 2000, all the patients were evaluated (no lost for follow-up). Wear was measured according to Livermore with a digitizer. Results: At 72 ± 14 months of follow-up (5–9 years), 51% of the patients (20/39) did not come back for the intermediate controls. No factor could be accounted for that, except being a male (p=0.04). Wear was rather severe (average 0.23 mm/year). 37% of the patients with a severe wear (> 0.2 mm/year) did not come back for intermediate controls and were all free of symptoms. There were 3 revisions because of wear (> 2 mm) and 3 others hips are awaiting revision because of wear. Three of these 6 patients did not come back for intermediate examination. The 28mm heads induced more wear than the 22mm heads (p=0.008). No other factor was related to how severe wear was (age, sex, activity or thickness of polyethylene inserts). Moreover the zirconia did not reduce wear. Conclusion: In spite of the recommendations, half of the patients did not respect the medical supervision schedule and that was unpredictable. We recommend a careful follow-up of young patients after a THA in order to detect wear (usually free of symptoms). Moreover our study underlined that the zirconia wasn’t so efficient to prevent wear phenomena.

Aims: Henri Dejour promoted a mechanism involving a third femoral condyle to achieve Posterior Stabilization (PS) in total knee arthroplasty (TKA) introducing the HLS II prosthesis. This retrospective study was conducted to assess the behavior of such PS mechanism. Methods: Between 1992 and 1993, 105 HLS II prostheses (94 patients) were consecutively inserted (78% arthrosis, 19% rheumatoid arthritis). No patient was lost for follow-up but 14 had died, 6 were unable to walk (severe neurological disorder), 4 were contacted by phone. Consequently, 70 patients (77 TKA) mean aged 66 years (22–79) were assessed after a mean follow-up of 7 years (6–8). All the components were fixed with cement and patellar resurfacing was always performed. Results: The knee IKS score increased from de 27 points [0–63] before surgery to 81 [21–100] at follow-up and functional IKS from 35 points [0–75] to 64 [0–100] (p< 0,0001). Similarly range of motion improved from 114° [60°–140°] to 116° [80°–135°] (NS). At follow-up, 86% of the patients were able to practice stairs (13% without support and 28% in alternative manner) against 52% before surgery (1% without support and 1% in alternative manner) (p=0,001). Tibial bone-cement radiolucencies were observed without loosening in 30% (all < 1 mm and non- progressive) mainly related to severe preoperative varus deformation (p = 0.01). One late infection required reoperation. Ninety months survival was 97% ± 1.3% with reoperation related to infection or mechanical disorder as end-point. Conclusion: The posterior stabilization, by means of a third condyle, allowed a satisfactory range of flexion and improved ability to practice stairs. Mid-term follow-up did not identified adverse effects of this PS mechanism on component fixation or knee stability.

Purpose of the study: There are few reports onlong-term outcome after Bankart procedure. The purpose of this study was to determine the rate of recurrent dislocation, the clinical results and the incidence of glenohumeral osteoarthritis after a minimum 10-year follow-up.

Material and methods: Ninety-seven Bankart procedures were performed in 97 patients between 1972 and 1986 for treatment of anterior shoulder instability with recurrent dislocations. We retrospectively reviewed 74 patients and obtained 64 complete radioclinical evaluations for an average follow-up of 16 years. Clinical evaluation was based on the G. Walch and the Duplay group score but for easier comparisons, we also calculated the Rowe et al. score. Radiographical evaluation was established on the Samilson and Prieto classification but real glenohumeral osteoarthritis with joint narrowing was noted independently as grade four. We also studied the contralateral shoulder.

Results: At last follow-up, 7 shoulders (9.5%) had recurrent dislocation, but two of them occurred subsequent to severe trauma over 18 months. Most patients (95%) were satisfied or very satisfied. Six patients (8.1%) had persistent apprehension but in some it was not due to anterior apprehension. According to the Duplay score (or the Rowe score), 25 shoulders (44.6%) had an excellent result (35/61.4%) 16 (28.6%) a good result (7/12.3%), 11 (19.7%) a fair result (11.19.3) and 4 (5.4%) a poor result (4/7%). Operated shoulders were pain free for 75% and painful for forced movements only for 25%. External rotation at 90° of abduction was reduced by 8.7 ± 15.7°. There was no limitation of internal rotation. Patients returned to preoperative sports activities at the same level for 70.9% and at a lower level for 12.7%. According to the Samilson classification, 7 (13%) of the shoulders had grade 2 and 2 (3.7%) had grade 3 glenohumeral osteoarthritis. We found 4 cases (7.4%) of real glenohumeral osteoarthritis (grade four) and 2 of these patients had contralateral osteoarthritis of a non unstable shoulder. There was no perioperative complication.

Discussion: In our hands the Bankart procedure is appeared as a safe procedure with a low rate of glenohumeral osteoarthritis and a high rate of patient satisfaction.

Purpose: The aims of this work were to assess very long-term outcome and to assess functional course of talocrural arthrodesis as well as to determine the clinical and radiological impact on adjacent joints.

Material and method: Fifty-two talocrural arthrodeses performed in 52 patients between 1963 and 1981 served as the reference population for this study. Clinical and radiological assessment of this series of patients was conducted in 1983, then again in 1999. Twenty-five talocrural arthrodeses in 25 patients were reviewed at a mean 23 years (19 to 36 years) (six patients were lost to follow-up, 20 had died, one had had leg amputation. The arthrodesis had been performed for advanced degenerative joint disease or to correct for post-traumatic deformity, or in two cases, for neurological varus equinus. Functional outcome was evaluated on the Duquennoy and Stahl score (100 points) that was also used for the intermediary assessment. Radiographic assessment included the position of the arthrodesis, the status of the adjacent joints (subtalar and mediotarsal), and residual motion of the forefoot.

Results: At 23 years follow-up, 13 patients (52%) had good (five patients) or very good (eight patients) function and 12 (48%) had fair function, according to the 100 point scale. There were no patients with poor function. Patients without intercurrent conditions (neurological or heart disease, obesity) had good or very good function in 80% of the cases. Patients were very satisfied in 48% of the cases, having “forgotten” their ankle. At the seven-year follow-up analysis, 15 patients (60%) had a very good (ten patients) or good (five patients) result, seven had a fair result, and three had a poor result. Comparing the 7-year and 23-year assessments did not disclose any significant degradation of the result (p = 0.07). Intercurrent conditions explained the two functional degradations, but there were also three functional improvements over the same period. Talocrural arthrodesis induced stiffness in the subtalar joint in all cases, associated with severe osteoarthritis with little clinical expression. There was a slow degradation of the mediotarsal joint but hypermobility useful for good function was maintained in 45% of the cases (mean 24°). Fair results were related to development of subtalar osteoarthritis, malposition of the arthrodesis in the frontal plane (rear foot varus) and presence of intercurrent conditions (cardiovascular, neurological disease).

Conclusion: Talocrural arthrodesis is a safe and reliable procedure for the treatment of destroyed joints. This palliative surgery can restore satisfactory function which persists in the long term.

Purpose: A new radiographic classification of the femoral trochlea was proposed by David Dejour in 1998 to quantify the severity of bony dysplasia. The purpose of this work was to evaluate the reproducibility of this classification system and to determine its contribution to the identification of trochlea with a high-risk of femoropatellar instability.

Material and methods: Nine independent observers (one resident, four junior surgeons, four senior surgeons) with no knowledge of the patient’s history read 68 strict lateral views of knees with femoropatellar instability (53 objective instabilities (OI) and 15 potential instabilities (PI)). The classification system includes four types determined with three signs: crossing (defining the dysplasia and present in all four types), supratrochlear spike, double contour. The four types are: type A crossing alone, type B crossing and spike, type C crossing and double contour, type D crossing, spike and double contour. The kappa test was used to assess reproducibility and chi square test to analyse data by category.

Results: Twenty-one radiographs were excluded by one or several observers due to insufficient quality or the impossibility to identify the signs of the new classification. Interob-server reproducibility assess on 47 radiographs was fair (kappa = 0.48). The crossing sign was identified by the nine observers on the 47 radiographs. Reproducibility of identification of the spike was good (κ= 0.62), but was fair for the double contour (κ = 0.51). there was no difference in reproducibility by level of experience of the observers. The new classification system was not correlated with severity of femoropatellar instability: presence of spike 80% OI, absence of spike 67% OI; presence of double contour 74% OI, absence of double contour 75% OI.

Discussion, conclusion: This new classification system is more reproducibly than the former 3-type system proposed by Henri Dejour. The crossing sign and the spike are the most reproducible signs. There presence is however insufficient to quantify the dysplasia and predict the severity of the femoropatellar instability. A quantitative measure of the depth of the trochlea, which shows excellent reproducibility (interclass coefficient 0.65) could be added to better quantify the morphological anomaly and determine the most adapted treatment.

Purpose: The purpose of this prospective work was to determine: 1) the frequency of iliopsoas conflicts with prosthetic acetabular material in patients with a painful total hip arthroplasty, and 2) to determine the diagnostic features of this conflict and the results of an adapted therapeutic approach.

Material and methods: This prospective study was conducted between 1988 and 2000 in 206 painful total hip arthroplasties. Nine patients (4.%), mean age 50 years (38–65) had a highly suggestive clinical presentation (eight press-fit cups without cement, one metal-backed cemented cup). Mean delay to onset after implantation of the total hip arthroplasty was 7.3 months (1–48). The predominant clinical sign was groin pain triggered by active flexion of the hip with pain from 3° to 70°. For these nine patients, there was no sign of loosening and puncture had ruled out infection. The diagnosis was confirmed by sedation of the pain after extra-articular infiltration in contact with the anterior acetabular rim (rim overhang in six out of nine cases) under computed tomographic guidance.

Results: Therapeutic infiltration (xylocaine-slow release corticosteroids) under computed tomographic guidance provided complete sedation in four out of nine cases and partial sedation in one. Four patients experienced recurrence leading to terminal tenotomy of the iliopsoas which provided complete sedation in three and partial sedation in one. In all, seven of the nine patients achieved complete pain relief (four after infiltration including one recurrent case, and three after tenotomy). Physical examination at last follow-up did not disclose any loss of flexion amplitude of the thigh.

Discussion and conclusion: The delay to symptom onset was variable, but a symptom-free interval was always observed after implantation of the total hip arthroplasty. An anatomic factor (anterior cup rim) was not indispensable for diagnosis as it was confirmed in one of three cases with a positive infiltration test without rim overhang. The infiltration test was essential to confirm diagnosis and constituted the first therapeutic attempt which allowed complete cure in four of the nine patients. Tenotomy, indicated in case of recurrence, achieved complete cure in three of the four cases. The cup does not have to be changed necessarily to treat iliopsoas conflicts since infiltration or simple tenotomy provided complete cure in seven out of nine cases.

Purpose: Changes in the lever arm of the abductors is not always perfectly controlled during implantation of total hip arthroplasties. Its possible effect on the development of prothesis dislocation is not known. The purpose of this study was to evaluate the influence of the lever arm and its modifications on the development of prosthetic instability.

Material and methods: We analysed prospectively 73 total hip arthroplasties implanted via the posterolateral approach. The study group was composed of a consecutive series of 45 dislocated prostheses and a control group of 28 stable prostheses selected at random. The following measurements were made on the anteroposterior x-ray: 1) lever arm of the abductors, 2) femoral offset. These measures were compared with the healthy contralateral hip and when this hip was diseased or had a prosthesis, with the pre-implantation x-rays.

Results: None of the studied parameters was statistically different between the dislocated and stable prostheses. However, in the dislocated prostheses, the lever arm of the abductors before insertion of the prosthesis was shorter than in the control group (p = 0.04) suggesting the presence of a group of hips “at risk”. There was a correlation between the offset values and the lever arm values for the stable prostheses and for the healthy contralateral hips in both groups. Conversely, this balance was not found in the dislocated hips. The lever arm/offset ratio was calculated to determine if the ideal ratio influenced hip stability. This ratio was not directly related to the development of dislocation, but it was decreased for dislocated hips. This ratio was ideal for 75% of the stable prostheses and for only 53% of the dislocated prostheses.

Conclusions: We concluded that: 1) hips “at risk” of dislocation would have a shorter lever arm, 2) the lever arm or the femoral offset do not have a direct effect on dislocation, and 3) stable hip prostheses have a balance similar to that in healthy hips identified by a correlation between the lever arm and the femoral offset. We thus emphasise the importance of respecting these parameters although they are probably not the only factors influencing prosthesis stability. Allowable variations are small, demanding careful and precise operation planning.

Purpose: The aim of this study was to follow a prospective cohort of young subjects with total hip arthroplasties (THA) in order to determine 1) how they complied with instructions concerning the surveillance of their implant, and 2) determine factors affecting non-compliance and the potential consequences in terms of wear.

Material and methods: We implanted fifty Harris I cups with the ABG I stem in 15 young men and 24 young women (11 bilateral implantations) between 1991 and 1995. These patients were a non-consecutive prospective series of patients under 60 years of age (mean age at implantation was 38 ± 11 years, range 15–58 years). The ABG I stem was implanted when the femoral canal was cylindrical, other implants were chosen for other morphologies. 28-mm femoral heads were used in 38 cases (30 zircone, 8 chromium-cobalt, polyethylene thickness 8.6 mm, range 8.3–12.4). The patients were informed of the need for regular surveillance with controls at two months, and one year and then every two years. We re-evaluated all the patients again in 2000. Radiographic wear was assessed according to Livermore using a numeric table (OrthoGraphics).

Results: At mean follow-up of 72 ± 14 months (5–9 years), 20 of the 39 patients (25/50 implants) had not attended the intermediary consultations, but all were seen again for the systematic review made in 2000. There was no particular factor related to non-attendance other than male gender (p = 0.04). Wear was rather severe. Mean values were: linear wear 1.32 mm, 0.23 mm/an. Thirty-seven percent of the patients who had severe annual linear wear (> 0.2 mm) had not attended consultations, all were asymptomatic. The 28 mm head produced more volumetric wear than the 22 mm heads (p = 0.008). There was no other factor correlated with severity of wear (age, sex, activity, polyethylene thickness). This systematic revision led to: three replacements due to excessive wear > 2 mm and replacements planned for three others with > 2 mm wear. These six patients were asymptomatic and three of the six had never attended the planned visits.

Conclusion: Despite our recommendations, half of the patients did not comply with the planned surveillance protocol. Non-compliance being inevitable, we recommend regular systematic recalls to detect severe wear early, even in asymptomatic patients. In addition, our study demonstrated that zircone heads contribute little and confirmed the interest of 22 mm heads to limit wear phenomena.

We reviewed radiographs and CT scans of 38 total hip arthroplasties which had dislocated (36 posteriorly; 2 anteriorly) and compared the alignment of the prosthetic components with those of 14 uncomplicated arthroplasties. No difference was found between the alignment of the prosthetic components in the two groups. In the seven patients who had reoperations, the cause of dislocation diagnosed by CT was confirmed in only two cases (one retroversion of the cup and one protruding osteophyte). Muscular imbalance rather than malposition of the components was the major factor determining dislocation. CT allows accurate measurement of cup and neck anteversion but contributes little to preoperative planning.