We evaluated an operative technique, described by the Exeter Hip Unit, to assist accurate introduction of the femoral component. We assessed whether it led to a reduction in the rate of leg-length discrepancy after total hip arthroplasty (THA).

A total of 100 patients undergoing THA were studied retrospectively; 50 were undertaken using the test method and 50 using conventional methods as a control group. The groups were matched with respect to patient demographics and the grade of surgeon. Three observers measured the depth of placement of the femoral component on post-operative radiographs and measured the length of the legs.

There was a strong correlation between the depth of insertion of the femoral component and the templated depth in the test group (R = 0.92), suggesting accuracy of the technique. The mean leg-length discrepancy was 5.1 mm (0.6 to 21.4) pre-operatively and 1.3 mm (0.2 to 9.3) post-operatively. There was no difference between Consultants and Registrars as primary surgeons. Agreement between the templated and post-operative depth of insertion was associated with reduced post-operative leg-length discrepancy. The intra-class coefficient was R ≥ 0.88 for all measurements, indicating high observer agreement. The post-operative leg-length discrepancy was significantly lower in the test group (1.3 mm) compared with the control group (6.3 mm, p < 0.001).

The Exeter technique is reproducible and leads to a lower incidence of leg-length discrepancy after THA.

Cite this article: Bone Joint J 2015;97-B:154–9.

Objectives

Acetabular component orientation in total hip arthroplasty (THA) influences results. Intra-operatively, the natural arthritic acetabulum is often used as a reference to position the acetabular component. Detailed information regarding its orientation is therefore essential. The aim of this study was to identify the acetabular inclination and anteversion in arthritic hips.

Radiological evidence of post-traumatic osteoarthritis (PTOA) after fracture of the tibial plateau is common but end-stage arthritis which requires total knee arthroplasty is much rarer.

The aim of this study was to examine the indications for, and outcomes of, total knee arthroplasty after fracture of the tibial plateau and to compare this with an age and gender-matched cohort of TKAs carried out for primary osteoarthritis.

Between 1997 and 2011, 31 consecutive patients (23 women, eight men) with a mean age of 65 years (40 to 89) underwent TKA at a mean of 24 months (2 to 124) after a fracture of the tibial plateau. Of these, 24 had undergone ORIF and seven had been treated non-operatively. Patients were assessed pre-operatively and at 6, 12 and > 60 months using the Short Form-12, Oxford Knee Score and a patient satisfaction score.

Patients with instability or nonunion needed total knee arthroplasty earlier (14 and 13.3 months post-injury) than those with intra-articular malunion (50 months, p < 0.001). Primary cruciate-retaining implants were used in 27 (87%) patients. Complication rates were higher in the PTOA cohort and included wound complications (13% vs 1% p = 0.014) and persistent stiffness (10% vs 0%, p = 0.014). Two (6%) PTOA patients required revision total knee arthroplasty at 57 and 114 months. The mean Oxford knee score was worse pre-operatively in the cohort with primary osteoarthritis (18 vs 30, p < 0.001) but there were no significant differences in post-operative Oxford knee score or patient satisfaction (primary osteoarthritis 86%, PTOA 78%, p = 0.437).

Total knee arthroplasty undertaken after fracture of the tibial plateau has a higher rate of complications than that undertaken for primary osteoarthritis, but patient-reported outcomes and satisfaction are comparable.

Cite this article: Bone Joint J 2015;97-B:532–8.

In this in vitro study of the hip joint we examined which soft tissues act as primary and secondary passive rotational restraints when the hip joint is functionally loaded. A total of nine cadaveric left hips were mounted in a testing rig that allowed the application of forces, torques and rotations in all six degrees of freedom. The hip was rotated throughout a complete range of movement (ROM) and the contributions of the iliofemoral (medial and lateral arms), pubofemoral and ischiofemoral ligaments and the ligamentum teres to rotational restraint was determined by resecting a ligament and measuring the reduced torque required to achieve the same angular position as before resection. The contribution from the acetabular labrum was also measured. Each of the capsular ligaments acted as the primary hip rotation restraint somewhere within the complete ROM, and the ligamentum teres acted as a secondary restraint in high flexion, adduction and external rotation. The iliofemoral lateral arm and the ischiofemoral ligaments were primary restraints in two-thirds of the positions tested. Appreciation of the importance of these structures in preventing excessive hip rotation and subsequent impingement/instability may be relevant for surgeons undertaking both hip joint preserving surgery and hip arthroplasty.

Cite this article: Bone Joint J 2015; 97-B:484–91.

Hypovitaminosis D has been identified as a common risk factor for fragility fractures and poor fracture healing. Epidemiological data on vitamin D deficiency have been gathered in various populations, but the association between vertebral fragility fractures and hypovitaminosis D, especially in males, remains unclear. The purpose of this study was to evaluate serum levels of 25-hydroxyvitamin D (25-OH D) in patients presenting with vertebral fragility fractures and to determine whether patients with a vertebral fracture were at greater risk of hypovitaminosis D than a control population. Furthermore, we studied the seasonal variations in the serum vitamin D levels of tested patients in order to clarify the relationship between other known risk factors for osteoporosis and vitamin D levels. We measured the serum 25-OH D levels of 246 patients admitted with vertebral fractures (105 men, 141 female, mean age 69 years, sd 8.5), and in 392 orthopaedic patients with back pain and no fractures (219 men, 173 female, mean age 63 years, sd 11) to evaluate the prevalence of vitamin D insufficiency. Statistical analysis found a significant difference in vitamin D levels between patients with vertebral fragility fracture and the control group (p = 0.036). In addition, there was a significant main effect of the tested variables: obesity (p < 0.001), nicotine abuse (p = 0.002) and diabetes mellitus (p < 0.001). No statistical difference was found between vitamin D levels and gender (p = 0.34). Vitamin D insufficiency was shown to be a risk factor for vertebral fragility fractures in both men and women.

Cite this article: Bone Joint J 2015;97-B:89–93.

The indications for reverse shoulder arthroplasty (RSA) continue to be expanded. Associated impairment of the deltoid muscle has been considered a contraindication to its use, as function of the RSA depends on the deltoid and impairment of the deltoid may increase the risk of dislocation. The aim of this retrospective study was to determine the functional outcome and risk of dislocation following the use of an RSA in patients with impaired deltoid function. Between 1999 and 2010, 49 patients (49 shoulders) with impairment of the deltoid underwent RSA and were reviewed at a mean of 38 months (12 to 142) post-operatively. There were nine post-operative complications (18%), including two dislocations. The mean forward elevation improved from 50° (sd 38; 0° to 150°) pre-operatively to 121° (sd 40; 0° to 170°) at final follow-up (p < 0.001). The mean Constant score improved from 24 (sd 12; 2 to 51) to 58 (sd 17; 16 to 83) (p < 0.001). The mean Single Assessment Numeric Evaluation score was 71 (sd 17; 10 to 95) and the rate of patient satisfaction was 98% (48 of 49) at final follow-up.

These results suggest that pre-operative deltoid impairment, in certain circumstances, is not an absolute contraindication to RSA. This form of treatment can yield reliable improvement in function without excessive risk of post-operative dislocation.

Cite this article: Bone Joint J 2013;95-B:1106–13.

Necrotising soft-tissue infections (NSTIs) of the upper limb are uncommon, but potentially life-threatening. We used a national database to investigate the risk factors for amputation of the limb and death.

We extracted data from the Japanese Diagnosis Procedure Combination database on 116 patients (79 men and 37 women) who had a NSTI of the upper extremity between 2007 and 2010.

The overall in-hospital mortality was 15.5%. Univariate analysis of in-hospital mortality showed that the significant variables were age (p = 0.015), liver dysfunction (p = 0.005), renal dysfunction (P < 0.001), altered consciousness (p = 0.049), and sepsis (p = 0.021). Logistic regression analysis showed that the factors associated with death in hospital were age over 70 years (Odds Ratio (OR) 6.6; 95% confidence interval (CI) 1.5 to 28.2; p = 0.011) and renal dysfunction (OR 15.4; 95% CI 3.8 to 62.8; p < 0.001).

Univariate analysis of limb amputation showed that the significant variables were diabetes (p = 0.017) mellitus and sepsis (p = 0.001). Multivariable logistic regression analysis showed that the factors related to limb amputation were sepsis (OR 1.8; 95% CI 1.5 to 24.0; p = 0.013) and diabetes mellitus (OR 1.6; 95% CI 1.1 to 21.1; p = 0.038).

For NSTIs of the upper extremity, advanced age and renal dysfunction are both associated with a higher rate of in-hospital mortality. Sepsis and diabetes mellitus are both associated with a higher rate of amputation.

Cite this article: Bone Joint J 2014;96-B:1530–4.

We present the validation of a translation into Danish of the Oxford ankle foot questionnaire (OxAFQ). We followed the Isis Pros guidelines for translation and pilot-tested the questionnaire on ten children and their parents. Following modifications we tested the validity of the final questionnaire on 82 children (36 boys and 45 girls) with a mean age of 11.7 years (5.5 to 16.0) and their parents. We tested the reliability (repeatability (test–retest), child–parent agreement, internal consistency), feasibility (response rate, time to completion, floor and ceiling effects) and construct validity. The generic child health questionnaire was used for comparison. We found good internal consistency for the physical and the school and play domains, but lower internal consistency for the emotional domain. Overall, good repeatability was found within children and parents as well as agreement between children and parents. The OxAFQ was fast and easy to complete, but we observed a tendency towards ceiling effects in the school and play and emotional domains. To our knowledge this is the first independent validation of the OxAFQ in any language. We found it valid and feasible for use in the clinic to assess the impact on children’s lives of foot and/or ankle disorders. It is a valuable research tool.

Cite this article: Bone Joint J 2015;97-B:420–6.

The purpose of this study was to describe the radiological characteristics of a previously unreported finding: posterior iliac offset at the sacroiliac joint and to assess its association with pelvic instability as measured by initial displacement and early implant loosening or failure. Radiographs from 42 consecutive patients with a mean age of 42 years (18 to 77; 38 men, four women) and mean follow-up of 38 months (3 to 96) with Anteroposterior Compression II injuries, were retrospectively reviewed. Standardised measurements were recorded for the extent of any diastasis of the pubic symphysis, widening of the sacroiliac joint, static vertical ramus offset and a novel measurement (posterior offset of the ilium at the sacroiliac joint identified on axial CT scan). Pelvic fractures with posterior iliac offset exhibited greater levels of initial displacement of the anterior pelvis (anterior sacroiliac widening, pubic symphysis diastasis and static vertical ramus offset, p < 0.001,0.034 and 0.028, respectively). Pelvic fractures with posterior ilium offset also demonstrated higher rates of implant loosening regardless of fixation method (p = 0.05). Posterior offset of the ilium was found to be a reliable and reproducible measurement with substantial inter-observer agreement (kappa = 0.70). Posterior offset of the ilium on axial CT scan is associated with greater levels of initial pelvic displacement and early implant loosening.

Cite this article: Bone Joint J 2014;96-B:1535–9.

A total of 219 hips in 192 patients aged between 18 and 65 years were randomised to 28-mm metal-on-metal uncemented total hip replacements (THRs, 107 hips) or hybrid hip resurfacing (HR, 112 hips). At a mean follow-up of eight years (6.6 to 9.3) there was no significant difference between the THR and HR groups regarding rate of revision (4.0% (4 of 99) vs 5.8% (6 of 104), p = 0.569) or re-operation rates without revision (5.1% (5 of 99) vs 2.9% (3 of 104), p = 0.428). In the THR group one recurrent dislocation, two late deep infections and one peri-prosthetic fracture required revision, whereas in the HR group five patients underwent revision for femoral head loosening and one for adverse reaction to metal debris. The mean University of California, Los Angeles activity scores were significantly higher in HR (7.5 (sd 1.7) vs 6.9 (sd 1.7), p = 0.035), but similar mean Western Ontario and McMaster Universities Osteoarthritis Index scores were obtained (5.8 (sd 9.5) in HR vs 5.1 (sd 8.9) in THR, p = 0.615) at the last follow-up. Osteolysis was found in 30 of 81 THR patients (37.4%), mostly in the proximal femur, compared with two of 83 HR patients (2.4%) (p < 0.001). At five years the mean metal ion levels were < 2.5 μg/l for cobalt and chromium in both groups; only titanium was significantly higher in the HR group (p = 0.001). Although revision rates and functional scores were similar in both groups at mid-term, long-term survival analysis is necessary to determine whether one procedure is more advantageous than the other.

Cite this article: Bone Joint J 2013;95-B:1464–73.

Total knee arthroplasty (TKA) is known to lead to a reduction in periprosthetic bone mineral density (BMD). In theory, this may lead to migration, instability and aseptic loosening of the prosthetic components. Bisphosphonates inhibit bone resorption and may reduce this loss in BMD. We hypothesised that treatment with bisphosphonates and calcium would lead to improved BMD and clinical outcomes compared with treatment with calcium supplementation alone following TKA. A total of 26 patients, (nine male and 17 female, mean age 67 years) were prospectively randomised into two study groups: alendronate and calcium (bisphosphonate group, n = 14) or calcium only (control group, n = 12). Dual energy X-ray absorptiometry (DEXA) measurements were performed post-operatively, and at three months, six months, one, two, four, and seven years post-operatively.

Mean femoral metaphyseal BMD was significantly higher in the bisphosphonate group compared with controls, up to four years following surgery in some areas of the femur (p = 0.045). BMD was observed to increase in the lateral tibial metaphysis in the bisphosphonate group until seven years (p = 0.002), and was significantly higher than that observed in the control group throughout (p = 0.024). There were no significant differences between the groups in the central femoral metaphyseal, tibial medial metaphyseal or diaphyseal regions of interest (ROI) of either the femur or tibia.

Bisphosphonate treatment after TKA may be of benefit for patients with poor bone quality. However, further studies with a larger number of patients are necessary to assess whether this is clinically beneficial.

Cite this article: Bone Joint J 2015;97-B:337–45.

Our understanding of the origin of hip pain in degenerative disorders of the hip, including primary osteoarthritis, avascular necrosis and femoroacetabular impingement (FAI), is limited. We undertook a histological investigation of the nociceptive innervation of the acetabular labrum, ligamentum teres and capsule of the hip, in order to prove pain- and proprioceptive-associated marker expression. These structures were isolated from 57 patients who had undergone elective hip surgery (44 labral samples, 33 ligamentum teres specimens, 34 capsular samples; in 19 patients all three structures were harvested). A total of 15 000 histological sections were prepared that were investigated immunohistochemically for the presence of protein S-100, 68 kDa neurofilament, neuropeptide Y, nociceptin and substance P. The tissues were evaluated in six representative areas.

Within the labrum, pain-associated free nerve ending expression was located predominantly at its base, decreasing in the periphery. In contrast, the distribution within the ligamentum teres showed a high local concentration in the centre. The hip capsule had an almost homogeneous marker expression in all investigated areas.

This study showed characteristic distribution profiles of nociceptive and pain-related nerve fibres, which may help in understanding the origin of hip pain.

Cite this article: Bone Joint J 2013;95-B:770–6.

Our aim was to compare the one-year post-operative outcomes following retention or removal of syndesmotic screws in adult patients with a fracture of the ankle that was treated surgically. A total of 51 patients (35 males, 16 females), with a mean age of 33.5 years (16 to 62), undergoing fibular osteosynthesis and syndesmotic screw fixation, were randomly allocated to retention of the syndesmotic screw or removal at three months post-operatively. The two groups were comparable at baseline.

One year post-operatively, there was no significant difference in the mean Olerud–Molander ankle score (82.4 retention vs 86.7 removal, p = 0.367), the mean American Orthopedic Foot and Ankle Society ankle-hindfoot score (88.6 vs 90.1, p = 0.688), the mean American Academy of Orthopedic Surgeons foot and ankle score (96.3 vs 94.0, p = 0.250), the mean visual analogue pain score (1.0 vs 0.7, p = 0.237), the mean active dorsiflexion (10.2° vs 13.0°, p = 0.194) and plantar flexion (33.6° vs 31.3°, p = 0.503) of the ankle, or the mean radiological tibiofibular clear space (5.0 mm vs 5.3 mm, p = 0.276) between the two groups. A total of 19 patients (76%) in the retention group had a loose and/or broken screw one year post-operatively.

We conclude that removal of a syndesmotic screw produces no significant functional, clinical or radiological benefit in adult patients who are treated surgically for a fracture of the ankle.

Cite this article: Bone Joint J 2014;96-B:1699–1705.

This study sought to determine the medium-term patient-reported and radiographic outcomes in patients undergoing surgery for hallux valgus. A total of 118 patients (162 feet) underwent surgery for hallux valgus between January 2008 and June 2009. The Manchester-Oxford Foot Questionnaire (MOXFQ), a validated tool for the assessment of outcome after surgery for hallux valgus, was used and patient satisfaction was sought. The medical records and radiographs were reviewed retrospectively. At a mean of 5.2 years (4.7 to 6.0) post-operatively, the median combined MOXFQ score was 7.8 (IQR:0 to 32.8). The median domain scores for pain, walking/standing, and social interaction were 10 (IQR: 0 to 45), 0 (IQR: 0 to 32.1) and 6.3 (IQR: 0 to 25) respectively. A total of 119 procedures (73.9%, in 90 patients) were reported as satisfactory but only 53 feet (32.7%, in 43 patients) were completely asymptomatic. The mean (SD) correction of hallux valgus, intermetatarsal, and distal metatarsal articular angles was 18.5° (8.8°), 5.7° (3.3°), and 16.6° (8.8°), respectively. Multivariable regression analysis identified that an American Association of Anesthesiologists grade of > 1 (Incident Rate Ratio (IRR) = 1.67, p-value = 0.011) and recurrent deformity (IRR = 1.77, p-value = 0.003) were associated with significantly worse MOXFQ scores. No correlation was found between the severity of deformity, the type, or degree of surgical correction and the outcome. When using a validated outcome score for the assessment of outcome after surgery for hallux valgus, the long-term results are worse than expected when compared with the short- and mid-term outcomes, with 25.9% of patients dissatisfied at a mean follow-up of 5.2 years.

Cite this article: Bone Joint J 2015;97-B:208–14.

The aim of this study was to examine the loading of the other joints of the lower limb in patients with unilateral osteoarthritis (OA) of the knee. We recruited 20 patients with no other symptoms or deformity in the lower limbs from a consecutive cohort of patients awaiting knee replacement. Gait analysis and electromyographic recordings were performed to determine moments at both knees and hips, and contraction patterns in the medial and lateral quadriceps and hamstrings bilaterally. The speed of gait was reduced in the group with OA compared with the controls, but there were only minor differences in stance times between the limbs. Patients with OA of the knee had significant increases in adduction moment impulse at both knees and the contralateral hip (adjusted p-values: affected knee: p < 0.01, unaffected knee p = 0.048, contralateral hip p = 0.03), and significantly increased muscular co-contraction bilaterally compared with controls (all comparisons for co-contraction, p < 0.01).

The other major weight-bearing joints are at risk from abnormal biomechanics in patients with unilateral OA of the knee.

Cite this article: Bone Joint J 2013;95-B:348–53.

Intra-operative, peri-articular injection of local anaesthesia is an increasingly popular way of controlling pain following total knee replacement. At the same time, the problems associated with allogenic blood transfusion have led to interest in alternative methods for managing blood loss after total knee replacement, including the use of auto-transfusion of fluid from the patient’s surgical drain. It is safe to combine peri-articular infiltration with auto-transfusion from the drain. We performed a randomised clinical trial to compare the concentration of local anaesthetic in the blood and in the fluid collected in the knee drain in patients having either a peri-articular injection or a femoral nerve block. Clinically relevant concentrations of local anaesthetic were found in the fluid from the drains of patients having peri-articular injections (4.92 μg/ml (sd 3.151)). However, none of the patients having femoral nerve blockade had detectable levels. None of the patients in either group had clinically relevant concentrations of local anaesthetic in their blood after re-transfusion.

The evidence from this study suggests that it is safe to use peri-articular injection in combination with auto-transfusion of blood from peri-articular drains during knee replacement surgery.

We reviewed the rate of revision of unicompartmental knee replacements (UKR) from the New Zealand Joint Registry between 1999 and 2008. There were 4284 UKRs, of which 236 required revision, 205 to a total knee replacement (U2T) and 31 to a further unicompartmental knee replacement (U2U). We used these data to establish whether the survival and functional outcome for revised UKRs were comparable with those of primary total knee replacement (TKR). The rate of revision for the U2T cohort was four times higher than that for a primary TKR (1.97 vs 0.48; p < 0.05). The mean Oxford Knee Score was also significantly worse in the U2T group than that of the primary TKR group (30.02 vs 37.16; p < 0.01). The rate of revision for conversion of a failed UKR to a further UKR (U2U cohort) was 13 times higher than that for a primary TKR.

The poor outcome of a UKR converted to a primary TKR compared with a primary TKR should contra-indicate the use of a UKR as a more conservative procedure in the younger patient.

There is conflicting evidence about the functional outcome and rate of satisfaction of super-elderly patients (≥ 80 years of age) after carpal tunnel decompression.

We compiled outcome data for 756 patients who underwent a carpal tunnel decompression over an eight-year study period, 97 of whom were super-elderly, and 659 patients who formed a younger control group (< 80 years old). There was no significant difference between the super-elderly patients and the younger control group in terms of functional outcome according to the mean (0 to 100) QuickDASH score (adjusted mean difference at one year 1.8; 95% confidence interval (CI) -3.4 to 7.0) and satisfaction rate (odds ratio (OR) 0.78; 95% CI 0.34 to 1.58). Super-elderly patients were, however, more likely to have thenar muscle atrophy at presentation (OR 9.2, 95% CI 5.8 to 14.6). When nerve conduction studies were obtained, super-elderly patients were more likely to have a severe conduction deficit (OR 12.4, 95% CI 3.0 to 51.3).

Super-elderly patients report functional outcome and satisfaction rates equal to those of their younger counterparts. They are more likely to have thenar muscle atrophy and a severe nerve conduction deficit at presentation, and may therefore warrant earlier decompression.

Cite this article: Bone Joint J 2014; 96-B:1234–8.

Malpositioning of the trochanteric entry point during the introduction of an intramedullary nail may cause iatrogenic fracture or malreduction. Although the optimal point of insertion in the coronal plane has been well described, positioning in the sagittal plane is poorly defined.

The paired femora from 374 cadavers were placed both in the anatomical position and in internal rotation to neutralise femoral anteversion. A marker was placed at the apparent apex of the greater trochanter, and the lateral and anterior offsets from the axis of the femoral shaft were measured on anteroposterior and lateral photographs. Greater trochanteric morphology and trochanteric overhang were graded.

The mean anterior offset of the apex of the trochanter relative to the axis of the femoral shaft was 5.1 mm (sd 4.0) and 4.6 mm (sd 4.2) for the anatomical and neutralised positions, respectively. The mean lateral offset of the apex was 7.1 mm (sd 4.6) and 6.4 mm (sd 4.6), respectively.

Placement of the entry position at the apex of the greater trochanter in the anteroposterior view does not reliably centre an intramedullary nail in the sagittal plane. Based on our findings, the site of insertion should be about 5 mm posterior to the apex of the trochanter to allow for its anterior offset.

Cite this article: Bone Joint J 2014;96-B:1274–81.

We undertook a retrospective cohort study to determine clinical outcomes following the revision of metal-on-metal (MoM) hip replacements for adverse reaction to metal debris (ARMD), and to identify predictors of time to revision and outcomes following revision. Between 1998 and 2012 a total of 64 MoM hips (mean age at revision of 57.8 years; 46 (72%) female; 46 (72%) hip resurfacings and 18 (28%) total hip replacements) were revised for ARMD at one specialist centre. At a mean follow-up of 4.5 years (1.0 to 14.6) from revision for ARMD there were 13 hips (20.3%) with post-operative complications and eight (12.5%) requiring re-revision.

The Kaplan–Meier five-year survival rate for ARMD revision was 87.9% (95% confidence interval 78.9 to 98.0; 19 hips at risk). Excluding re-revisions, the median absolute Oxford hip score (OHS) following ARMD revision using the percentage method (0% best outcome and 100% worst outcome) was 18.8% (interquartile range (IQR) 7.8% to 48.3%), which is equivalent to 39/48 (IQR 24.8/48 to 44.3/48) when using the modified OHS. Histopathological response did not affect time to revision for ARMD (p = 0.334) or the subsequent risk of re-revision (p = 0.879). Similarly, the presence or absence of a contralateral MoM hip bearing did not affect time to revision for ARMD (p = 0.066) or the subsequent risk of re-revision (p = 0.178).

Patients revised to MoM bearings had higher rates of re-revision (five of 16 MoM hips re-revised; p = 0.046), but those not requiring re-revision had good functional results (median absolute OHS 14.6% or 41.0/48). Short-term morbidity following revision for ARMD was comparable with previous reports. Caution should be exercised when choosing bearing surfaces for ARMD revisions.

Cite this article: Bone Joint J 2014;96-B:1600–9.