Survival analysis is an important tool for assessing the outcome of total joint replacement. The Kaplan-Meier method is used to estimate the incidence of revision of a prosthesis over time, but does not account appropriately for competing events which preclude revision. In the presence of competing death, this method will lead to statistical bias and the curve will lose its interpretability. A valid comparison of survival results between studies using the method is impossible without accounting for different rates of competing events. An alternative and easily applicable approach, the cumulative incidence of competing risk, is proposed. Using three simulated data sets and realistic data from a cohort of 406 consecutive cementless total hip prostheses, followed up for a minimum of ten years, both approaches were compared and the magnitude of potential bias was highlighted. The Kaplan-Meier method overestimated the incidence of revision by almost 4% (60% relative difference) in the simulations and more than 1% (31.3% relative difference) in the realistic data set. The cumulative incidence of competing risk approach allows for appropriate accounting of competing risk and, as such, offers an improved ability to compare survival results across studies.

Although the importance of sound statistical principles in the design and analysis of data has gained prominence in recent years, biostatistics, the application of statistics to the analysis of biological and medical data, is still a subject which is poorly understood and often mishandled. This review introduces, in the context of orthopaedic research, the terminology and the principles involved in simple data analysis, and outlines areas of medical statistics that have gained prominence in recent years. It also lists and provides an insight into some of the more common errors that occur in published orthopaedic journals and which are frequently encountered at the review stage in papers submitted to the Journal of Bone and Joint Surgery.

We present an update of the clinical and radiological results of 62 consecutive acetabular revisions using impacted morsellised cancellous bone grafts and a cemented acetabular component in 58 patients, at a mean follow-up of 22.2 years (20 to 25). The Kaplan-Meier survivorship for the acetabular component with revision for any reason as the endpoint was 75% at 20 years (95% confidence interval (CI) 62 to 88) when 16 hips were at risk. Excluding two revisions for septic loosening at three and six years, the survivorship at 20 years was 79% (95% CI 67 to 93). With further exclusions of one revision of a well-fixed acetabular component after 12 years during a femoral revision and two after 17 years for wear of the acetabular component, the survivorship for aseptic loosening was 87% at 20 years (95% CI 76 to 97). At the final review 14 of the 16 surviving hips had radiographs available. There was one additional case of radiological loosening and four acetabular reconstructions showed progressive radiolucent lines in one or two zones.

Acetabular revision using impacted large morsellised bone chips (0.5 cm to 1 cm in diameter) and a cemented acetabular component remains a reliable technique for reconstruction, even when assessed at more than 20 years after surgery.

In a series of 450 patients over 70 years of age with displaced fractures of the femoral neck sustained between 1995 and 1997 treatment was randomised either to internal fixation or replacement. Depending on age and level of activity the latter was either a total hip replacement or a hemiarthroplasty. Patients who were confused or bed-ridden were excluded, as were those with rheumatoid arthritis. At ten years there were 99 failures (45.6%) after internal fixation compared with 17 (8.8%) after replacement. The rate of mortality was high at 75% at ten years, and was the same in both groups at all times. Patient-reported pain and function were similar in both groups at five and ten years. Those with successfully healed fractures had more hip pain and reduction of mobility at four months compared with patients with an uncomplicated replacement, and they never attained a better outcome than the latter patients regarding pain or function.

Primary replacement gave reliable long-term results in patients with a displaced fracture of the femoral neck.

This retrospective cohort study was conducted to investigate whether operative treatment of patients with a pertrochanteric femoral fracture outside working hours is associated with an increased risk of complications and higher mortality. During the study period 165 patients were operated on outside working hours and 123 were operated on during working hours (08.00 to 17.00). There was no difference in the rate of early complications (outside working hours 33% versus working hours 33%, p = 0.91) or total complications during follow-up (outside working hours 40% versus working hours 41%, p = 0.91). Both in-hospital mortality (outside working hours 12% versus working hours 11%, p = 0.97) and mortality after one year (outside working hours 29% versus working hours 27%, p = 0.67) were comparable. Adjustment for possible confounders by multivariate logistic regression analysis revealed no increased risk of complications when patients were operated on outside working hours.

On the basis of these data, there is no medical reason to postpone operative reduction and fixation in patients with a proximal femoral fracture until working hours.

The long-term success of total knee replacement is multifactorial, including factors relating to the patient, the operation and the implant. The purpose of this study was to examine the 20-year survival of the cemented Anatomical Graduated Component (AGC) total knee replacement. Between 1983 and 2004, 7760 of these were carried out at our institution. Of these, 6726 knees which received the non-modular metal-backed tibial component with compression-moulded polyethylene and had a minimum two-year follow-up were available for study. In all, 36 knees were followed over 20 years with a survival of the tibial and femoral components together of 97.8% (95% confidence interval (CI) 0.9851 to 0.9677), with no implants being revised for polyethylene wear or osteolysis. Age > 70 was associated with increased survival (99.6%, 95% CI 99.0 to 99.8) (p < 0.0001) but pre-operative valgus alignment reduced survival (95.1%, 95% CI 90.0 to 97.6) (p = 0.0056). Age < 55 (p = 0.129), pre-operative varus alignment (p = 0.707), osteonecrosis (p = 0.06), rheumatoid arthritis (p = 0.247), and gender (p = 0.666) were not statistically associated with failure.

We attribute the success of the AGC implant to its relatively unconstrained articular geometry and the durability of a non-modular metal-backed tibial component with compression moulded polyethylene.

We report the outcome at a minimum of five years of 110 consecutive metal-on-metal Birmingham Hip Resurfacing arthroplasties in 98 patients. The procedures were performed between October 1999 and June 2002 by one surgeon. All patients were followed up clinically and radiologically. The mean follow-up was 71 months (60 to 93). Revision of either component was defined as failure.

The mean Harris Hip score at follow-up was 96.4 (53 to 100). The mean Oxford hip score was 41.9 (16 to 57) pre-operatively and 15.4 (12 to 49) post-operatively (p < 0.001). The mean University of California Los Angeles activity score was 3.91 (1 to 10) pre-operatively and 7.5 (4 to 10) post-operatively (p < 0.001).

There were four failures giving a survival at five years of 96.3% (95% confidence interval 92.8 to 99.8). When applying a new method to estimate narrowing of the femoral neck we identified a 10% thinning of the femoral neck in 16 hips (14.5%), but the relevance of this finding to the long-term outcome remains unclear.

These good medium-term results from an independent centre confirm the original data from Birmingham.

Between 1989 and 1992 we had 102 knees suitable for unicompartmental knee replacement (UKR). They were randomised to receive either a St Georg Sled UKR or a Kinematic modular total knee replacement (TKR). The early results demonstrated that the UKR group had less complications and more rapid rehabilitation than the TKR group. At five years there were an equal number of failures in the two groups but the UKR group had more excellent results and a greater range of movement.

The cases were reviewed by a research nurse at 8, 10 and 12 years after operation. We report the outcome at 15 years follow-up. A total of 43 patients (45 knees) died with their prosthetic knees intact. Throughout the review period the Bristol knee scores of the UKR group have been better and at 15 years 15 (71.4%) of the surviving UKRs and 10 (52.6%) of the surviving TKRs had achieved an excellent score. The 15 years survivorship rate based on revision or failure for any reason was 24 (89.8%) for UKR and 19 (78.7%) for TKR. During the 15 years of the review four UKRs and six TKRs failed.

The better early results with UKR are maintained at 15 years with no greater failure rate. The median Bristol knee score of the UKR group was 91.1 at five years and 92 at 15 years, suggesting little functional deterioration in either the prosthesis or the remainder of the joint. These results justify the increased use of UKR.

The rate and mode of early failure in 463 Birmingham hip resurfacings in a two-centre, multisurgeon series were examined. Of the 463 patients two have died and three were lost to follow-up. The mean radiological and clinical follow-up was for 43 months (6 to 90).

We have revised 13 resurfacings (2.8%) including seven for pain, three for fracture, two for dislocation and another for sepsis. Of these, nine had macroscopic and histological evidence of metallosis. The survival at five years was 95.8% (95% confidence interval (CI) 94.1 to 96.8) for revision for all causes and 96.9% (95% CI 95.5 to 98.3) for metallosis.

The rate of metallosis related revision was 3.1% at five years. Risk factors for metallosis were female gender, a small femoral component, a high abduction angle and obesity. We do not advocate the use of the Birmingham Hip resurfacing procedure in patients with these risk factors.

We evaluated the cost and consequences of proximal femoral fractures requiring further surgery because of complications. The data were collected prospectively in a standard manner from all patients with a proximal femoral fracture presenting to the trauma unit at the John Radcliffe Hospital over a five-year period. The total cost of treatment for each patient was calculated by separating it into its various components. The risk factors for the complications that arose, the location of their discharge and the mortality rates for these patients were compared to those of a matched control group. There were 2360 proximal femoral fractures in 2257 patients, of which 144 (6.1%) required further surgery. The mean cost of treatment in patients with complications was £18 709 (£2606.30 to £60.827.10), compared with £8610 (£918.54 to £45 601.30) for uncomplicated cases (p < 0.01), with a mean length of stay of 62.8 (44.5 to 79.3) and 32.7 (23.8 to 35.0) days, respectively. The probability of mortality after one month in these cases was significantly higher than in the control group, with a mean survival of 209 days, compared with 496 days for the controls. Patients with complications were statistically less likely to return to their own home (p < 0.01).

Greater awareness and understanding are required to minimise the complications of proximal femoral fractures and consequently their cost.

Second-generation metal-on-metal bearings were introduced as a response to the considerable incidence of wear-induced failures associated with conventional replacements, especially in young patients. We present the results at ten years of a consecutive series of patients treated using a metal-on-metal hip resurfacing. A distinct feature of the bearings used in our series was that they had been subjected to double-heat treatments during the post-casting phase of their manufacture. In the past these bearings had not been subjected to thermal treatments, making this a unique metal-on-metal bearing which had not been used before in clinical practice.

We report the outcome of 184 consecutive hips (160 patients) treated using a hybrid-fixed metal-on-metal hip resurfacing during 1996. Patients were invited for a clinicoradiological follow-up at a minimum of ten years. The Oxford hip score and anteroposterior and lateral radiographs were obtained. The mean age at operation was 54 years (21 to 75). A series of 107 consecutive hips (99 patients) who received the same prosthesis, but subjected to a single thermal treatment after being cast, between March 1994 and December 1995, were used as a control group for comparison.

In the 1994 to 1995 group seven patients (seven hips) died from unrelated causes and there were four revisions (4%) for osteolysis and aseptic loosening. In the 1996 group nine patients died at a mean of 6.9 years after operation because of unrelated causes. There were 30 revisions (16%) at a mean of 7.3 years (1.2 to 10.9), one for infection at 1.2 years and 29 for osteolysis and aseptic loosening. Furthermore, in the latter group there were radiological signs of failure in 27 (24%) of the 111 surviving hips.

The magnitude of the problem of osteolysis and aseptic loosening in the 1996 cohort did not become obvious until five years after the operation. Our results indicate that double-heat treatments of metal-on-metal bearings can lead to an increased incidence of wear-induced osteolysis.

We carried out a prospective study of 132 patients (159 knees) who underwent closed-wedge high tibial osteotomy for severe medial compartment osteoarthritis between 1988 and 1997. A total of 94 patients (118 knees) was available for review at a mean of 16.4 years (16 to 20). Seven patients (7.4%) (11 knees) required conversion to total knee replacement. Kaplan-Meier survival was 97.6% (95% confidence interval 95.0 to 100) at ten years and 90.4% (95% confidence interval 84.1 to 96.7) at 15 years. Excellent and good results as assessed by the Hospital for Special Surgery knee score were achieved in 87 knees (73.7%). A pre-operative body mass index > 27.5 kg/m² and range of movement < 100° were risk factors predicting early failure.

Although our long-term results were satisfactory, strict indications for osteotomy are required if long-term survival is required.

Between 1966 and 2001, 1254 patients underwent excision of a bone tumour with endoprosthetic replacement. All patients who had radiotherapy were identified. Their clinical details were retrieved from their records.

A total of 63 patients (5%) had received adjunctive radiotherapy, 29 pre-operatively and 34 post-operatively. The mean post-operative Musculoskeletal Tumor Society scores of irradiated patients were significantly lower (log-rank test, p = 0.009). The infection rate in the group who had not been irradiated was 9.8% (117 of 1191), compared with 20.7% (6 of 29) in those who had pre-operative radiotherapy and 35.3% (12 of 34) in those who radiotherapy post-operatively. The infection-free survival rate at ten years was 85.5% for patients without radiotherapy, 74.1% for those who had pre-operative radiotherapy and 44.8% for those who had post-operative radiotherapy (log-rank test, p < 0.001). The ten-year limb salvage rate was 89% for those who did not have radiotherapy and 76% for those who did (log-rank test, p = 0.02).

Radiotherapy increased the risk of revision (log-rank test, p = 0.015). A total of ten amputations were necessary to control infection, of which nine were successful. Radiotherapy may be necessary for the treatment of a bone sarcoma but increases the risk of deep infection for which amputation may be the only solution.

We wished to determine whether simultaneous bilateral sequential total knee replacement (TKR) carried increased rates of mortality and complications compared with unilateral TKR in low- and high-risk patients.

Our study included 2385 patients who had undergone bilateral sequential TKR under one anaesthetic and 719 who had unilateral TKR. There were no significant pre-operative differences between the groups in terms of age, gender, height, weight, body mass index, diagnosis, comorbidity and duration of follow-up, which was a mean of 10.2 years (5 to 14) in the bilateral and 10.4 years (5 to 14) in the unilateral group.

The peri-operative mortality rate (eight patients, 0.3%) of patients who had bilateral sequential TKR was similar to that (five patients, 0.7%) of those undergoing unilateral TKR. In bilateral cases the peri-operative mortality rate (three patients, 0.4%) of patients at high risk was similar to that (five patients, 0.3%) of patients at low risk as it was also in unilateral cases (two patients, 1.0% vs three patients, 0.6%). There was no significant difference (p = 0.735) in either the overall number of major complications between bilateral and unilateral cases or between low- (p = 0.57) and high-risk (p = 0.61) patients. Also, the overall number of minor complications was not significantly different between the bilateral and unilateral group (p = 0.143).

Simultaneous bilateral sequential TKR can be offered to patients at low and high risk and has an expected rate of complications similar to that of unilateral TKR.

We investigated whether simultaneous bilateral sequential total hip replacement (THR) would increase the rate of mortality and complications compared with unilateral THR in both low- and high-risk groups of patients.

We enrolled 978 patients with bilateral and 1666 with unilateral THR in the study. There were no significant pre-operative differences between the groups in regard to age, gender, body mass index, diagnosis, comorbidity as assessed by the grading of the American Society of Anesthesiologists (ASA), the type of prosthesis and the duration of follow-up. The mean follow-up was for 10.5 years (5 to 13) in the bilateral THR group and 9.8 years (5 to 14) in the unilateral group.

The peri-operative mortality rate of patients who had simultaneous bilateral THR (0.31%, three of 978 patients) was similar to that of patients with unilateral THR (0.18%, three of 1666 patients). The peri-operative mortality rate of patients in the bilateral group was similar in high risk and low risk patients (0.70%, two of 285 patients vs 0.14%, one of 693 patients) and this was also true in the unilateral THR group (0.40%, two of 500 patients vs 0.09%, one of 1166 patients). Patients with bilateral THR required more blood transfusions and a longer hospital stay than those in the unilateral THR group. There was no significant difference (p = 0.32) in the overall number of complications between the groups. This was also true for the low-risk (p = 0.81) vs high-risk (p = 0.631) patients.

Our findings confirm that simultaneous sequential bilateral THR is a safe option for patients who are considered to be either high or low risk according to the ASA classification.

Metal-on-metal hip resurfacing is commonly performed for osteoarthritis in young active patients. We have observed cystic or solid masses, which we have called inflammatory pseudotumours, arising around these devices. They may cause soft-tissue destruction with severe symptoms and a poor outcome after revision surgery. The aim of this study was to determine the incidence of and risk factors for pseudotumours that are serious enough to require revision surgery.

Since 1999, 1419 metal-on-metal hip resurfacings have been implanted by our group in 1224 patients; 1.8% of the patients had a revision for pseudotumour. In this series the Kaplan-Meier cumulative revision rate for pseudotumour increased progressively with time. At eight years, in all patients, it was 4% (95% confidence interval (CI) 2.2 to 5.8). Factors significantly associated with an increase in revision rate were female gender (p < 0.001), age under 40 (p = 0.003), small components (p = 0.003), and dysplasia (p = 0.019), whereas implant type was not (p = 0.156). These factors were inter-related, however, and on fitting a Cox proportional hazard model only gender (p = 0.002) and age (p = 0.024) had a significant independent influence on revision rate; size nearly reached significance (p = 0.08). Subdividing the cohort according to significant factors, we found that the revision rate for pseudotumours in men was 0.5% (95% CI 0 to 1.1) at eight years wheras in women over 40 years old it was 6% (95% CI 2.3 to 10.1) at eight years and in women under 40 years it was 13.1% at six years (95% CI 0 to 27) (p < 0.001).

We recommend that resurfacings are undertaken with caution in women, particularly those under 40 years of age but they remain a good option in young men. Further work is required to understand the aetiology of pseudotumours so that this complication can be avoided.

We have reviewed 70 Harris-Galante uncemented acetabular components implanted as hybrid hip replacements with cemented stems between 1991 and 1995 in 53 patients whose mean age was 40 years (19 to 49). The mean follow-up was for 13.6 years (12 to 16) with no loss to follow-up. We assessed the patients both clinically and radiologically.

The mean Oxford hip score was 20 (12 to 46) and the mean Harris hip score 81 (37 to 100) at the final review. Radiologically, 27 hips (39%) had femoral osteolysis, 13 (19%) acetabular osteolysis, and 31 (44%) radiolucent lines around the acetabular component. Kaplan-Meier survival curves were constructed for the outcomes of revision of the acetabular component, revision of the component and polyethylene liner, and impending revision for progressive osteolysis.

The cumulative survival for revision of the acetabular component was 94% (95% confidence interval 88.4 to 99.7), for the component and liner 84% (95% confidence interval 74.5 to 93.5) and for impending revision 55.3% (95% confidence interval 40.6 to 70) at 16 years.

Uncemented acetabular components with polyethylene liners undergo silent lysis and merit regular long-term radiological review.

We reviewed 111 hemispherical Duraloc series-500 acetabular components with a minimum follow-up of 12 years. The mean clinical and radiological follow-up was 13.4 years (12 to 15). A Profile hydroxyapatite-coated anatomical femoral component was used in each case. Six patients had a late dislocation, for whom the polyethylene liner was exchanged. Each acetabular component was well fixed and all femoral components showed signs of bone ingrowth. The mean rate of femoral head penetration was 0.10 mm/year (0.021 to 0.481). The probability of not developing femoral cortical hypertrophy and proximal osteopenia by 12 years was 80.2% (95% confidence interval, 72.7 to 87.6) and 77.5% (95% confidence interval, 69.7 to 85.2), respectively. Despite these good clinical results, further follow-up is needed to determine whether these prostheses will loosen with time.

We have analysed the pattern of symptoms in patients presenting with synovial sarcoma to identify factors which led to long delays in diagnosis. In 35 children, the early symptoms and the results of clinical and radiological investigation were reviewed, along with the presumed diagnoses. The duration of symptoms was separated into patient delay and doctor delay.

Only half of the patients had one or more of the four clinical findings suggestive of sarcoma according to the guidance of the National Institute for Clinical Excellence at the onset of symptoms. Of the 33 children for whom data were available, 16 (48.5%) presented with a painless mass and in ten (30.3%) no mass was identified. Seven (21.2%) had an unexplained joint contracture. Many had been extensively investigated unsuccessfully. The mean duration of symptoms was 98 weeks (2 to 364), the mean patient delay was 43 weeks (0 to 156) and the mean doctor delay was 50 weeks (0 to 362). The mean number of doctors seen before referral was three (1 to 6) and for 15 patients the diagnosis was obtained after unplanned excision. Tumours around the knee and elbow were associated with a longer duration of symptoms and longer doctor delay compared with those at other sites. Delays did not improve significantly over the period of our study of 21 years, and we were unable to show that delay in diagnosis led to a worse prognosis.

Our findings highlight the variety of symptoms associated with synovial sarcoma and encourage greater awareness of this tumour as a potential diagnosis in childhood.

We present the early clinical and radiological results of Articular Surface Replacement (ASR) resurfacings in 214 hips (192 patients) with a mean follow-up of 43 months (30 to 57). The mean age of the patients was 56 years (28 to 74) and 85 hips (40%) were in 78 women.

The mean Harris hip score improved from 52 (11 to 81) to 95 (27 to 100) at two years and the mean University of California, Los Angeles activity score from 3.9 (1 to 10) to 7.4 (2 to 10) in the same period. Narrowing of the neck (to a maximum of 9%) was noted in 124 of 209 hips (60%). There were 12 revisions (5.6%) involving four (1.9%) early fractures of the femoral neck and two (0.9%) episodes of collapse of the femoral head secondary to avascular necrosis. Six patients (2.8%) had failure related to metal wear debris. The overall survival for our series was 93% (95% confidence interval 80 to 98) and 89% (95% confidence interval 82 to 96) for hips with acetabular components smaller than 56 mm in diameter.

The ASR implant has a lower diametrical clearance and a subhemispherical acetabular component when compared with other more frequently implanted metal-on-metal hip resurfacings. These changes may contribute to the higher failure rate than in other series, compared with other designs. Given our poor results with the small components we are no longer implanting the smaller size.