Background

Medical complications and death are rare events following elective orthopaedic surgery. Diagnostic and operative codes are routinely collected on every patient admitted to hospital in the English NHS (hospital episode statistics, HES). This is the first study investigating rates of these events following total joint replacement (TJR) on a national scale in the NHS.

Patient-reported outcome measures (PROMs) are increasingly being used to assess functional outcome and patient satisfaction. They provide a framework for comparisons between surgical units, and individual surgeons for benchmarking and financial remuneration. Better performance may bring the reward of more customers as patients and commissioners seek out high performers for their elective procedures. Using National Joint Registry (NJR) data linked to PROMs we identified 22 691 primary total knee replacements (TKRs) undertaken for osteoarthritis in England and Wales between August 2008 and February 2011, and identified the surgical factors that influenced the improvements in the Oxford knee score (OKS) and EuroQol-5D (EQ-5D) assessment using multiple regression analysis. After correction for patient factors the only surgical factors that influenced PROMs were implant brand and hospital type (both p < 0.001). However, the effects of surgical factors upon the PROMs were modest compared with patient factors. For both the OKS and the EQ-5D the most important factors influencing the improvement in PROMs were the corresponding pre-operative score and the patient’s general health status. Despite having only a small effect on PROMs, this study has shown that both implant brand and hospital type do influence reported subjective functional scores following TKR. In the current climate of financial austerity, proposed performance-based remuneration and wider patient choice, it would seem unwise to ignore these effects and the influence of a range of additional patient factors.

Introduction

Unlike the NJR, no surgeon driven national database currently exists for ligament surgery in the UK and therefore information on outcome and adverse event is limited to case series.

Introduction

The National Institute for Health and Clinical Effectiveness recommends both low molecular weight heparin (LMWH) and Rivaroxaban for venous thromboembolic (VTE) prophylaxis following lower limb arthroplasty. Despite evidence in the literature that suggests Rivaroxaban reduces VTE events, there are emerging concerns from the orthopaedic community regarding an increase in wound complications following its use.

Background

Following an anterior cruciate ligament (ACL) injury, the affected knee is known to experience bone loss and is at significant risk of becoming osteoporotic. Surgical reconstruction is performed to attempt to restore the function of the knee and theoretically restore this bone density loss. Cross-sectional analysis of the proximal tibia using peripheral quantitative computed tomography (pQCT) enables localised analysis of bone mineral density (BMD) changes. The aim of this study was to establish the pattern of bone density changes in the tibia pre- and post- ACL reconstruction using pQCT image analysis.

Bearing diameter and acetabular component orientation have been shown to be important variables effecting blood metal ion levels following hip resurfacing arthroplasty. So far no studies on bilateral hip resurfacings have taken into account these variables.

We examined the serum ion results of patients under the care of two experienced hip resurfacing surgeons who carry out ion analysis as part of routine post operative care. Surgeon 1: Patients were implanted exclusively with a “third generation” resurfacing device. Surgeon 2: Patients were implanted with the same “third generation” device and also a low clearance “fourth generation” resurfacing device. Only ion results from patients who were 12 months post surgery were included. Bilateral patients were matched to unilateral patients according to the surgeon performing the operation; the resurfacing system implanted and cup inclination and anteversion angles. The ion data from each bilateral group was tested against the corresponding unilateral groups using the Mann Whitney U test for non parametric data. Significance was drawn at p<0.05.

Surgeon 1: There were 310 patients with unilateral joints and 50 with bilateral joints. There were no significant differences with regard to time to follow up, activity levels, joint sizes or cup orientations. Serum chromium (Cr) and cobalt (Co) concentrations were significantly greater in the bilateral group (p<0.001). Median ion levels were greater in the bilateral group by a factor of >2 in the smallest joint sizes and <2 in the largest joints. Surgeon 2: There were 11 patients with bilateral third generation resurfacing joints and 50 with unilateral joints of the same design. The same relationship as described above was identified. There was a notable difference in the fourth generation implant group (n=13 bilateral, n=100 unilateral). Median ion levels for patients receiving bilateral joints of sizes <47mm were ten times greater than in the corresponding unilateral group.

Bearing diameter and component design are critical factors in determining metal ion levels following bilateral hip resurfacings. Surgeons must consider the potential implications of gross increases in metal ion levels prior to performing bilateral hip resurfacings in smaller patients.

Background

Total Knee Replacement (TKR) is technically demanding, time consuming and has higher complication rates in super obese (BMI>45) patients. Bariatric surgery can be considered for such patients prior to TKR although its effect on complications is unknown.

From August 2009, all doctors were subject to the European Working Time Directive (EWTD) restrictions of 48 hours of work per week. Changes to rota patterns have been introduced over the last two years to accommodate for these impending changes, sacrificing ‘normal working hours’ training opportunities for out-of-hours service provision. We have analysed the elogbook data to establish whether operative experience has been affected.

A survey of trainees (ST3-8) was performed in February 2009 to establish shift patterns in units around the UK. All operative data entered into the elogbook during 2008 at these units was analysed according to type of shift (24hr on call with normal work the following day, 24hr on call then off next working day, or shifts including nights).

66% of units relied on traditional 24hrs on call in February 2009. When compared with these units, trainees working shifts had 18% less operative experience (564 to 471 operations) over the six years of training, with a 51% reduction in elective experience (288 to 140 operations). In the mid years of training, between ST3-5, operative experience fell from 418 to 302 operations (25% reduction) when shifts were introduced.

This national data reflects the situation in UK hospitals in 2009, prior to the implementation of a maximum of 48 hours. It is expected that most hospitals will need to convert to shift-type working patterns to fall within the law. This could have significant implications for elective orthopaedic training in the UK.

Background: Diagnostic and operative codes are routinely collected on every patient admitted to National Health Service (NHS) hospitals in England and Wales (hospital episode statistics, HES). The data allows for linkage of post-operative complications and primary operative procedures, even when patients are re-admitted following a successful discharge. Morbidity and mortality data on foot and ankle surgery (F& A) has not previously been available in large numbers for NHS patients.

Methods: All HES data for a 44-month period prior to August 2008 was analysed and divided into four groups – hindfoot fusion, ankle fracture surgery, ankle replacement and a control group. The control group was of first metatarsal osteotomy, which is predominantly day case surgery where no above ankle cast is used. The incidence of pulmonary embolism (PE) and all cause mortality (MR) within 90 days, and a return to theatre (RTT, as a complication of the index procedure) within 30 days was calculated for each group.

Results: 7448 patients underwent a hindfoot fusion. PE, RTT and MR were 0.11%, 0.11% and 0.12% respectively. 58732 patients had operative fixation of an ankle fracture. PE, RTT and MR were 0.16%, 0.08% and 0.35%. 1695 patients had an ankle replacement. PE, RTT and MR were 0.06%, 0.35% and zero. 35206 patients underwent a first metatarsal osteotomy. PE, RTT and mortality rates were 0.02%, 0.01% and 0.03%.

Discussion: There is controversy regarding the use of venous thrombo-embolic (VTE) prophylaxis in foot and ankle surgery. Non-fatal PE in F& A surgery has previously been reported as 0.15%. NICE guidelines recommend low molecular weight heparin (LMWH) for all inpatient orthopaedic surgery. 94% of F& A surgeons prescribe LMWH to post operative elective inpatients in plaster according to a previous British Orthopaedic foot and ankle society survey. VTE events, RTT and mortality rates for all groups were extremely low, including inpatient procedures requiring prolonged immobilisation. We question the widespread use of LMWH.

Background: There is a paucity of published data with regard to the wear of failed metal on metal (MoM) resurfacing devices.

Materials and Methods: MoM components retrieved from patients from two independent centres experiencing failure secondary to ARMD were analysed using a Mitutoyo Legex 322 coordinate measuring machine (CMM) which has an accuracy of 0.8 microns. Between 4000 – 6000 points were taken on each explant, dependent on the size of the bearing surface. Maximum wear depths and total volumetric wear were calculated. These values were compared to those from control samples retrieved following uncomplicated fractures/femoral collapse secondary to avascular necrosis (after calculating equivalent yearly wear rates).

Results: 58 ARMD components were analysed. This included 22 36mm MoM THRs (DePuy Pinnacle), 28 DePuy ASRs and 8 Zimmer Duroms. There were 30 resurfacing fracture/avascular necrosis controls. Volumetric wear rates and maximum wear depths of ARMD resurfacing components were significantly greater than the resurfacing control group for both the ASR and Duroms (p< 0.05) however 2 ARMD components exhibited similar amounts of wear compared to controls. Wear rates of the ARMD THR group were significantly lower than the ARMD resurfacing group (p< 0.05).

Conclusions: Increased articular wear is associated with an increased incidence of local adverse effects including tissue necrosis, joint effusions and fractures. However, there are a minority of patients who can develop tissue necrosis in the absence of accelerated wear, implying a spectrum of sensitivity This is reflected in the incidence of ARMD in the patient groups at the main study centre: > 5% in the ASR group and approx 1% in the THR group. We believe this indicates a failure of adequate lubrication and the resultant negative effects in larger bearing devices.

Background: Blood metal ion levels have been shown to correlate with wear of retrieved components. Increased articular wear is associated with early failure. It is not clear what the management of patients with raised metal ions who remain asymptomatic should be.

Methods: A prospective study of the DePuy ASR resurfacing device was commenced in 2004 at our independent centre. Blood and serum metal ion analysis is carried out routinely. We assessed the clinical outcome of all patients with cobalt (Co) concentrations > 5μg/L (n=25). We consider this value to be a clear indicator of a poorly performing bearing surface. Patients who remained asymptomatic were brought back for repeat ions testing and clinical assessment. A parallel study was carried out to examine the relationship between the wear depth of retrieved explants (assessed using a coordinate measuring) and blood ion levels

Results: Increases in chromium(Cr) and Co levels from the second blood sample correlated with the first sample (p< 0.05). The four patients with the highest Cr Co levels went on to suffer spontaneous femoral neck fractures between 3 and 4 years post op (pseudotumours in two of these cases). Of the remaining 25, 5 were revised for worsening pain (large effusions and areas of tissue necrosis were seen at revision) and 10 were under investigation for new onset of pain. Wear depths correlated well with blood ion concentrations (p< 0.05) substantiating the results of another centre.

Conclusion: 60% of patients with raised metal ion levels went on to develop complications within two years of the blood test. This suggests patients may only have temporary tolerances to excessive concentrations of metal wear debris. One novel association appears to be spontaneous delayed femoral fracture. Results suggest increased ion concentrations associated with small joint size/sub optimal cup orientation will continue to rise and these patients should be followed up closely.

Background: Definitive cup position for the reduction of blood metal ion levels has yet to be established.

Methods: Samples for serum metal ion analysis are taken routinely from patients under the care of the two senior authors of this paper. Both are high volume experienced hip resurfacing surgeons, one based in England, the other in Belgium. Metal ion results from two centres from patients with unilateral joints were correlated to size and orientation of femoral and acetabular components, UCLA activity score, age, time post surgery and post operative femoral head/neck ratios. EBRA software was used to assess cup inclination and anteversion on standing radiographs.

Cup orientation in vivo was compared to explant analysis of 60 retrieved resurfacing components using a coordinate measuring machine.

Results: Three resurfacing devices were studied. There were 620 results in total. Only femoral size and cup inclination/anteversion were found to have any effect on ion levels. In all devices, metal ion levels were inversely related to femoral size (p< 0.05). The device providing the smallest acetabular coverage arc was associated with the highest metal ion levels. Consistent throughout the implants, lowest ion levels were associated with cups with radiological inclination of 40–50° and anteversion 10–20°. Cup inclination angles lower than 40° were associated with posterior edge loading and likely sub-luxation of the femoral component.

Conclusion: The greater the coverage angle provided by the acetabular component, the greater the tolerance to suboptimal position. Lowest ion levels were found in well positioned lower clearance devices. Cup inclination/ante-version angles of 45/15+/−5° were associated with low ion levels in all three devices. We do not recommend cups to be placed with inclination angles below 40°.

In our independent centre, from 2002 to 2009, 155 BHRs (mean F/U 60 months) have been implanted as well as 420 ASR resurfacings and 75 THRs using ASR XL heads on SROM stems (mean F/U 35)

During this period we have experienced a number of failures with patients complaining of worsening groin pain at varying lengths of time post operatively. Aspiration of the hip joints yielded a large sterile effusion on each occasion. At revision, there were copious amounts of green grey fluid with varying degrees of necrosis. There were 17 failures of this nature in patients with ASR implants (12 females) and 0 in the BHR group. This amounts to a failure of 3.5% in the ASR group.

Tissue specimens from revision surgery showed varying degrees of “ALVAL” as well as consistently high numbers of histiocytes. Particulate metal debris was also a common finding.

The mean femoral size and acetabular anteversion and inclination angles of the ARMeD group/all asymptomatic patients was 45/49mm (p< 0.001), 27/20°(p< 0.001) and 53/48°(p< 0.08). Median blood chromium(Cr) and cobalt(Co) was 29 and 69 μg/L respectively in the ARMeD group versus 3.9 and 2.7 μg/L in the asymptomatic patients (n=160 with ion levels). Explant analysis confirmed greater rates of wear than expected.

Lymphocyte proliferation studies involving ARMeD patients showed no hyper reactivity to Cr and Co in vitro implying that these adverse clinical developments are mediated by a toxic reaction or a localised immune response.

Our overall results suggest that the reduced arc of cover of the fourth generation ASR cup has led to an increased failure rate secondary to the increased generation of metal debris. This failure rate is 7% in ASR devices with femoral components _47mm.

Introduction: In August 2007, the Department of Health initiative Modernising Medical Careers was implemented. This was a system of reform and development in postgraduate medical education and training. In preparation for the changes, the SAC for T& O outlined a new curriculum. The emphasis of early training, StR years 1 and 2, was to be trauma. We aim to identify how effectively the SAC proposals are being applied, and what difference this makes to the trainees’ operative experience? Furthermore, how do the new posts compare to the historic SHO models?

Methods: A survey carried out by BOTA allowed us to assess post compliance with the SAC recommendations. A compliant job was defined as trauma based for 50% or more of working time. Consent was obtained to evaluate the eLogbooks of trainees in compliant and non-compliant jobs, along with registrars who had previously held traditional SHO grade posts. Overall operative experience over a specified 4 month time period was examined, with focus on routine trauma procedures.

Results: The results of the BOTA and SAC survey revealed that 45% of the new orthopaedic posts were compliant with curriculum guidelines. The eLogbooks of 92 individuals were analysed; 28 historical posts, 34 compliant and 30 non-compliant. The mean total number of recorded entries by trainees in the 4 month period was 73.2 in the historic group, 90.5 in the compliant and 87.3 in the non-compliant job group. The corresponding numbers of trauma operations were 35.7, 48.4 and 41.5.

Conclusions: Operative experience has improved since the introduction of the new curriculum. The new posts are offering more operative and in particular trauma exposure than traditional SHO jobs. If jobs can be restructured such that they all comply with the SAC, educational opportunities in the early years will be maximised.

Background: From August 2009, all doctors were subject to the European Working Time Directive (EWTD) restrictions of 48 hours of work per week. Changes to rota patterns have been introduced over the last two years to accommodate for these impending changes, sacrificing ‘normal working hours’ training opportunities for out-of-hours service provision. We have analysed the elogbook data to establish whether operative experience has been affected.

Methods: A survey of trainees (ST3-8) was performed in February 2009 to establish shift patterns in units around the UK. All operative data entered into the elogbook during 2008 at these units was analysed according to type of shift (24hr on call with normal work the following day [traditional on call], 24hr on call then off next working day, or shifts including nights).

Results: 66% of units relied on traditional 24hrs on call in February 2009. When compared with these units, trainees working shifts had 18% less operative experience (564 to 471 operations) over the six years of training, with a 51% reduction in elective experience (288 to 140 operations). In the mid years of training, between ST3-5, operative experience fell from 418 to 302 operations (25% reduction) when shifts were introduced.

Discussion: The eLogbook is a powerful resource that provides accurate data for the purpose of supporting orthopaedic training. This national data reflects the situation in UK hospitals in 2009, prior to the implementation of a maximum of 48 hours. It is expected that most hospitals will need to convert to shift-type working patterns to fall within the law. This could have significant implications for elective orthopaedic training in the UK.

Background: Diagnostic and operative codes are routinely collected on every patient admitted to hospital in England and Wales (hospital episode statistics, HES). Linked data allows post-operative complications to be associated with the primary operative procedure, even if patients are re-admitted following a successful discharge. Morbidity and mortality data on shoulder surgery have not previously been available in large numbers.

Methods: All HES data for a 42-month was analysed and divided into three groups – elective shoulder replacement (total or hemiarthroplasty), shoulder arthroscopy (all procedures), and proximal humerus fracture surgery (internal fixation or replacement). Incidence of pulmonary embolism (PE), deep venous thrombosis (DVT) and mortality within 90 days was established.

Results: For elective shoulder replacement (10735 patients), 90-day DVT, PE and mortality rates were 0.07%, 0.11% and 0.36% respectively. Mortality in patients over 75 years was 0.9%. For arthroscopic procedures (66344 patients), 90-day DVT, PE and mortality rates were 0.01%, 0.01% and 0.03%. For proximal humerus fracture surgery (internal fixation or replacement, 4968 patients) 90-day DVT, PE and mortality rates were 0.20%, 0.38% and 2.98%. Mortality in patients over 75 years old was 6.6%.

Discussion: Venous thromboembolic (VTE) prophylaxis is rarely used for upper limb surgery. PE and mortality rates for shoulder replacement and proximal humerus fracture surgery are lower those for patients receiving chemical prophylaxis after hip replacement. Further investigation into the cause of high mortality rates following fracture surgery in patients over 75 years old is required. VTE prophylaxis may be required in this age group.

Since 2003 Trauma and Orthopaedic trainees in the UK and Ireland have routinely submitted data recording their operative experience electronically via the eLog-book. This provides evidence of operative experience of individuals and national comparisons of trainee, trainer, hospital and training programme performance. We have analysed trauma surgery data and established standards for training.

By January 2008 there were over 4 million operations logged. Operations performed and uploaded since 2003 have been included. Each trainee’s work is analysed by ‘year-in-training’. Data on levels of supervision, missed opportunities (where the trainee assisted rather than performed the operation) was analysed. The average number of trauma operations performed annually by trainees was 109, 120, 110, 122, 98 and 84 (total 643) for YIT one (=ST3) to six (=ST8) respectively. There were only 22% of missed opportunities throughout six years of training. A high level of experience is gained in hip fracture surgery (121 operations) and forearm (30), wrist (74) and ankle (47) operative stabilisation over the six years. However, the average number of tibial intra-medullary nails (13), external fixator applications (12) and childrens’ elbow supracondylar fracture procedures (4) performed is low. We are also able to identify trainees performing fewer operations than required during their training (two standard deviations or more below the mean for their YIT). We expect a trainee to have performed at least 255, 383, 473, and 531 trauma operations at the end of YIT three to six respectively.

The eLogbook is a powerful tool which can provide accurate information to support in-depth analysis of trainees, trainers, and training programmes. This analysis has established a baseline which can be used to identify trainees who are falling below the required operative experience.

Tarsal fractures are rare in children. Clinical and radiographic evaluation of these injuries can be difficult. We present a retrospective study documenting all tarsal fractures presenting to an inner-city children’s hospital in the UK over a fifteen year period.

Of 70 case notes retrieved from the hospital database, 7 patients were excluded due to inadequate data. This resulted in 69 tarsal injuries in 63 patients being included. Mean age at presentation was 9.3 years (2.5 – 13.9). 80% were male. 72% were calcaneal fractures, 12% cuboid, 9% navicular, 4% talus and 2% medial cuneiform. The main method of diagnosis was plain x-rays. Cause of injury was predominantly fall from height, crush or road traffic accident. 25% had another associated lower extremity injury. Three patients had bilateral tarsal injuries. Only 3% had upper limb injuries and there were no injuries with spinal involvement. Calcaneal fractures were treated with a short leg cast for a mean time of 4.1 weeks (2–6). Mean time to recovery was 5.7 weeks (2–20). Mean time to discharge from clinic was 7 weeks (2–40). There were two patients with open fractures requiring surgical debridement. One patient with a talar fracture had percutaneous fixation. Only one patient re-presented with pain following discharge. X-rays showed healing avascular necrosis of the proximal talus.

Tarsal fractures are rare, usually benign and most require simple immobilisation for only a short period of time. Surgical intervention is only occasionally required in complex injuries. Complications and long term problems are rare, even following open injuries.

This report documents the clinical and histological findings, the operative management and the explant analyses of patients with early aseptic failure of large metal-on-metal (MonM) bearing joints.

Three hundred and fifty patients have been implanted with the ASR bearing surface (resurfacing or a modular THR) by a single surgeon at an independent centre since 2004. Six patients (all female) have been revised secondary to aseptic failure. All complained of severe groin pain exacerbated by straight leg raise and routine investigations were unable to establish a cause. Large amounts of sterile, highly viscous green fluid were aspirated from the hip joints in each case. Gross swelling of the pseudocapsule and a similar green fluid surrounding the implants were found at revision surgery. Histological examination of periprosthetic tissue samples showed changes consistent with ALVAL/metallosis, and analysis of the fluid revealed dense numbers of inflammatory cells. Symptoms in patients revised to ceramic-on-ceramic bearings improved post operatively. This was not the case with those reimplanted with MoM joints. Data from a subset of 76 patients (all unilateral resurfacings) showed that malaligned cups (anteverted > 20° +/− inclination angle > 45°) were associated with significantly higher whole blood metal ion levels than cups positioned within this range. All the patients with early aseptic failure had malaligned cups. Independent explant analysis revealed significant increases in the surface roughness values of the articular surfaces.

Our results suggest that some patients develop a significant inflammatory reaction to metal implants. This may be due to high levels of metal debris around the implant as a result of accelerated wear. Poor component alignment was found in all our patients with early aseptic failure.

The first United Kingdom In-training Examination (UKITE) was held in 19 training programmes during December 2007. The aims of the project are to offer a national, online examination (providing immediate results to trainees) and to allow practice for the ‘real’ FRCS T& O examination with similarly formatted questions based on the UK T& O curriculum. All Speciality training years (StR2, StR3 and above, and all SpRs) and all deaneries will participate in the future. A total of 450 trainees sat this first examination. This is an online exam (accessed through the eLogbook/OCAP website) which is voluntary and has no bearing on RITA outcomes. To take part in the exam trainees were asked to provide 3 questions for a bank which can be used in subsequent years for both UKITE and the FRCS T& O examination.

The exam was 3 hours and questions were of multi-choice (MCQ) and extended matching question (EMQ) style covering all aspects of orthopaedics. Correct answers and explanations were available to the trainee after their answer had been submitted. Final scores ranged from 35% to 92%. Mean scores increased from 48% (StR2) to 73% in SpR year 5 (StR7 equivalent). This was followed by a drop off in performance in the final year of training. Three candidates had equal top scores at 92%. 97% stated they would sit the UKITE again and 93% felt there was educational benefit. The length and level of difficulty of the exam was felt to be satisfactory. Improvements were suggested for question quality. The UKITE is a powerful tool for self-assessment of trainees. This analysis establishes a baseline for future years.