Introduction

Hip and knee replacements are predictable orthopaedic procedure with excellent clinical outcomes. Discharging or leaking wounds affect length of hospital stay, affect bed planning and predispose to superficial and potentially deep wound infection. Predictable wound healing therefore remains the first hurdle. This trial aims to study the effectiveness of portable disposable incisional negative pressure wound therapy (NPWT) dressings in hip and knee replacements.

This trial aims to study the effectiveness of portable disposable incisional negative pressure wound dressings in hip and knee replacements and the impact on wound healing, length of stay and wound complications.

To discuss the rationale, selection criteria, indications, and results of using large diameter ceramic heads in revision hip arthroplasty.

We routinely use Biolox family of ceramic heads and acetabular liners in patients undergoing revision total hip replacements. We present our experience in using ceramic articular bearings over the last 20 years and the switch to larger diameter ceramic heads. We also present our rationale for using a large diameter ceramic head instead of a large metal head.

We reviewed a total of 689 revision arthroplasties over this time period and we report the outcome of large bearing couples with case examples in primary and revision scenarios. Furthermore we compared a subset of patients (110) with large diameter ceramic heads – Biolox Delta 36mm to patients who had metal on metal (large head 42 mm and above) bearing couples. The performance of the ceramic bearing couples will be discussed along with the functional outcome of these patients. We found no difference in the functional, clinical sports activities (UCLA and Tegner scores) between patients who had large metal bearing couples and large ceramic couples. Complication rate was less with the ceramic bearing revision arthroplasties, as was patient satisfaction.

Ceramic bearing couples have stood the test of time and have demonstrated an excellent long term wear properties. The recent introduction of the large diameter couples proves to be an excellent alternative if not the first choice in young, complex primary and revision case scenarios.

Introduction: The Optiplug^® bio-absorbable cement restrictor, marketed by Biomet inc., is manufactured from ‘PolyActive’ – a polymer of poly(ethylene glycol) and butylene terephthalate. Biodegradation is thought to be by a combination of hydrolysis and oxidation.

The potential benefit – eliminating the need for restrictor removal at future revision surgery – led to Optiplug becoming our cement restrictor of choice over the last 5 years.

Anecdotally we have seen marked osteolysis around the distal cement mantle in a number of follow up radiographs in these patients. To date we have not seen an associated peri-prosthetic fracture.

We undertook a retrospective, radiographic study to determine incidence, severity and progression of this osteolysis over the first 5 years of follow up.

Method: 100 patients for whom 5 year follow up had been undertaken were identified from the departmental database. Patients with loose prostheses and or infection were excluded as were those who had undergone revision surgery.

Radiographs from the immediate post operative period, twelve months and five years follow up visits were identified and reviewed.

Osteolysis was quantified by calculating the ratio of maximum medullary diameter to the overall cortical diameter of the bone. Comparison was made over time and, where radiographs allowed, to the immediately adjacent femur.

Results: 87% of radiographs showed greater than 10% thinning of the cortex at 1 year cf. immediate post op. 5 cases showed greater than 33% thinning. These changes do not appear to progress or regress significantly between 1 and 5 years.

Discussion: While marked osteolytic changes appear to be uncommon, some degree of cortical thinning was almost universal in this series. The zone immediately distal to the cement mantle is commonly involved in peri-prosthetic fractures. Any weakening in this area is undesirable.

Background: There are no evidence based guidelines on the surveillance of cemented total hip arthroplasty. We reviewed the outcomes of those patients undergoing this procedure in 1996 & 1997.

Methods: The patients were identified from theatre log books. The follow up date was then retrieved from the electronic patient record system used at our institution. From these we recorded the age, sex, side of procedure, evidence of radiological loosening & time of revision surgery. The data with regards to radiological evidence of loosening & revision surgery were then analysed using a ® statistical software package. From this we were able to plot Kaplan-Meier survival & hazard plots.

Results: We identified 425 primary total hip arthroplasties. Using radiological evidence of loosening as the end point we found that there was an initial peak and a peak at 8 years. There was a 10 year survival rate of 85.8%. Using revision surgery as the end point we found that there was, again, an initial peak & a peak at 8 years. There was a 10 year survival rate of 91.5%.

Conclusions: Once the patient has made it through the first post operative year they do not need to be followed up again until 8 years assuming they remain asymptomatic.

Background: There are no evidence based guidelines on the surveillance of cemented total hip arthroplasty. We reviewed the outcomes of those patients undergoing this procedure in 1996 & 1997.

Methods: The patients were identified from theatre log books. The follow up date was then retrieved from the electronic patient record system used at our institution. From these we recorded the age, sex, side of procedure, evidence of radiological loosening & time of revision surgery. The data with regards to radiological evidence of loosening & revision surgery were then analysed using a XXX statistical software package. From this we were able to plot Kaplan-Meier survival & hazard plots.

Results: We identified 425 primary total hip arthroplasties. Using radiological evidence of loosening as the end point we found that there was a peak initially and a peak at 8 years. There was a 10 year survival rate of 85.8%. Using revision surgery as the end point we found that there was, again, an initial peak & a peak at 8 years. There was a 10 year survival rate of 91.5%.

Conclusions: Once the patient has made it through the first post operative year they do not need to be followed up again until 8 years assuming they remain asymptomatic.

Introduction: Primary total hip replacement remains one of the commonest orthopaedic procedures performed. It is yet to be clearly demonstrated whether use of a postoperative drain is of benefit in these procedures.

Methods: We carried out a prospective randomised study comparing the use of autologous reinfusion drains, closed suction drains or no drain to determine their influence on allogenic blood transfusion requirements, length of hospital stay and infection rates. Stratification was carried out for confounding factors.

Results: 153 patients were recruited into the study and randomised to one of the three closely matched groups. There was no significant difference between the mean intra-operative blood loss or post-operative haemaglo-bin levels between the 3 groups. 42% of the suction drain group required post-operative transfusion as compared to 17% of the reinfusion drain group and 12% of the group with no drains. This difference was highly significant (P=0.02) Mean time for the wound to become dry was 3 days, 3.9 days and 4 days in the no drain, re-transfusion drain and suction drain groups respectively. This difference was statistically significant (P=0.03). There was no statistically significant difference in the mean length of inpatient stay.

Discussion: This study demonstrates a significantly higher transfusion rate with closed suction drains compared to reinfusion drains or no drains. With the drive to reduce hospital stay our study supports the considered use of no drain or a reinfusion drain.

Objective: The adequacy of the cement mantle around various designs of impaction-grafted stems has been compared and deemed inadequate around the Exeter system. Yet, good clinical results have been reported. The conventional wisdom of solid cement mantles has been also been questioned in recent reports by the low migration and high survival rates of stems inserted with a very thin cement mantle – the so called ‘French paradox’. We performed this study specifically to address two questions

Does cement mantle thickness affect cement penetration depth during impaction grafting? and

Materials and Methods: 12 composite femurs were prepared to mimic cavitary defect. Impaction grafting was done with morcellized freshly frozen porcine femoral condyles using Exeter X-change system. The size of tamp and prosthesis were independently varied creating tamp/stem mismatch to produce cement mantles with a nominal thickness of 0, 1, 2, 3 or 4 mm. Cyclical loading was done at 1 Hz for 2500 cycles at 2500 N. From the displacement data measured by 6 linear displacement transducers we calculated subsidence and retroversion. The solid cement mantle and the penetration depth into the graft were then measured along 16 points in each cut section of the femurs done at 1.5 cm intervals.

Results: There was a high correlation between tamp/stem mismatch (nominal mantle thickness) and actual mantle thickness (r=0.84). Average cement penetration into the graft for each prosthesis varied between 0.3 and 2.0 mm. Largest variations were proximally, where average penetration varied between 0.4 and 3.5 mm. A thicker solid cement mantle gave on average less cement penetration (r=−0.62). Stem subsidence after cyclic loading ranged from 0.4 to 2.5 mm and correlated significantly with tamp size (r=0.59, p< .05). However, better correlations were found with solid mantle thickness (r=0.90, p< 0.05) and cement penetration depth (r=−0.81). Stem retroversion after cyclic loading ranged from 0.1 to 2.0 degrees and correlated negatively with stem size (r=−0.53) but did not correlate with tamp size. Correlations with solid mantle thickness and cement penetration depth were not better than those with tamp size.

Discussion: Our study shows that a thinner mantle is associated with deeper cement penetration into the graft. This probably is due to the higher cement pressure generated during stem insertion when there is less space for the cement to escape. Better mechanical interlock with the higher cement penetration possibly explains the reduced subsidence with thin cement mantles. Our study also shows that stem retroversion is associated with stem size only, and is larger for thinner stems. This could be explained by thinner stems providing less resistance to torsional forces.

The adequacy of cement mantles around some impaction-grafting systems has been criticised yet good clinical results have been reported. This study investigates this contradiction by asking

Does cement mantle thickness affect cement penetration depth?

Twelve artificial femora were prepared to simulate cavitary defects. Porcine cancellous bone was morselized. The defect was reconstructed by impaction grafting, using a size 0, 1 or 2 tamp. Bone cement was injected, and a size 0, 1 or 2 Exeter stem inserted. By using all nine tamp/prosthesis combinations, 0–4 mm thick cement mantles were produced. Femora were positioned in a testing machine and loaded with 2500 cycles of 2500 N. Prosthesis subsidence and retroversion were measured. Each femur was sliced transversely and the sections digitised. Solid cement mantle thickness and cement penetration depth were measured using image analysis. Correlation analysis was used to find if tamp/stem mismatch (nominal mantle thickness) influenced actual solid mantle thickness and cement penetration. We then analysed if tamp size, stem size, solid mantle thickness or cement penetration determined stem subsidence and retroversion.

Cement mantles were produced with an average thickness of 1.7–2.2 mm, with largest variations proximally (1.5–2.8 mm). Average cement penetration was 0.3–2.0 mm, with largest variations proximally (0.4–3.5 mm). Thicker solid mantles gave less penetration (r=−0.62). Stem subsidence ranged from 0.4–2.5 mm and correlated significantly with tamp size (r=0.59, p< .05). Better correlations were found with solid mantle thickness (r=0.90, p< 0.05) and cement penetration depth (r=−0.81). Stem retroversion ranged from 0.1–2.0 degrees and correlated with stem size (r=−0.53) but not with tamp size.

Tamp/stem mismatch determined the thickness of the solid cement mantle around impaction-grafted stems, and thinner mantles were associated with deeper cement penetration. Thinner mantles and deeper penetration were associated with reduced stem subsidence. Stem retroversion was associated with stem size only, and larger for thinner stems. Thinner cement mantles will therefore be associated with deeper penetration and reduced stem subsidence upon loading. This association may explain the good long-term results of impaction-grafted Exeter stems, despite deficient solid cement mantles.

Introduction: etiology of late infection after arthroplasty can be difficult to establish. Histology is the gold standard for infection in patients without inflammatory arthritis but diagnosis in inflammatory arthritis depends on culture (Atkins et al). Real-time PCR offers a rapid and direct assessment for staphylococci and enterococci infection but has not been widely assessed.

The aims of this study were

to develop the Roche lightcycler Staphylococcal and Enterococcal PCR kits to facilitate diagnosis of hip and knee prosthetic infections

Methods: uplicate, multiple tissue samples were taken (with separate sterile instruments) at the 1^st stage of revision after informed consent. One set were cultured and results interpreted by the Oxford criteria. The second set were extracted using the Qiagen DNA kit, purified (in-house method) and tested using the Roche lightcycler kits.

Results:53 patients undergoing 2 stage revision for suspected infection were recruited.15 (28.3%) had negative histology and no inflammatory arthritis; 3 with single positive cultures and negative PCR – considered contaminants.

29 patients had non-inflammatory arthritis. 14/18 (77.8%) with positive cultures had staphylococci +/or enterococci isolated and 10 PCR results correlated. The other 11 patients had negative cultures.

9 patients had inflammatory arthritis. Six were culture negative and of the other three, 2 were positive for staphylococci on culture with 1 positive by PCR.

Discussion: Negative staphylococcal PCR correlates with the isolation of staphylococci from only one sample. This agrees with the Oxford criteria that such samples may be considered contaminants. Additional positives detected by staphylococcal PCR alone are rare.

Enterococcal PCR confirmed culture positivity in 2/3 patients. An additional 5 positive PCR’s were obtained from patients’ culture negative for enterococci. It is not clear if these are false positives or more sensitive detection of enterococcal isolation.

Purpose: To investigate whether cement mantle thickness influence early migration of the stem after impaction grafting

Methods: Twelve artificial femora were prepared to mimic cavitary defects. After compacting morselized bone into the cavities, Exeter stems were cemented in place. By using all combinations of three sizes tamps and stems (0, 1 and 2), we created cement mantles of 0, 1, 2, 3 and 4 mm thickness. Bones with stems were placed in a testing machine and loaded cyclically to 2,500 N while measuring stem migration. Statistical analysis was by regression analysis. Outcomes were stem subsidence and retroversion, predictors were mantle thickness, tamp size and stem size.

Results: Average stem subsidence after 2500 cycles when using size 1 tamp and stem (2 mm mantle) was 0.94 mm. Cement mantle thickness significantly influ-enced stem subsidence (r=0.68, p=0.015). For a 0 mm mantle, subsidence was 0.59 mm and for a 4 mm mantle it was 2.54 mm. Cement mantle thickness also signifi-cantly influenced stem retroversion (r=0.62, p=0.031). Cement mantle thickness was a better predictor than tamp or stem size.

Discussion: Concern exists that inadequate cement mantles may affect stability of impaction-grafted stems. In our study, larger difference between tamps and stems gave substantially more subsidence and rotation, whereas a smaller difference reduced them. Concerns over thin mantles may have been premature.

Introduction: The failure rate of cemented hip replacements is about 1% per year, mainly due to aseptic loosening. PMMA acts as a grout, therefore high pressure is needed to ensure fixation. Various plug designs are used to increase pressure. No data is available on their ability to occlude the canal. Factors including canal size, canal shape and cement viscosity may affect performance. The two aims of this study are (I) to determine the effect of cement viscosity, canal shape and canal size on the ability of cement restrictors to withstand cementation pressures, and (II) to determine which of the currently commercially available designs of cement restrictor is able to withstand cementation pressures, regardless of values of other potentially influential factors.

Methods: Artificial femoral canals were drilled in oak blocks. Circular canals had diameters of 12 or 17.5 mm. Oval canals had short axes equal to the diameter of the circular canals and long axes 1.3 times longer. This ellipticity of 1.3 is average for human femoral canals. One of four types of cement plugs (Hardinge, DePuy, UK; Exeter, Stryker, UK; Amber Flex, Summit Medical, UK; and OptiPlug, Scandimed, Sweden) was inserted. A pressure transducer was fitted in the canal just proximal to the plug. Bone cement (Palacos LV-40 low viscosity or Palacos R-20 high viscosity, both Schering Plough, UK) was prepared in a mixing device for 1 min at 21°C, and inserted in the artificial canal after 4 minutes. A materials testing machine was used to generate pressure in the cement. Cement pressure and plug position were measured. All combinations of canal size and shape, plug design and cement viscosity were pre-selected according to a D-optimal experimental design which was optimised to perform a four-way ANOVA to analyse the four main factors plus the interactions between plugs and the other three factors. A total of 23 experiments was performed.

Results: Average cement pressures achieved differed between implants (OptiPlug 448±66 kPa, Hardinge 142±66, Exeter 705±66, Amber Flex 475±72; p=0.002, all mean±SEM). They also differed between canal sizes (12 mm 529±49, 18 mm 356±47; p=0.03), canal shapes (Round 631±45, Oval 254±51; p=0.004) and cement viscosity (High 535±54, Low 350±43; p=0.03). No significant interaction between factors was found.

Discussion and Conclusion: All plugs resisted lower pressures in large canals, oval canals or with low viscosity cement. When comparing plugs, these different circumstances should therefore be taken into account. Of the four tested, the Exeter plug performed best in all adverse circumstances. The OptiPlug and AmberFlex, which are both resorbable, had an intermediate performance. The Hardinge plug performed worse.

Aim: To review the early complications associated with staged revision hip Arthroplasty utilising the Biomet antibiotic loaded cement spacer.

Method: We report on 80 consecutive staged revision hip replacements using the Biomet antibiotic loaded cement system in our institution over 3 years (1999–2002), performed by three consultant surgeons, with a minimum 1 year follow up.

Results: Our patients had an average age of 68 (range 48–90) years, with an equal sex distribution.

The median time between the first and second stage was 147 (range 50–619) days.

Fractures of the Biomet antibiotic loaded acrylic spacer occurred in 11% revisions when associated with an increase in time between stages and there was a 7% dislocation rate.

Patients did not receive a revision prosthesis in 19% cases and had early recurrent sepsis following their two stage procedure in 6%. Three patients had a single episode of dislocation of their revision hip prosthesis within a month postoperatively. Two patients had a proximal DVT and one patient had a pulmonary embolus. The mortality within eight weeks was 7%, rising to 10% within a year. This may be related to patient sepsis and comorbidities or the energy expenditure required to mobilise following a first stage procedure that we have analysed.

Conclusions: The risks of staged revision hip surgery for infection are substantial when considering the time involved, the energy expenditure required to mobilise following a first stage, the possibility of not achieving a revision hip prosthesis and the mortality rate.

The Biomet antibiotic loaded cement system articulates and maintains soft tissue length in the majority of patients for the duration required between stages.

Modern cementation techniques in hip arthroplasty are enhanced by the use of a cement restrictor. Failure of cemented hip replacements is commonly caused by aseptic loosening. Cement plugs which occlude the medullary canal are widely used to increase cementation pressures. Many plug types with variable performance exist. Ideally, plug performance should be sufficient regardless of other factors. All plug designs are circular in cross section, yet the vast majority of human femora are of oval section, the average ellipticity for human femora being 1.3. This study aims to determine (I) the effect of cement viscosity, canal shape and canal size on plug performance and (II) which designs of cement restrictor are able to withstand cementation pressures, regardless of values of other potentially influential factors.

Methods: Artificial femoral canals were drilled in oak blocks. Canals had diameters of 12 or 17.5 mm and oval or circular cross section. Four synthetic plug types (Hardinge, Exeter, Summit and OptiPlug.) and a bone plug (human allograft, Sulzer instrumentation) were tested. The effect of canal diameter, canal shape and low or regular cement viscosity was assessed.

Results: Maximal pressures achieved varied significantly between plugs. (OptiPlug 448±66 kPa, Hardinge 142±66, Exeter 705±66, Amber Flex 475±72, Bone plug 502±97 kPa; p=0.002, all mean±SEM). Al plugs performed worse in canals of increased size and of elliptical canal cross section (12 mm 529±49, 18 mm 356±47; p=0.03), canal shapes (Round 631±45, Oval 254±51; p=0.004). Cement viscosity had no statistical effect.

Discussion: Elliptical canal cross section and increased canal diameter adversely affects performance of all plug designs. Of the five tested, the Exeter and bone plugs performed best in all adverse circumstances. The Opti-Plug and AmberFlex, which are both resorbable, had an intermediate performance. The Hardinge plug performed worse. These factors should be considered when selecting plug design.

Introduction: Impaction grafting has become a popular technique to revise implants. The Norwegian Arthroplasty Registry reports its use for a third of all revisions. Yet, the technique is seen as demanding. A particular challenge is to achieve sufficient mechanical stability of the construction. This work tests two hypotheses: (1) Graft compaction is an important determinant of mechanical stability, and (2) Graft compaction depends on compaction effort and graft properties. Methods: Impaction grafting surgery was simulated in laboratory experiments using artificial bones with realistic elastic properties (Sawbones, Malmö, Sweden). Bone stock was restored with compacted morsellised graft, and the joint reconstructed with a cemented implant. The implant was loaded cyclically and its migration relative to bone measured. In a second study, morsellised bone of various particle sizes and bone densities, with or without added ceramic bone substitutes, was compacted into a cylindrical mould by impaction of a plunger by a dropping weight. Plunger displacement was measured continuously. Results: Initial mechanical stability of the prostheses correlated most strongly with degree of graft compaction achieved. Graft compaction to similar strength was achieved with less energy for morsellised bone with larger particles, higher density, or bone mixed with ceramic substitutes. Conclusion: Initial mechanical stability of impaction-grafted joint reconstructions depends largely on degree of graft compaction achieved by the surgeon. Compaction depends partly on the vigour of impaction, and partly on graft quality. Higher bone density, larger particle size and mixing with ceramic particles all help to facilitate graft compaction, giving a stronger compacted mass with less effort.