Summary Statement

This project proves that Patellofemoral (PF) joint degeneration is not a contraindication to medial unicompartmental knee replacement.

Introduction

Unicompartmental knee arthroplasty (UKA) is a recognized procedure for treatment of medial compartment osteoarthritis. Patellofemoral (PF) joint degeneration is widely considered to be a contraindication to medial unicompartmental knee replacement. We examined the validity of this preconception using information gathered prospectively on consecutive patients who underwent UKA using the Repicci II® UKA prosthesis for medial compartment osteoarthritis.

The advantages of unicompartmental knee arthroplasty (UKA) include its bone preserving nature, lower relative cost and superior functional results. Some temporary pain has been reported clinically following this procedure. Could this be related to bone remodeling? A validated bone remodeling algorithm may have the answers…

A 3D geometry of an intact human cadaveric tibia was generated using CT images. An all poly unicompartmental implant geometry was positioned in an inlay and onlay configuration on the tibia and the post-operative models created. An adaptive bone remodeling algorithm was used with finite element modeling to predict the bone remodeling behavior surrounding the implant in both scenarios. Virtual DEXA images were generated from the model and bone mineral density (BMD) was measured in regions of interest in the AP and ML planes. BMD results were compared to clinical results.

The bone remodelling algorithm predicted BMD growth in the proximal anterior regions of the tibia, with an inward tendency for both inlay and onlay models. Looking in the AP plane, a maximum of up to 7% BMD growth was predicted and in the ML plane this was as high as 16%. Minimal BMD loss was observed, which suggests minimal disturbance to the natural bone growth following UKA.

Positron emission tomography (PET) scans showed active hot spots in the antero- medial regions of the tibia. These results were consistent with the finite element modeling results.

Bone remodeling behavior was found to be sensitive to sizing and positioning of the implant.

The adaptive bone remodeling algorithm predicted minimal BMD loss and some BMD growth in the anterior region of the tibia following UKA. This is consistent with patient complaint and PET scans.

Repair of chronic Achilles tendon rupture is technically complex. Flexor jallucis longus (FHL) and peroneus brevis (PB) tendon transfers have been described, but the mechanical properties of these tissues have not been well reported.

The FHL, PB and tendo achilles (TA) tendons were harvested from 17 fresh frozen human cadavers free of gross pathology (mean age 69 years). Samples were tested in uniaxial tension at 100% per minute. Samples were secured using special jigs for the bony aspect or by freezing the tendons in cryogrips using liquid carbon dioxide. The peak load (N), linear stiffness (N/mm) and energy to peak load (N*mm) were determined. Mechanical data was analysed using one way analysis of variance (ANOVA) followed by a Games Howell multiple comparison post-hoc test.

Fifty one tendons were harvested and mechanical testing was successfully completed in all samples apart from one PB that slipped from the grips during testing (sample was omitted from the analysis). The mean ultimate loads differed for each group, with the TA tendons being the strongest (1724.5 N ± 514.3) followed by FHL (511.0 N ± 164.3) and PB (333.1 N ± 137.2) (P< 0.05). Similar results were found with respect to energy, with TA tendons absorbing the most energy followed by FHL and PB (P< 0.05). Stiffness for the TA tendons (175.5 N/mm ± 94.8) was greater than FHL (43.3 N/mm ± 14.1) and PB (43.6 N/mm ± 18.9), which did not differ from each other.

FHL is stronger than PB, but have similar stiffness. The mechanical properties of PB and FHL were both inferior to TA. Graft stiffness appears to be an important variable rather than ultimate load based on the clinical success of both techniques.

Introduction: The Ionising Radiations Medical Exposure Regulations Act 2000 has established diagnostic reference levels for radiological examinations, however at present there are no national guidelines available for orthopaedic trauma surgery. There may be a role for the introduction of diagnostic reference levels at a local level therefore we studied patient area dose and screening time for orthopaedic trauma operations performed in the Regional Trauma Centre in Northern Ireland.

Methods: Retrospectively data was retrieved from written radiography records in the Royal Victoria Hospital, during the period of 1st January 2007 to 31st December 2007 for all orthopaedic trauma cases in which an image intensifier was used. The screening time, patient area dose (PAD), details of the operation, patient age, sex, month of the operation and grade of the operating surgeon (trainee or consultant), were recorded.

Results: 1709 cases using image intensifier were reviewed. 137 cases were excluded due to incomplete data. 319 hips were screened for insertion of sliding hip screw, mean screening time was 0.51min with a mean PAD of 145cGycm2. 127 femoral nails were inserted with an average screening time of 1.84min and mean PAD of 310cGycm2. 166 tibias were screened for application of Ilizarov frame or insertion of tibial nail, average screening time was 3.00min with a mean PAD of 48cGycm2. 129 spinal cases were screened with an average screening time of 0.80mins and mean PAD of 37.9cGycm2. Consultants had lower screening times and mean PADs than trainees with 0.63min versus 1.01min and 65.8cGycm2 versus 70.9cGycm2.

Conclusions: The average screening times and mean PADs compared favourably with local reference guides for image intensifier cases and with other published series. Every trauma unit should have local reference ranges for orthopaedic imaging cases and regular audit should be carried out to assess compliance with these ranges.

Introduction: Measurements taken from digital radiographs and corrected for the labelled scale do not correspond to actual clinical dimensions accurately.

Methods: The routine radiographs taken of 30 hip hemi-arthoplasty patients were used to make measurements in the region of the hip. 15 patients were compared in two hospitals using different digital imaging processing. The images were scaled to 100% (actual size), three blinded measurements were taken of the prosthetic femoral head and the average calculated. These results were compared to the actual size of the implanted prosthesis. Further analysis was carried out to determine true magnification.

Results: A consistent difference was observed between measured dimensions at 100% scale and known actual prosthesis dimensions. It was identified that the magnification factor ranged from 110 to 115% with at average magnification factor of 113%. No significant difference was observed between the two hospital systems.

Discussion: Many factors influence radiographic magnification, including variations in distances from x-ray source to patient and from x-ray source to x-ray plate. Angulation and degree of divergence of the x-ray beam and rotation of a limb also influence the magnification which occurs. Accurate measurements can only be taken from digital x-rays if appropriate calibration is performed.

Introduction: The Ionising Radiations Medical Exposure Regulations Act 2000 has established diagnostic reference levels for radiological examinations, however at present there are no national guidelines available for orthopaedic trauma surgery. There may be a role for the introduction of diagnostic reference levels at a local level therefore we studied patient area dose and screening time for orthopaedic trauma operations performed in the Regional Trauma Centre in Northern Ireland.

Methods: Retrospectively data was retrieved from written radiography records in the Royal Victoria Hospital, during the period of 1st January 2007 to 31st December 2007 for all orthopaedic trauma cases in which an image intensifier was used. The screening time, patient area dose (PAD), details of the operation, patient age, sex, month of the operation and grade of the operating surgeon (trainee or consultant), were recorded.

Results: 1709 cases using image intensifier were reviewed. 137 cases were excluded due to incomplete data. 319 hips were screened for insertion of sliding hip screw, mean screening time was 0.51min with a mean PAD of 145cGycm2. 127 femoral nails were inserted with an average screening time of 1.84min and mean PAD of 310 cGycm2. 166 tibias were screened for application of Ilizarov frame or insertion of tibial nail, average screening time was 3.00min with a mean PAD of 48cGycm2. 129 spinal cases were screened with an average screening time of 0.80mins and mean PAD of 37.9cGycm2. Consultants had lower screening times and mean PADs than trainees with 0.63min versus 1.01min and 65.8cGycm2 versus 70.9cGycm2.

Conclusions: The average screening times and mean PADs compared favourably with local reference guides for image intensifier cases and with other published series. Every trauma unit should have local reference ranges for orthopaedic imaging cases and regular audit should be carried out to assess compliance with these ranges.

Introduction: Following the introduction of a hip hemiarthroplasty monobloc a number of unusual problems occurred in our unit. These included the need for narrow stem prostheses, on table revision for excessive anteversion and intra-operative fracture, revision for irreducible dislocation and excision arthroplasty. It was noted that the trial component fitted easily whereas there were problems with insertion of the actual prostheses.

Methods: To investigate the issue of differences in the trial component versus the actual prosthesis the following methods were employed. Digital micrometer measurements were taken from the trial and real components. Differences in the antero-posterior and medial-lateral dimensions of the components were analysed. Cross-sectional area analysis was also performed. The effect of femoral canal diameter on the insertion depth of trial and component was investigated using a canal model with digital image processing. The changes in cross-sectional geometry were investigated with respect to canal width, insertion depth and the consequential effect of component version.

Results: It was shown that the trial was smaller than the actual prosthesis, particularly in the medial-lateral plane. No significant difference was observed between real components. A critical region of the component was identified which corresponded to a canal width of 10mm. Components not fully seated at this point drifted in to excessive anteversion with impaction.

Discussion: This study identified that an unacceptable difference existed between the trial and real prosthesis. The study explained the effect of a narrow canal in terms of increased complications which resulted in increased morbidity for patients. Consequently the use of this implant was abandoned in narrow femoral canal cases and an alternative implant has been selected for use in fracture patients requiring hemiarthroplasty in our unit.

Introduction: Measurements taken from digital radiographs and corrected for the labelled scale do not correspond to actual clinical dimensions accurately.

Methods: The routine radiographs taken of 30 hip hemiarthoplasty patients were used to make measurements in the region of the hip. 15 patients were compared in two hospitals using different digital imaging processing. The images were scaled to 100% (actual size), three blinded measurements were taken of the prosthetic femoral head and the average calculated. These results were compared to the actual size of the implanted prosthesis. Further analysis was carried out to determine true magnification.

Results: A consistent difference was observed between measured dimensions at 100% scale and known actual prosthesis dimensions. It was identified that the magnification factor ranged from 110 to 115% with at average magnification factor of 113%. No significant difference was observed between the two hospital systems.

Discussion: Many factors influence radiographic magnification, including variations in distances from x-ray source to patient and from x-ray source to x-ray plate. Angulation and degree of divergence of the x-ray beam and rotation of a limb also influence the magnification which occurs. Accurate measurements can only be taken from digital x-rays if appropriate calibration is performed.

Introduction: Many minor procedures are performed daily in fracture clinic and often carried out under intravenous sedation and analgesia. These occur between 9 am and 8 pm.

Methods: All minor and intermediate cases logged in the theatre register at the fracture clinic were reviewed from February 2007 to February 2008. The following items were recorded: month, age of patient, diagnosis, procedure, grade of doctor or nurse performing the procedure and the method of anaesthesia and the time of day the procedure was carried out. Specifically the use of intravenous sedation, intravenous morphine, use of Entonox, and local or regional anaesthetic techniques were recorded. Records were classified by the time of day the procedure occurred – any procedure starting at or after 1730 hours was regarded as an ‘out-of-hours’ procedure.

Results: The age range of patients undergoing procedures ranged from 13 to 100 years, with a mean age of 49 years. 576 procedures were performed, 529 by SHOs, 35 by registrars, 4 by consultants and 8 by nursing staff. The commonest procedure was distal radius manipulation under anaesthesia (MUA) with 227 cases, followed by ankles with 175 cases. Intravenous sedation was used in 473 cases with 95 of these occurring ‘out-of-hours’. Entonox was used in 33 cases. A regional or local anaesthetic technique was used in 38 cases. No anaesthetic or analgesia was required in 10 cases.

Discussion: The ability to perform minor and intermediate procedures is dependent on being to provide safe and appropriate anaesthesia. If the current procedures could not be performed in fracture clinic this would put significant pressures on operating theatre time and would have inherent cost implications. However, in order to provide safe and appropriate anaesthesia for these procedures adequate and contempory staff training is required, especially when intravenous sedation is utilised.

Treatment of chronic Achilles tendon ruptures can be technically demanding due to tendon retraction, atrophy and short distal stumps. Although rare, re-rupture following surgical treatment is a major late complication.

Biomechanical studies on the strength of reconstructed Achilles tendon using autologous tendon grafts have not been well documented.

This study examined the time zero in vitro mechanical properties of a reconstructed Achilles tendon (TA) using the peroneus brevis (PB) or the flexor hallucis longus (FHL) tendons in a human cadaver model (n=17).

The TA was reconstructed using the same technique for all specimens. Biomechanical testing was performed using an MTS 858 Bionix testing machine and structural properties (failure load, stiffness and mode of failure) were determined.

Average failure load was significantly higher in the PB-group (p=0.0116) (PB: 343.82 N (+/− 124.90 N, FHL: 241.54 N (+/− 82.17 N)). There was no significant difference in stiffness (p=0.212), (PB: 16.53 N/mm (+/− 6.25 N/mm), FHL: 14.00 N/mm (+/− 3.84 N/mm)) or energy (p=0.075).

Mode of failure was the same for all specimens, with the tendon graft cutting through either the distal or proximal TA-stump. Reinforcement of these stumps could lead to increased failure loads. Based on the biomechanical data, the present study supports the use of either FHL or PB to reconstruction chronic TA tendon ruptures. The greater failures loads for PB may not be clinically relevant considering the peak loads. The addition of the suturing pattern, whilst is does reconstruct the tendon, does not provide a similar ability to resist the load.

Introduction: Repair of chronic Achilles tendon rupture is technically complex. Flexor Hallucis Longus (FHL) and Peroneus Brevis (PB) Tendon transfers have been described, but the mechanical properties of these tissues have not been well reported.

Methods: The FHL, PB and tendo Achilles (TA) tendons were harvested from 17 fresh frozen human cadavers free of gross pathology (mean age 69 years). Samples were tested in uniaxial tension at 100% per minute. Samples were secured using special jigs for the bony aspect or by freezing the tendons in cryogrips using liquid carbon dioxide. The peak load (N), linear stiffness (N/mm) and energy to peak load (N*mm) were determined. Mechanical data was analysed using one way analysis of variance (ANOVA) followed by a Games Howell multiple comparison post-hoc test.

Results: 51 tendons were harvested. Mechanical testing was successfully completed in all samples apart from one PB that slipped from the grips during testing (sample was omitted from the analysis). The mean ultimate loads differed for each group, with the TA tendons being the strongest (1724.5 N ± 514.3) followed by FHL (511.0 N ± 164.3) and PB (333.1 N ± 137.2) (P< 0.05). Similar results were found with respect to energy, with TA tendons absorbing the most energy followed by FHL and PB (P< 0.05). Stiffness for the TA tendons (175.5 N/mm ± 94.8) was greater than FHL (43.3 N/mm ± 14.1) and PB (43.6 N/mm ± 18.9), which did not differ from each other.

Conclusions: FHL is stronger than PB, but have similar stiffness. The mechanical properties of PB and FHL were both inferior to TA. Graft stiffness appears to be an important variable rather than ultimate load based on the clinical success of both techniques.

Introduction: The curing of polymethylmethacrylate cement is an exothermic reaction, with temperatures reaching 80oC. Thus contact with cement can result in thermal injury. During orthopaedic procedures bone cement may be inadvertently left lying on surgical drapes. This study aims to investigate the effect of drapes on the contact temperature of such cement.

Methods: The experiment was carried out in an orthopaedic theatre with a constant temperature and humidity. Polymethylmethacrylate cement was prepared and pieces of cement of a specific size were investigated. Experiment one measured the contact temperature of curing cement in isolation. Experiment two and three investigated the effect of one and two layers of surgical drape respectively on contact temperature.

Results: The mean maximum temperature reached in experiments one, two and three was 75oC, 71oC and 69oC respectively. The maximum temperature was reached quickest in experiment three. In considering the time period that the contact temperature remained above 40oC this was longest in experiment three at 750s followed by experiment 2 and then experiment 1 at 525s and 330s respectively. Contact temperature graphs for each experiment have been plotted in figure.1

Discussion: One might think that drapes provide a protective barrier between a patient’s skin and bone cement, however this is incorrect. The drapes insulate the exothermic reaction of curing bone cement which tends to accelerate the reaction. Furthermore the insulating effect of drapes causes curing bone cement to lose heat at a slower rate resulting in the contact temperature remaining above 40oC for longer; this increases the potential for thermal injury. This study demonstrates that a specific effort should be made to ensure bone cement is not inadvertently left lying on surgical drapes.

Introduction: In patients presenting with an infected hip arthroplasty first-line antibiotics at our institution are Flucloxacillin and Fucidic Acid. It was observed that many patients needed their antibiotic regime changed once sensitivities became known. This was because of resistance to the first-line antimicrobial agents. This raised the question ‘How frequent is Flucloxacillin and Fucidic acid resistance in prosthetic hip infection and is there a more appropriate first-line antibiotic?’

Method: A computerised search of the Belfast Orthopaedic Information System (BOIS) identified all cases of infected hip arthroplasty from October 2001 to April 2005. All microbiology results of these patients were obtained and analysed to determine the infecting organism, sensitivity and resistance to Flucloxacillin, Fucidic acid and Teicoplanin.

Results: BOIS identified forty cases of infected hip arthroplasty, 35 primary arthroplasties and 5 revision arthroplasties. Of the 40 patients 36 had positive cultures, 3 had no growth and there was missing data on one patient. Flucloxacillin sensitivity occurred in 25% of patients with 58% resistance. With regard to Fucidic Acid sensitivity was 47 % with 44% resistance. There were no cases of Teicoplanin resistance. Teicoplanin sensitivity occurred in 78% of patients and with analysis by organism susceptibility sensitivity would measure 92% (8% were not tested against Teicoplanin).

Discussion: Flucloxacillin resistance is greater than sensitivity in infected hip arthroplasty in this study group. Fucidic Acid sensitivity and resistance are roughly equal. There are no cases of Teicoplanin resistance in our study population and with analysis with respect to the causative organism Teicoplanin sensitivity reaches 92%. It could be suggested therefore that the first-line agents of Flucloxacillin and Fucidic acid are inappropriate and that Teicoplanin may be a better choice.

Introduction and Aims: While clinical variables are considered important risk factors for post-arthroplasty VTE, the role of common genetic thrombophilic factors is less clear. The aims of this study were to determine if common thrombophilic genetic polymorphisms are independent risk factors for VTE post-arthroplasty; and if clinical variables are equally or more important.

Method: A prospective study of consecutive patients undergoing elective total hip or knee arthroplasty at a single institution, involving two surgeons. Patients were interviewed to assess clinical risks. Pre-operative blood samples were taken for Factor V Leiden (FVL), Pro-thrombin G20210A (PTH) and Methylenetetrahydrofolate reductase C677T (MTHFR) testing. All patients received routine enoxaparin prophylaxis and compression stockings. Intermittent pneumatic calf compression was also used by one surgeon. Presence of DVT was assessed using bilateral lower limb duplex ultrasonography (seven ± two days post-operatively) in all patients and performed in a vascular laboratory. Symptoms suggestive of pulmonary embolism were investigated by ventilation/perfusion lung scanning.

Results: A total of 569 patients were recruited with a median age of 67 years (range 20–90). Osteoarthritis was the main surgical indication. The overall incidence of post-operative venous thromboembolism (VTE) was 26%. Of thromboembolic events, 15% VTE were proximal DVT; 84% VTE were distal DVT and only one percent were pulmonary emboli. Prevalence of the thrombophilic genotypes was: 4.6% (heterozygous FVL mutation); 2.1% (heterozygous PTH); and 10.4% (homozygous C677T MTHFR mutation). Using univariate analysis, older age (p < 0.0005), total knee arthroplasty (p < 0.0005), recent surgery (p = 0.002), general anaesthesia (p = 0.013), operation time in minutes (p < 0.0005) and use of blood transfusions (p < 0.0005) were significantly associated with post-operative DVT. None of the thrombophilic genotypes were found to be significantly associated with post-operative DVT, however the frequency of FVL and PTH was highest in patients with proximal DVT and total hip arthroplasty patients with DVT. In multivariate analysis of both genetic and clinical thrombophilic factors, only age (p=0.02) and total knee arthroplasty (p< 0.0005) were found to be significant independent risk factors for post-operative VTE.

Conclusion: We conclude that clinical factors such as age and type of surgery (total knee arthroplasty) are independent risks for post-operative VTE in patients undergoing lower limb arthroplasty. FVL, PTH and MTHFR are not significant risk factors for post-operative VTE and screening for these mutations is not indicated.

Introduction The outcome of conventional treatment of isolated end-stage patello-femoral arthritis is unpredictable. Encouraging success rates have been documented in the literature with the use of patello-femoral arthroplasty. A prospective study was performed to review the early outcome of patello-femoral arthroplasties at St Vincent’s Private Hospital.

Methods The prospective study involved 32 patients over a period from 1999 to 2002. There were eight males and 24 females, with a mean age at surgery of 65 years. The mean follow up was 1.6 years (0.3 – 4 years). The patients were scored pre-operatively and at each post-operative visit using the Knee Society Score. Subjective results also recorded at follow-up included 1) the patient’s satisfaction in regard to pain and function, 2) whether expectations were met from the surgery and 3) if the patient would have the same operation on the other limb if the same symptoms existed. There were 31 knees available for post-operative analysis.

Results The mean knee score pre-operatively was 65.9 and post-operatively 84.3. The mean functional score pre-operatively was 75.7 and post-operatively 83.2. The mean pre-operative total knee score was 141.6 and improved to 167.5 post-operatively. The descriptive post operative Total Knee Score equates to 26 (83.9%) patients with excellent or good results and five (16.1%) patients with fair or poor results. Subjectively 87.2% of patients had an improvement in pain, 83.9% had an improvement in function, 80.7% had their expectations met and 77.5 % would consider the surgery again if required on the other leg.

Conclusions This review of the early experience of patello-femoral arthroplasty has shown patello-femoral arthroplasty to be a viable treatment option in the short term for end stage patello-femoral arthritis.

Introduction Unicompartmental knee arthroplasty (UKA) is well established in the treatment of OA of the knee, but has not been performed in large numbers compared with total knee arthroplasty. However, with the development of minimally invasive surgery, numbers of procedures are increasing rapidly. This study examines the results of minimally invasive UKA performed by one surgeon since august 1998.

Methods A consecutive series of 388 knees in 360 patients operated between August 1988 and February 2003 were evaluated using a prospective database. All surgeries were performed by the senior author using a minimally invasive technique in a day surgery unit. General anaesthesia was used in all cases with local anaesthesia intra-operatively, combined with an anaesthetic infusion pump. No patient received parenteral narcotics. Ninety-seven percent were medial and three percent were lateral arthroplasties.

Results Post-operatively no patients were lost to follow-up which ranged from two months to 4.5 years. Average age was 66 years. The average IKS score improved from 75 to 158 post-operatively. Most patients retained their pre-operative range of motion which averaged 120°. Average length of stay was 1.57 days with 41% of patients discharged the same day. There were five failures, due in part to osteoporosis and overcorretion. These were revised successfully to a ‘primary’ type knee prosthesis. Satisfaction rate subjectively was 98%.

Conclusions Mid term results of UKA using the Repicci technique of minimally invasive surgery with rapid mobilisation and early discharge has ahcieved excellent results for unicompartmental OA of the knee. The procedure is better tolerated with a low complication rate and higher patient satisfaction than total knee replacement in this gourp of patients.